University of Southern California Dissertations and Theses

The impact of family structure and family process on adolescent sexual behavior: race and gender variations

(USC Thesis Other)

The impact of family structure and family process on adolescent sexual behavior: race and gender variations

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THE IMPACT OF FAMILY STRUCTURE AND FAMILY PROCESS
ON ADOLESCENT SEXUAL BEHAVIOR: RACE AND GENDER
VARIATIONS
by
Thomas I-Ping Lo
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(SOCIOLOGY)
December 2007
Copyright 2007 Thomas I-Ping Lo
ii
Acknowledgements
Many people have supported me in some way during the dissertation process.
First, my greatest gratitude goes to my dissertation chair, Dr. Harry “Edward”
Ransford. He has dedicated a tremendous amount of his time and energy to help
make this dissertation possible. This dissertation truly would not have been possible
without his valuable assistance.
Second, I would like to thank the rest of my dissertation committee members,
Dr. Timothy Biblarz and Dr. Merril Silverstein. I appreciate all of their critiques for
this study. In addition, I am truly grateful about their overwhelming emotional
support for this project.
Third, I would like to thank my father, Tain Lai Lo, my mother, I Ching Lo,
and my two sisters, Lucy and Mary Lo for emotionally supporting me during my
graduate studies and for sacrificing some of their time so that I could pursue my
educational goals.
Finally, I would like to thank my colleagues Don Naylor and Maria Elena
Ruiz, and my friends, in no particular order, William T. Liu, Leon Liao, Michael S.
Lee, William Liao, and Wen Hung Jen for being supportive of my endeavors. They
always believed that I could accomplish my goals. I am grateful that they were
always there for emotional support while during my graduate studies.
iii
Table of Contents
Acknowledgements ii
List of Tables vi
List of Figures ix
Abstract x
Chapter 1: Introduction 1
Family Structure vs. Family Process Debate 2
Dissertation Objectives 5
Brief Overview of the Explanations 11
Family Transitions 14
Parental Supervision 15
Parental Attitudes 16
Family Support 17
Peer Influence 17
Functional and Conflict Perspectives on Socialization 18
Study Questions 19
Study Significance 20
Preview of Chapters 26
Chapter Summary 27
Chapter 2: Theories 30
Family Transition Explanation 30
Parental Supervision Explanation 37
Parental Attitudes Explanation 42
Family Support Explanation 46
Peer Influence Explanation 50
Chapter 3: Methodology 53
Description of the Dataset 53
Study Sample 54
Study Measures: Dependent Variable 55
Study Measures: Independent Variables 55
Demographic Variables 56
Family Transition Measure 58
Parental Supervision Measure 59
Parental Attitude Measure 61
Family Support Measure 63
Peer Influence Measure 63
iv
Data Analysis 64
Statistical Explanations 64
Chapter 4: Total Sample Analyses 67
Hypotheses 67
Descriptive Results 68
Test of Hypotheses Results 74
Discussion 78
Unwedded Mother Household Discussion 80
Divorced/Separated Mother Household Discussion 81
Stepparent Household Discussion 83
Relative Importance of the Five Explanations Discussion 84
Chapter Summary 90
Chapter 5: Adolescent Sexual Activity By Gender and Race: 93
Theories and Rationales
Gender Differences: Rationales 93
Race Differences: Black Adolescents 96
Race Differences: Hispanic Adolescents 100
Race Differences: Asian Adolescents 101
Family Transitions: Gender & Race Differences 102
Parental Supervision: Gender & Race Differences 105
Parental Attitude: Gender & Race Differences 107
Family Support: Gender & Race Differences 111
Peer Influence: Gender & Race Differences 113
Chapter 6: Gender Context Analyses 115
Hypotheses 115
Descriptive Results 116
Test of Hypotheses Results 118
Discussion 125
Chapter Summary 129
Chapter 7: Racial Context Analyses 133
Hypotheses 134
Descriptive Results 135
Test of Hypotheses Results 137
Discussion 149
Non-Traditional Household Discussion 149
Family and Peer Influence Variables Discussion 153
Relative Importance of the Five Explanations Discussion 156
Chapter Summary 158
v
Chapter 8: Race-Gender Context Analyses 162
Hypotheses 163
Descriptive Results 164
Test of Hypotheses Results 175
Discussion 196
Non-traditional Households: White Adolescents 196
Non-traditional Households: Black Adolescents 199
Non-traditional Households: Hispanic Adolescents 201
Summary of Non-Traditional Households Discussion 202
Family and Peer Influence Variables Discussion 204
Relative Importance of the Five Explanation Discussion 209
Chapter Summary 211
Chapter 9: Hierarchically Ordered Regression Analyses 214
Analyses 215
Discussion 247
Chapter Summary 251
Chapter 10: Conclusion 253
Answers to Questions Posed in the Introduction 254
Some Significant Findings from this Study 270
Significant Family Structure Findings 271
Significant Family and Peer Influence Variables Findings 276
Study Limitations and Issues for Further Research 278
Research Implications 283
Social Policy and Counseling Implications 284
Final Comments 288
References 290
vi
List of Tables
Table 4-1: Mean Scores for the Demographic, Family Influence, and 69
Peer Influence Variables By Adolescent Sexual Status
Table 4-2: Mean Scores for the Adolescent Sexual Activity, Demographic, 70
Family Influence, and Peer Influence Variables By Household
Status
Table 4-3: Logistic Regressions of Sexual Activity on the Demographic, 75
Family Influence, and Peer Influence Variables for the Total
Adolescent Sample
Table 6-1: Mean Scores for the Sexual Activity, Demographic, Family 117
Influence, and Peer Influence Variables By Adolescent’s
Gender
Table 6-2: Logistic Regression of Adolescent Sexual Activity on the 119
Demographic, Family Transition, and Parental Supervision
Variables By Gender
Table 6-3: Logistic Regression of Adolescent Sexual Activity on the 121
Demographic, Parental Attitudes, Family Support, and Peer
Influence Variables By Gender
Table 6-4: Summary of the Family Structure and Adolescent Sexual 124
Activity Relationship By Gender
Table 7-1: Mean Scores for the Adolescent Sexual Activity, Demographic, 136
Family Influence, and Peer Influence Variables Within Race
Categories
Table 7-2: Logistic Regression of Adolescent Sexual Activity on the 138
Demographic Variables Within Race Categories
Table 7-3: Logistic Regression of Adolescent Sexual Activity on the 140
Demographic and Family Transition Variables Within Race
Categories
Table 7-4: Logistic Regression of Adolescent Sexual Activity on the 141
Demographic and Parental Supervision Variables Within Race
Categories
vii
Table 7-5: Logistic Regression of Adolescent Sexual Activity on the 144
Demographic and Parental Attitude Variables Within Race
Categories
Table 7-6: Logistic Regression of Adolescent Sexual Activity on the 145
Demographic and Family Support Variables Within Race
Categories
Table 7-7: Logistic Regression of Adolescent Sexual Activity on the 146
Demographic and Peer Influence Variable Within Race
Categories
Table 7-8: Summary of the Family Structure and Adolescent Sexual 148
Activity Relationship Within Race Categories
Table 8-1: Sexual Activity, Demographic, Family Influence, and Peer 165
Influence Variables Mean Scores For White Adolescents
Table 8-2: Sexual Activity, Demographic, Family Influence, and Peer 166
Influence Variables Mean Scores For Black Adolescents
Table 8-3: Sexual Activity, Demographic, Family Influence, and Peer 167
Influence Variables Mean Scores For Hispanic and Asian
Adolescents
Table 8-4: Mean Scores for the Sexual Activity, Demographic, Family 170
Influence, and Peer Influence Variables Among Male
Adolescents
Table 8-5: Mean Scores for the Sexual Activity, Demographic, Family 173
Influence, and Peer Influence Variables Among Female
Adolescents
Table 8-6: Logistic Regression of Adolescent Sexual Activity on the 176
Demographic Variables Within Race-Gender Categories
Table 8-7: Logistic Regression of Adolescent Sexual Activity on the 179
Family Transition Variable Within Race-Gender Categories
Table 8-8: Logistic Regression of Adolescent Sexual Activity on the 181
Parental Supervision Variables Within Race-Gender
Categories
Table 8-9: Logistic Regression of Adolescent Sexual Activity on the 186
Parental Attitude Variables Within Race-Gender Categories
viii
Table 8-10: Logistic Regression of Adolescent Sexual Activity on the 188
Family Support Variable Within Race-Gender Categories
Table 8-11: Logistic Regression of Adolescent Sexual Activity on the Peer 191
Influence Variable Within Race-Gender Categories
Table 8-12: Summary of the Family Structure and Adolescent Sexual 193
Activity Relationship Within Race-Gender Categories
Table 9-1: Hierarchical Ordered Regression of Sexual Activity on the 216
Demographic, Family Influence, and Peer Influence Variables
for the Total Sample
Table 9-2: Hierarchical Ordered Regression of Adolescent Sexual Activity 220
on the Demographic, Family Influence, and Peer Influence
Variables for Males
Table 9-3: Hierarchical Ordered Regression of Adolescent Sexual Activity 223
on the Demographic, Family Influence, and Peer Influence
Variables for Females
Table 9-4: Hierarchical Ordered Regressions of Sexual Activity on the 226
Demographic, Family Influence, and Peer Influence Variables
for White Adolescents
Table 9-5: Hierarchical Ordered Regression of Sexual Activity on the 230
Demographic, Family Influence, and Peer Influence Variables
for Black Adolescents
Table 9-6: Hierarchical Ordered Regression of Adolescent Sexual Activity 233
on the Demographic, Family Influence, and Peer Influence
Variables for White Males
Table 9-7: Hierarchical Ordered Regression of Adolescent Sexual Activity 237
on the Demographic, Family Influence, and Peer Influence
Variables for Black Males
Table 9-8: Hierarchical Ordered Regression of Adolescent Sexual Activity 240
on the Demographic, Family Influence, and Peer Influence
Variables for White Females
Table 9-9: Hierarchical Ordered Regression of Adolescent Sexual Activity 244
on the Demographic, Family Influence, and Peer Influence
Variables for Black Females
ix
List of Figures
Figure 1-1: Proposed Model for the Relationship Between Family 6
Structure and Adolescent Sexual Activity
Figure 1-2: Proposed Hierarchically Ordered Regression Model Between 12
the Five Explanations and Adolescent Sexual Activity
x
Abstract
Using the 1994-1995 National Study of Adolescent Health dataset, I
conducted two investigations on the impact of family structure and family influences
on sexual activity for white, black, Hispanic, and Asian adolescents. The first
investigation determined the best explanation (family transitions, parental
supervision, parental attitudes, family support, and peer influence) for the
relationships between non-traditional households (unwedded mother,
divorced/separated mother, and stepparent households) and sexual activity. The
second investigation determined which of these explanations are either directly or
indirectly associated with sexual activity. These analyses were conducted on the
total sample and within gender, race, and race-gender subgroups.
Initially, parental attitudes, parental supervision, and family transition
explanations best mediated the unwedded, divorced/separated, and stepparent
household effects respectively, although they only provided a partial explanation for
the relationship. Within gender subgroups, similar results were found except that the
stepparent household effect for males is fully explained by the family transition
explanation. Within adolescent’s race and race-gender subgroups, the best
explanation varied by family configuration and subgroup contexts.
Experiencing family transitions, mother-child activities, mother-child
problem and sex discussions, parental meal presence and strictness, mother’s
religiosity and disapproval of adolescent sexuality, family support, and acquaintance
with deviant peers were associated with sexual activity for the total sample. Gender
xi
analyses revealed similar results except that the family support effect for females is
significantly stronger than for males.
The second investigation found that mother-child problem discussions and
sex discussions, mother’s disapproval of adolescent sexuality, family support, and
acquaintance with deviant peers are directly associated with adolescent activity while
family transition, mother’s religiosity, mother-child activities, and parental strictness
are indirectly associated with sexual activity. Indirect relationships were only found
for white males and females.
Important findings include 1) unwedded mother and two biological parent
households share many similar characteristics, 2) contrary to the literature, sexual
activity risk was no greater for some non-traditional households than two parent
households and 3) parental monitoring variables were not while parental interaction
variables were associated with sexual activity. Policy and counseling
recommendations are suggested based on these findings. This study demonstrated
that the family structure- adolescent sexual activity relationship is highly
contextualized for each non-traditional household.
1
Chapter 1: Introduction
According to several studies, the percentage of sexually experienced
adolescents living in the United States increased significantly between the early
1970s to the mid 1990s. For example, Moore (1992) found that the percentage of 15
to 19 year old sexually experienced females increased from 29% in 1970 to 52% in
1988. In addition, Parrillo, Felts, & Mikow-Porto (1997) found that approximately
half of the adolescents during the early 1990s had engaged in sexual activity by the
time they reached 12
th
grade. Furthermore, Laumann et al. (1994) found that the age
at which adolescents first engage in sexual activity had also declined during this
period. However, researchers have found that these trends have been trending
downward during the first few years of the 21
st
century (Papillo et al, 2003).
Many adults, such as parents, educators, health care professionals, and
counselors are concerned about both of these trends because sexually active
adolescents may experience severe social and economic consequences that can
diminish the prospect of a healthy and productive adulthood. Although many
sexually active adolescents have managed to avoid these negative consequences,
some do not. The social consequences of engaging in sexual activity during puberty
include acquiring a sexually transmitted disease among adolescents of both genders
(Hayes, 1987), early pregnancy for females (Hayes, 1987; Brooks-Gunn &
Furstenberg, 1989), and early fatherhood for males (Sonnenstein, 1986).
Adolescents who become pregnant or experience early fatherhood are also more
likely than those who do not encounter these events to drop out of high school (Card
2
& Wise, 1978; Hayes, 1987; Brooks-Gunn & Furstenberg, 1989), to have limited
employment opportunities, and to receive lower wages throughout one’s
employment history (Card & Wise, 1978; Hofferth & Moore, 1979).
The Family Structure vs. Family Process Debate
One of the major debates within the family literature that has been occurring
over the past 30 years is determining whether the family structure (the organization
of the family) or family process (the method in which family members interact with
each other) has a greater influence on adolescent sexual activity. The early family
researchers mainly focused on how the household organization may influence
adolescent sexual activity because this factor strongly guides how family members
interact with each other. According to this perspective, adolescent sexual activity
may be a function of household attributes such as parental marital status, the child’s
ordinal position, household size, and the sibling’s gender (Rogers, 1983; Miller et al.,
1987).
Briefly, parental marital status is associated with adolescent sexual activity
because it may be perceived as a proxy for a parent’s availability to monitor his or
her children’s activities (Newcomer & Udry, 1987). Married parents are more likely
to monitor their children’s activities than divorced or unmarried parents because
there are two adult figures present in these households. Children who reside with
married parents are less likely to engage in sex during puberty than those who reside
3
with a single parent since they have more opportunities to prevent their children from
engaging in deviant behaviors.
A child’s ordinal position is associated with adolescent sexual activity
because of attitude differences between older and younger siblings (Reiss, 1967).
Younger siblings are more likely than older siblings to possess permissive attitudes
including those regarding premarital sex. Consequently, younger siblings are more
likely than older siblings to engage in sexual activity during puberty. The reason for
the attitude difference is due to the older siblings having to bear more responsibility
for other family members because of their position within the family.
Household size is associated with adolescent sexual activity because it
defines the amount of time and money that parents can invest into their children
according to the resource dilution hypothesis (Powell & Steelman, 1993). Since
these resources are limited, a child who has many siblings is more likely to be
apportioned a smaller share of these goods compared to those who have fewer
siblings. Consequently, a child who has many siblings may feel less support from
his or her parents than those without many siblings.
Finally, sibling’s gender is associated with adolescent sexual activity through
the masculinization of the family environment. Rogers (1983) found support for the
masculinization hypothesis where a male dominated environment would be
supportive of premarital sexual activity if one views sex as a “masculine” activity (p.
82). Thus, a child with many male siblings may be more likely to engage in sexual
activity during puberty because his or her brothers could (1) have more discussions
4
about sex with them, (2) help to provide male and female friends who may be
potential partners for them, and (3) act as role models regarding sexuality.
Although many studies have shown family structure to be strongly associated
with adolescent sexual activity, the trend in the family literature has shifted toward
investigating how interactions between family members, or the family process, may
also account for this issue. Based on social learning theory, the family process
perspective focuses on the importance of family interactions in the learning and
subsequent performance of behaviors (Miller, Forehand, & Kotchick, 1999).
According to these researchers, the shift in focus to investigating family process
occurred because the family structure perspective has not played a prominent role in
the theoretical framework to understand, and the practical efforts to alter, the risk for
adolescent sexual activity due to the perception that the household organization is
static and not immediately susceptible to change through therapeutic intervention. In
addition, this perspective argues that it is easier to change the family processes rather
than changing the family structure when altering adolescent outcomes. Some of the
family processes found to be associated with sexual activity include the quality of the
parent-child interaction, whether or not the parents monitor their children’s activities,
and the parent’s attitudes regarding adolescent sexual activity (Miller, Forehand, &
Kotchick, 1999).
A number of researchers, such as Davis & Friel (2001), have recently
attempted to unite the family structure and family process perspectives by noting that
family processes can be viewed as interpretations or explanations for the family
structure and adolescent sexual activity relationship. In another words, these
5
researchers investigated whether family process could be the reason why adolescents
who reside in non-traditional households are more likely to be sexually active than
those who reside with two biological parents. According to Davis & Friel (2001),
the explanations for most of these non-traditional household effects could be divided
into three different family process rationales: familial control, familial involvement,
and familial culture. Familial control is the process of how the parents monitor and
discipline their children. Familial involvement is the process of how parents and
their children maintain a relationship with each other. Familial culture is the race
and gender context in which familial control and familial involvement operates.
Dissertation Objectives
There are two different investigations in this study. The goal of the first
investigation is to determine the best family process explanation for the increased
risk for sexual activity among children who reside in various non-traditional
households compared to those who reside with two biological parents. This
investigation, which follows the trend of uniting the family structure and family
process explanations, examines the family structure effect on adolescent sexual
activity for three non-traditional households by using four family process
explanations and a peer explanation as intervening interpretations. The three non-
traditional households that will be investigated in this study are unwedded mother,
divorced/separated mother, and stepparent households. Figure 1-1 presents the
6
Figure 1-1: Proposed Model for the Relationship Between Family Structure and
Adolescent Sexual Activity
The Five Explanations
Family Structure
Two Biological Parents
Unwedded Mother
Divorced/Separated Mother
Stepparents
Adolescent
Sexual
Activity
Parental Supervision
Parental Attitudes
Family Support
Family Transitions
Peer Influence
7
model for this investigation. The terms “measure” and “explanation” are used
interchangeably in this study.
Researchers have proposed a number of social influences to explain why
adolescents who reside in non-traditional households are more likely to engage in
sexual activity than those who reside with two biological parents (Brooks-Gunn &
Furstenberg; 1989; Miller & Moore, 1990). Some of these social influences include
family transitions, parental supervision, parental attitudes, family support, and peers.
However, there are several general drawbacks to these studies even though
researchers have found that these influences do individually account for the non-
traditional household effect.
The first drawback is that some studies, such as those by Wu & Thomson
(2001) and Jacobson & Crockett (2000), investigated only one or two of these family
influences when exploring this relationship. While methodologically sound, the
disadvantage of these investigations is that they only provide a limited perspective of
the family influence and adolescent sexual activity relationship. Since these studies
are limited in their scope, one would have to refer to other studies in order to
determine whether other family influences may also explain this association. It may
be tempting to conclude that the findings from one study may also be applicable to
another study if the samples from both studies share similar characteristics.
However, there may be subtle differences regarding the sample among several
studies even though they may have similar demographic characteristics. For
example, the neighborhood context may be different across several samples even
though these respondents may be of similar gender, race, and age such as the case
8
when one study focuses on rural adolescents while another study focuses on urban or
suburban adolescents. Consequently, one would not definitively know whether the
findings from one study could be applicable to another study even though most of
the demographic factors may be similar across study samples.
The second drawback is found in studies that use two or more explanations to
account for the family structure and adolescent sexual activity relationship. In these
studies, such as those by Davis & Friel (2001), the family influence explanations
would be analyzed simultaneously rather than individually. This type of analysis
would be perfectly acceptable if the focus of the study is whether the non-traditional
household effect can be completely explained away by several family influences.
However, the main weakness of these analyses is that it would be difficult to
determine which explanation has a greater influence for the family structure effect
for several non-traditional households.
Studies have shown that family members who reside in non-traditional
households face a variety of issues that are unique to each household (Furstenberg,
1987; Newcomer & Udry, 1987). This finding suggests that some family influences
might be a better explanation than others for the family structure and adolescent
sexual activity relationship for the various non-traditional households. Thus, it
would be difficult to determine whether there are subtle family environment
differences among these households if the family influences are examined
simultaneously rather than individually since these differences are blurred.
The first investigation addresses the first drawback by using the same sample
throughout this study when determining the best explanation for the family structure
9
and adolescent sexual activity relationship among three non-traditional households.
This method avoids the issue of having to determine whether the findings from one
study may be applicable to another study because of possible sample differences.
The first investigation addresses the second drawback by regressing the
family and peer influence explanations independently rather than in a simultaneous
manner. This method is congruent with the premise that the family structure and
adolescent sexual activity relationship is not the same across the various non-
traditional households because there are a variety of issues that are unique to each
household. Consequently, this method can reveal whether the relationship between
family structure and adolescent sexual activity differs across these non-traditional
households by demonstrating that certain social influences are better explanations
than others for this effect. These results can then be effectively used to address the
primary underlying cause that may increase an adolescent’s risk to engage in sexual
activity for a particular non-traditional household.
One way to determine the best explanation for the relationship between
family structure and adolescent sexual activity among several non-traditional
households when investigating multiple family influence explanations is to use
stepwise regression analyses. However, there are several weaknesses when using
this method. First, even though it is possible to rank the strength of the various
family influence explanations for this relationship by using stepwise regression
analyses, this method does not allow one to determine if a single explanation could
individually account for the family structure effect because the eventual result is to
examine the tandem effect of two or more explanations. Consequently, the eventual
10
results of these analyses are similar to those that analyze several family influences
simultaneously without going through a stepwise process.
The second weakness of using stepwise regression is issue of causal order
that one may need to address when regressing the explanations in a stepwise manner.
This issue is important because it establishes the time order in which these family
influences occur. However, it would be very difficult to determine which of the
family influences that will be explored in this study should come first, second, third,
etc. because establishing the time order of these influences may be debatable. Thus,
regressing each explanation individually avoids the need of having to address a
difficult issue.
The focus of the second investigation is to examine the family and peer
influence explanations as predictors of adolescent sexual activity in their own rights
rather than as mediators of family structure. Regardless of family structure, this
investigation views each of the five explanations as potentially powerful predictors
of adolescent sexual activity. In this second inquiry, demographic variables such as
gender, race, family structure, and socioeconomic status function as control variables
while family and peer influence variables are portrayed as independent variables.
The goal of the second investigation is to identify which explanations are
directly and indirectly associated with adolescent sexual activity when they are
analyzed simultaneously. This investigation uses hierarchically ordered regression
analyses where the measures are regressed in a stepwise manner. The reason for
doing this investigation is because researchers have noted that several of these
explanations may possibly mediate the effects of other explanations (Newcomer &
11
Udry, 1987; Wu & Thomson, 2001). However, the family literature has not fully
explored this premise. Figure 1-2 presents the model for this investigation.
This study will use the 1994-1995 National Study of Adolescent Health (Add
Health) Wave I dataset for the two investigations. The subjects for this study will be
white, black, Hispanic, and Asian adolescents of both genders and their biological
mother. These investigations will be performed on the total sample and within the
adolescent’s race, gender, and race-gender contexts using t-tests and logistic
regressions.
Brief Overview of the Explanations
As mentioned previously, some of the family influences found to be
associated with adolescent sexual activity include marital or family transitions,
parental supervision, parental sexual attitudes, and family support (Brooks-Gunn &
Furstenberg, 1989; Miller & Moore, 1990; Wu & Thomson, 2001). One can make
the argument that some of these influences should be examined simultaneously
rather than individually because they may share similar basis for explaining
adolescent sexual activity. For example, family transitions and parental supervision
levels or parental supervision levels and family support should be analyzed
simultaneously because of their high correlation with each other. However, one can
also make the argument that these influences could be analyzed separately because
each explanation has a relationship with adolescent sexual activity that is unique.
The next chapter will discuss this topic in further detail. Meanwhile, the next few
12
Figure 1-2: Proposed Hierarchically Ordered Regression Model Between the Five
Explanations and Adolescent Sexual Activity
Family Transitions
Parental Supervision
Parental Attitudes
Family Support
Peer Influence
Adolescent
Sexual
Activity
Demographic
Variables
Gender
Race
Family Structure
Socioeconomic Status
13
sections present a brief description of how these family influences may be directly
associated with adolescent sexual activity.
Family Transitions
There are several reasons why a family transition event, such as a household
being transformed from a two-parent family to a single parent family and vice versa,
may be associated with adolescent sexual activity. First, a family transition event
may indicate that family members are currently experiencing, or have experienced, a
turbulent home environment. The turbulent home environment may be due to family
members being forced to address issues when a household transforms from one form
to another. This process may lead parents and their children to experience frequent
stresses in their lives as they adjust to their new lifestyle. Some issues that could
contribute to the turbulent home environment for family members who live in
divorced households include financial problems and preoccupation by both parents
with the marital problem (Newcomer & Udry, 1987). In addition, family members
of stepparent households may face issues such as the stepparents being confused
about the appropriate roles for them in disciplining their stepchildren and stepparents
having difficulty in forging an emotional relationship with their stepchildren
(Cherlin, 1978; Amato, 1987; Furstenberg, 1987). Children who are exposed to
these issues may have an increased desire to leave home and seek an alternate
intimate relationship with people outside of the family (Thornton, 1991).
14
The second reason why family transition may be associated with adolescent
sexual activity is because these events provide children the opportunity to model
their parent’s non-marital behaviors. For families that are undergoing, or have
undergone, the transformation from a two-parent to a one-parent household, the
residing parent may be dating, or have dated, other adults during the transition
process. This dating activity may include sexual activity (Wallerstein & Kelly,
1980). Parents who engage in sexual activity while dating others may transmit the
message to their children that engaging in sexual behaviors outside of a marriage is
acceptable (Whitbeck, Simons, & Kao, 1994). Since children learn the appropriate
norms, behaviors, and attitudes from their parents (Thornton & Camburn, 1987),
those who have experienced a family transition event may be more likely to engage
in sexual activity by the time they reach young adulthood than those who have not
experienced these situations.
The third reason why family transitions may be associated with adolescent
sexual activity is because these events result in the formation of households where
the parents are less likely to monitor their children’s activities closely compared to
married parents (Newcomer & Udry, 1987; Thomson, McLanahan, & Curtin, 1992).
For the divorced/separated parent households, the absence of a parental figure from
the home may suggest that single parents monitor their children’s activities less
closely than parents of two parent households because they are either 1) working
longer hours at one job or have multiple jobs in order to maintain the household
income (Thornton, 1991) or 2) doing non-market activities, such as running
household errands, which could have been done by the absent parent (Biblarz &
15
Raftery, 1999). Children whose activities are monitored less closely by their parents
are given greater opportunities to do whatever they want including engaging in
sexual activity (Zelnick & Kantner, 1980).
Children who reside with stepparents may be less likely to engage in deviant
behaviors than those who reside with a single parent because of the increased
parental presence at home. However, there may be a reduction of a stepparent’s
parental authority if their stepchildren view him or her as being a “relative stranger”
(Beer, 1988). This reduced parental authority may result in stepparents having more
conflicts with their stepchildren when they try to discipline them (Amato, 1987).
The increased conflict between stepparents and the stepchildren may lead to the
children being more likely to engage in sexual activity during puberty.
Parental Supervision
Parental supervision may be associated with adolescent sexual activity
through several pathways. The first pathway is through the parental monitoring of
the children’s activities (Jacobson & Crockett, 2000). As mentioned previously,
children whose activities are being monitored less frequently by their parents have
greater opportunities to engage in sexual activity during puberty than those whose
activities are being monitored more frequently. The second pathway is through the
parent-child interactions (Miller, Forehand, & Kotchick, 1999). Increased parental
presence around the home may lead to parents having greater opportunities to
connect with their children through activities such as doing things together or talking
16
about one’s day. These interactions may help to foster a closer relationship between
parents and their children. Children who have a stronger connection with their
parents are less likely to engage in deviant activities, such as sexual behaviors,
because it would be easier for the child to accept their parent’s standard of
appropriate behavioral conduct (Wu & Thomson, 2001).
Parental Attitudes
This explanation suggests that children learn about their parent’s attitudes
and values through the frequent interactions with each other. The attitudes and
values that are conveyed from a parent to his or her children may influence their
decision to engage in sexual activity during puberty. Parents who hold conservative
sexual views may be more likely to transmit messages to their children that
discourage sexual activity during puberty than those who hold relatively liberal
values (Dittus & Jaccard, 2000). Consequently, children who are taught
conservative sexual values by their parents may be less likely to engage in sexual
activity before reaching young adulthood compared to those who are taught
relatively liberal sexual values. Measures that represent a parent’s sexual attitude
also include the parent’s religiosity, the parent’s view regarding adolescent sexual
activity, and frequency of sex discussions that the parent has with his or her children.
A parent’s view regarding adolescent sexual activity and his or her decision to talk
about sex with their children may also be independent from his or her religious
values.
17
Family Support
The family support explanation suggests that a child’s decision to engage in
sexual activity during puberty is influenced by his or her feelings of support by
immediate family members. Children experience numerous biological,
psychological, and social changes during puberty (Havighurst, 1967) that can be both
confusing and challenging for them. Those who struggle with these changes may
feel emotionally insecure about themselves if the immediate family members are
perceived as being unsupportive as they work through these issues. Consequently,
children who receive support from immediate family members during this time may
be less likely to seek intimate relationships from individuals outside of the household
compared to those with less supportive immediate family members.
Peer Influence
In addition to family influence explanations, the type of friends and peers that
a child spends time with may be associated with adolescent sexual activity. The
pathway in which peers may influence a child’s decision to engage in sexual activity
is through peer pressure or the feeling that one needs to conform to a particular
group’s ideas and values in order to feel connected to and become accepted by them.
Having acquaintances with friends who engage in deviant activities may provide
greater opportunities for children to engage in sexual behaviors during puberty
18
(Metzler et al., 1994; Furstenberg et al., 1987; Gibson & Kempf, 1990; Schuster et
al., 1998).
Functional and Conflict Perspectives of Socialization
It is my contention that the five social influences that will be explored in this
study are consistent with the classic functional and conflict perspectives on
childhood socialization. The functional perspective views the family as a vehicle
where childhood socialization occurs (Parsons & Bales, 1955; Bredemeier &
Stephenson, 1962). Parents teach the appropriate norms, behaviors, and attitudes to
their children early in life in order to produce conformity, to instill a sense of
identity, and to prepare them for life through processes such as modeling, reward and
punishment, and parental discipline (Thronton & Camburn, 1987; Maccoby &
Martin, 1983). This process fosters the development of an independent censor, or
conscience, which guides the child to act appropriately according to societal mores
(Mead, 1964). The conscience also provides a sense of discomfort whenever the
child violates societal values even when there is no one around to observe the
transgression. In addition, the conscience allows children to learn about the world
and helps them to act independently during adulthood. Thus, the functional
perspective emphasizes conformity to societal norms and values.
In contrast, the conflict perspective views the family as a power struggle
between the parents and their children. This struggle occurs when parents attempt to
maintain control over their children by not letting them into the privileges of adult
19
life (Collins, 1975). Initially, the power within the family mainly lies with the
parents since they are physically bigger and stronger than their children. In addition,
parents are able to accumulate numerous advantages over their children because they
were here first. However, children begin to emancipate themselves from their family
when they grow older despite continued attempts by their parents to maintain control
over them. The events that could contribute to this emancipation process include 1)
becoming more exposed to and influenced by peer group values and expectations, 2)
attaining full physical stature which allows them to physically defend themselves or
run away during conflict with their parents, and 3) developing resources such as
acquiring full mobility and learning skills that will help them to become independent
from their parents.
As developed in this study, the parental supervision, parental attitude, and the
family support explanations are more consistent with the functional interpretation
while family transition and deviant peer group explanations are more consistent with
the conflict interpretation. A number of places in this study will use these functional
and conflict interpretations to help elucidate the findings of the study.
Study Questions
Based on the dissertation objectives described in the previous section, this
study will attempt to answer to five main questions. Once again, these questions are
also applicable for the investigations within the adolescent’s gender, the adolescent’s
race, and the adolescent’s race-gender contexts:
20
1. Does the risk for adolescent sexual activity differ across the three non-
traditional household configurations?
2. Which of the five explanations is the best reason for the family structure and
adolescent sexual activity relationship for the three non-traditional
households?
3. Which variables among the five explanations are significantly associated with
adolescent sexual activity when the explanations are regressed individually?
4. Which of the five explanations is the most and least parsimonious adolescent
sexual activity explanation?
5. Pertaining to the second analysis of this study, which variables among the
five explanations are directly and indirectly associated with adolescent sexual
activity when they are considered simultaneously?
Study Significance
There are several reasons why this study has potential significance. First, this
study will provide insight into the relationship of two social trends that occurred for
a 30-year period beginning in the early 1970s: the increasing percentage of children
who have engaged in sexual activity during puberty and the increasing percentage of
children who do not reside with two biological parents. According to the U.S.
Census Bureau (2003), the percentage of children who reside in non-traditional
households during this period increased from 14% in 1970 to 27% in 2000. Since
children who reside in non-traditional households are more likely to engage in sexual
21
behaviors during puberty, this finding suggests that more children have become
vulnerable to the potentially adverse consequences of adolescent sexual activity
during this period. Discovering which explanation best account for the family
structure effect among three non-traditional households for the total sample and
within the adolescent’s gender, adolescent’s race, and adolescent’s race-gender
contexts may help educators and therapists to create targeted social policies and
interventions respectively that could help to either prevent or delay these behaviors.
Second, this study is significant because it will determine the best reason for
the family structure and adolescent sexual activity relationship among three non-
traditional households. Earlier studies that investigated this relationship, such as
those by Rogers (1983) and Young et al. (1991), greatly simplified the analysis by
classifying households into either one or two parent categories. Several assumptions
are made when using this simplified method to analyze the data. The first
assumption is that the present family configuration, as determined by the number of
parents in the household, has a greater influence on a child’s decision to engage in
sexual behaviors than the events that households have gone through in order to reach
their current family status. The second assumption is that the family environments
of non-traditional households for those that are of similar organizational structure are
presumed to be similar to each other. Both of these assumptions would eliminate the
need for one to investigate whether the family structure effect regarding adolescent
sexual activity differs by household configuration.
Later studies, such as those by Wu & Martinson (1993) and Moore (2001),
have found that there are differences in the family environment for the various one
22
and two parent households even though they may share similar structural
characteristics. The main factor that distinguishes one particular household from
another for those that are of similar organizational structure is the family transition
event. Family members who have undergone one or more family transition event(s)
may experience greater stresses in their lives because they are either currently feeling
or have felt the effects of a chaotic family environment. In addition, Wu &
Martinson (1993) and Wu & Thomson (2001) have argued that the impact of these
events on adolescent sexual activity could be underestimated when non-traditional
households are classified into just one and two parent configurations.
These reasons have resulted in a number of studies that started to account for
the family transition effect when investigating the relationship between family
structure and adolescent sexual activity. Researchers have used several methods in
their studies to measure this effect. The first method is the direct method where the
number of marital or family transition events that a child has experienced in his or
her life would be tallied. The age at which researchers begin the count for the
number of marital or family transition is fourteen (Wu & Martinson, 1993). The
result of the tally would then be used as a control variable. The second method is the
implied method where households are placed into categories that inherently describe
whether a family or marital transition event has taken place in the past. Researchers
who have used this method in their studies, such as those by Davis & Friel (2001)
and Moore (2001), assumed that the family members of unwedded parent households
have not experienced the family transition event in their lives even though they share
a similar family structure with divorced households. This premise also applies to the
23
two parent household category where the family members of stepparent households
would have at least experienced a family transition event in their lives while those
from two biological parent households presumably have not experienced these
events at all.
This investigation will use both the parent’s current marital status and a
family transition measure, as represented by the variable that measures the number of
previous serious and marriage-like relationships that the parent has had in the past 18
years, to determine whether there is a relationship between family transition event(s)
and adolescent sexual activity. The parent’s current marital status may also be an
independent indicator of whether a household had previously gone through at least
one family transition event. Meanwhile, the family transition measure that will be
used for this study represents the amount of family transition events that a child may
have possibly experienced during his or her lifetime. Chapter 3 presents an in-depth
discussion of the limitations for using this particular family transition measure as a
family influence explanation.
Third, this study is significant because it will examine the relationship
between unwedded mother households and adolescent sexual activity. Children who
reside with an unwedded parent presumably have not experienced a family
disruption event during their lifetime. Yet, children who reside in these households
are more likely to engage in sexual activity during puberty than those who reside in
two biological parent households (Moore, 2001). This type of household has been
understudied in the family literature because researchers tend to characterize one-
parent households as those of divorced parent families despite a growing number of
24
households headed by mothers who have not married in the past (Moore, 2001). The
discovery of this and other new types of households have encouraged researchers to
examine how the risk for adolescent sexual activity may differ in new family
environments (Davis & Friel, 2001; Moore, 2001). This study continues this trend
by analyzing the unwedded parent and divorced/separated parent households
separately in order to determine whether there are differences between these two
family configurations in regards to adolescent sexual activity.
Fourth, this study is significant because it will determine the best reason for
the family structure and adolescent sexual activity relationship among three non-
traditional households using five different explanations. An extensive review of the
literature has not revealed a single published study that has examined this
relationship using this type of analysis. As mentioned previously, a number of
researchers, such as Moore (2001) and Wu & Thomson (2001), have tested a number
of explanations that will be used in this study, but these investigations have only
focused on one or two explanations at a time. Consequently, these studies have only
provided a limited view of the relationship between family structure and adolescent
sexual activity. Investigating multiple explanations in a single study would allow
one to perform a more extensive examination of this relationship.
Fifth, this study is significant because it will employ two ranking analyses
that are not usually seen in other family studies. The first ranking analysis would
determine which explanation best mediates a particular non-traditional household
and adolescent sexual activity relationship. The second ranking analysis would
determine which explanation is the most and least parsimonious for adolescent
25
sexual activity. The reason for using model parsimony rather than model variance or
the R
2
in order to determine the explanations that account for this relationship will be
discussed in chapter 3.
Educators and counselors can use these results to identify and effectively
address the pertinent family issues in the hope of lowering adolescent sexual activity
rates. For example, if functional explanations explain these behaviors better than
conflict explanations, it could suggest that an effective treatment or social policy to
prevent children from engaging in or delaying sexual activity during puberty may
include placing greater emphasis on parents to either 1) develop a closer connection
with their children through doing activities together, 2) monitor their children’s
activities more closely, 3) talk to their children about safe sexual practices, or 4)
explain to their children their views about sex. However, if conflict explanations
explain these behaviors better than functional explanations, it could suggest that an
effective treatment or social policy to prevent children from engaging in or delaying
sexual activity may include placing a greater emphasis on parents to 1) develop a
stable household for their children so they do not feel unsupported during family or
marital transition events or 2) understand the peer group influences and reemphasize
the values that they want to convey to their children.
Finally, this study is significant because it will investigate whether some of
these explanations have either a direct or indirect association with adolescent sexual
activity. While many previous studies have investigated whether the explanations
have a direct association with adolescent sexual activity, researchers have noted that
some of these measures may have an indirect association with adolescent sexual
26
activity (Newcomer & Udry, 1987; Wu & Thomson, 2001). However, the family
literature has not fully explored the premise. This study investigates this premise in
order to develop a better understanding of the relationship between family influences
and adolescent sexual activity.
Preview of Chapters
Chapter 2 will present a literature review regarding the five explanations and
adolescent sexual activity relationship.
Chapter 3 will present an in-depth explanation regarding the National
Longitudinal Study of Adolescent Health dataset. This chapter also provides a
detailed description of the dependent variables, the independent variables, and the
statistical analyses that are used in this study.
Chapter 4 will analyze the three non-traditional households and adolescent
sexual activity relationship for the total sample. This chapter contains hypotheses,
the results of the analyses, and a discussion of the findings.
Chapter 5 will present an argument for investigating adolescent sexual
activity within the gender and race contexts. This chapter presents theories
explaining the reasons for gender and race differences regarding this issue. The
chapter also contains a literature review of how the five explanations and adolescent
sexual activity relationship may differ within the gender and race contexts.
Chapter 6 is similar to chapter 4 except that the data will be analyzed within
the adolescent’s gender context. This chapter also has a discussion of the findings.
27
Chapter 7 is similar to the previous chapter except that the data will be
analyzed within the adolescent’s race context.
Chapter 8 is similar to the previous two chapters except that the data will be
analyzed within the adolescent’s race-gender context.
Chapter 9 will present the results of the five explanations and adolescent
sexual activity relationship using hierarchically ordered regression analyses. This
chapter contains analyses for the total sample and within the adolescent’s gender and
race contexts. However, there is a limited analysis within the adolescent’s race-
gender context. This chapter also has a discussion of the findings.
Chapter 10 concludes the study. This chapter answers the questions posed in
the introduction, discusses several interesting findings of the study, and discusses the
limitations, implications, and directions for future research.
Chapter Summary
This study involves two main investigations. The first investigation is to
determine the best reason for the family structure and adolescent sexual activity
relationship among three non-traditional household configurations (unwedded
mother, divorced/separated mother, stepparent) using four family influence
explanations (family transition, parental supervision, parental attitude, and family
support) and a peer explanation. This investigation consists of conducting separate
multiple regression analyses for each of the explanations in order to allow for the
possibility that more than one of these explanations may mediate this relationship
28
and to avoid possible time order issues. In addition, this analysis can help one to
determine which explanations are the most and least parsimonious for adolescent
sexual activity.
The second investigation is a simultaneous examination of the five
explanations and adolescent sexual activity relationship. This investigation, which
places the five explanations at the forefront of the analyses, is based on the premise
some of the explanations in this study may possibly mediate other explanations. In
another words, this investigation tries to determine which family and peer influence
variables have a direct and an indirect association for this relationship. These
analyses consist of conducting stepwise regressions until all five of the measures are
eventually regressed into a single regression model.
These investigations will be conducted on the 1994-1995 National Study of
Adolescent Health (Add Health) Wave I dataset. This study will examine white,
black, Hispanic, and Asian adolescents of both genders and their biological mother.
These investigations will be conducted on the total sample and within the
adolescent’s gender, the adolescent’s race, and the adolescent’s race-gender contexts
using t-tests, logistic regressions, and ordinary least squares regressions.
There are several significant reasons for this study. Some of these reasons
include the need to investigate unwedded mother households since they are
understudied, the need to investigate which of the five measures best explain
adolescent sexual activity so that adults who are concerned about this issue can
effective address the underlying causes, and the need to investigate which family and
peer influence variables are directly and indirectly associated with adolescent sexual
29
activity so that one can understand which of these factors are important. In essence,
this study attempts to provide a more comprehensive examination of the relationship
between both family and peer influences and adolescent sexual activity.
30
Chapter 2: Theories
This chapter will provide in-depth discussions of how the five explanations
may be associated with adolescent sexual activity. Chapter 5 will provide another
literature review that explores how this relationship may differ within the
adolescent’s gender and race contexts.
Family Transition Explanation
The family transition explanation focuses on children’s responses to the
stresses that accompany family disruptions. This explanation assumes that children
find certain family events, such as divorces and remarriages, relatively more stressful
than other events, such as changing residence or attending a new school, because
there needs to be greater adjustments in their lives in order to adapt to the new
environment (Holmes & Rahe, 1967; Wu & Martinson, 1993). A major change
within the family creates turmoil within the household and often disrupts a child’s
sense of emotional security. Children may feel insecure in these situations because
they do not know whether they can rely on their parent to provide stability to the
home environment and emotional support during the times before, during, and after
the family transition event.
This explanation potentially measures the frequency of turbulent family
events that a child could find to be stressful during his or her lifetime. Resident
parents who have greater interactions with their children may help to mitigate the
31
stresses that they go through when a household transitions from one form to another.
However, a resident parent’s interactions with his or her child may not necessarily
mitigate all the stresses of the family transition event since the child can still have
lingering empty feelings regarding the relationship with the non-residential parent
since they are not present around the home as often as before. This feeling may be
more pronounced for children who have experienced multiple family transition
events during their lifetime.
Regardless of whether the transition event results in a parental figure loss or
gain, children respond to the family turmoil by either engaging in impulsive,
rebellious, and aggressive actions or by spending less time at home (Wallerstein &
Kelly, 1980; Wallerstein & Blakeslee, 1989). Children may also be looking to peers
for emotional support when immediate family members are not emotionally
available to them during these times (Coleman, 1981). The availability of peers
somewhat softens the impact of the transition event (Hetherington, 1979), but also
hastens a child’s entry into sexual activity (Hetherington & Clingempeel, 1992).
Researchers have described several reasons for the family transition effect.
The first is family disruption. Family disruptions, defined as the household
instability that accompanies marital disruption and subsequent marital transitions
through parental divorce, remarriage, and redivorce (Wu & Martinson, 1993;
Capaldi, Crosby, & Stoolmiller, 1996), are events that cause disorder in the
interactions between family members because these events would compel them to
adjust to a new family configuration. This disorder may also contribute to a
32
turbulent family environment by affecting the quality of interaction between the
parents and their children (Wu & Martinson, 1993).
Divorced parents and stepparents face different issues after the occurrence of
a family transition event. Divorced households are characterized by 1) emotional
turmoil, 2) preoccupation by both parents with the marital problem, 3) removal of
one parent from the home, 4) financial problems, 5) a shift of total day-to-day
responsibility for the supervision of children to the lone single parent when he or she
has enough other problems to keep him or her preoccupied, and 6) disruption of
established relationships between the adolescent and both parents (Newcomer &
Udry, 1987). Parents who struggle with these issues further contribute to a turbulent
family environment by having even greater difficulty with disciplining their
children’s behaviors (Newcomer & Udry, 1987).
However, integrating a family member into the household may also be a
stressful time for children because family members have to adjust to new roles and
expectations so the household can function smoothly. Stepparents may experience
problems such as 1) not knowing their and the non-custodial parents role in
disciplining children, 2) inability to create emotional attachments with the non-kin,
3) not knowing the fiscal and legal responsibilities regarding stepchildren, and 4) not
knowing the proper terms to describe the stepchildren (Cherlin, 1978; Amato, 1987;
Furstenberg, 1987). Stepchildren may also have a difficult time with accepting their
stepparent as a parental figure because they may view him or her as a “relative
stranger” (Beer, 1988). These issues can lead stepchildren to becoming less open
33
and rebelling against or finding ways to undermine their stepparent’s authority
during the integration process (Cherlin, 1978; Amato, 1987; Furstenberg, 1987).
A parent’s gender may be associated with how much of an impact the
disruptions may affect the lives of family members for households that have
transitioned or are in the process of transitioning from a two-parent to a single parent
family (Thomson, McLanahan, & Curtin, 1992). Gender plays a prominent role in
this situation because mothers are usually the custodial parent when they are
separated from their husbands. In general, divorced mothers may be similar to
married mothers regarding the adherence to traditional maternal roles. This
explanation contends that divorced or separated mothers exert weaker controls or
demands on their children because they were not the main parental figure to
discipline the children during the marriage. Since fathers usually carry more
authority compared to mothers during the marriage, divorced mothers may have a
more difficult time with disciplining their children’s behaviors when the separation
from a spouse occurs. This weakened parental authority may result in children
feeling that the family is chaotic and disorganized for those who reside with a single
mother compared to those who reside with two biological parents.
A selection bias may also explain how a family disruption affects a child’s
life. This somewhat controversial argument contends that households that have gone
through multiple family transitions may be selective for parents who possess and
express antisocial personalities and behaviors (Simons et al., 1999). Although many
divorced women do not display the risky behaviors that are seen as inappropriate or
insensitive to a majority of the population such as fighting, substance abuse, and
34
sexual promiscuity, some studies have shown that divorced parents are more likely
than married parents to display antisocial tendencies (Capaldi & Patterson, 1991;
Simons et al., 1996).
There are several reasons why parents with antisocial personalities are more
likely to have children who engage in deviant activities. First, antisocial parents may
possess poor parenting skills (Simons et al, 1999). These parents may not have
developed the problem-solving skills that are necessary when interacting with other
individuals (Wu & Thomson, 2001). Parents who do not have these problem-solving
skills may not be empathetic to another person’s concerns. Children who are in this
situation may experience additional stresses in their lives of having to contend with
issues relating to the family transition event along with who reside with a parent who
is uncaring about his or her needs. Consequently, children who reside in these
households may respond to this lack of parental emotional support by seeking
intimate relationship with others outside of the family. Indeed, researchers have
found that mothers’ antisocial behaviors are associated with poor parenting qualities
(Capaldi & Patterson, 1991; Simons et al., 1996).
The second pathway for the family transition effect is the modeling of a
parent’s behaviors by his or her children. Modeling is also another reason why
parents who possess antisocial personalities are more likely to have children who
engage in deviant activities (Simons et al., 1999). Children who reside with
antisocial parents have a greater opportunity to observe and follow these non-
normative behaviors, including sexual activity, during puberty.
35
In addition to antisocial behaviors, children may learn about sexual values
when their parents begin to date other adults during the family transition process or
after the family has transitioned from a two-parent to a single parent household
(Whitbeck, Simons, & Kao, 1994). Since many single parents engage in non-marital
sexual intercourse while dating other adults (Wallerstein & Kelly, 1980; Thornton &
Camburn, 1987), children may conclude from these behaviors that non-martial
sexual intercourse is acceptable. Children may also be highly aware about the
relationship’s sexual nature when a parent decides to cohabit with a new partner
(Wallerstein & Kelly, 1980). Thus, the witnessing of the parent’s sexual partner
presence and sexual relationships at home could accelerate a child’s sexual
involvement (Newcomer & Udry, 1984; Thornton & Camburn, 1987; Moore &
Chase-Lansdale, 2001).
Besides transmitting sexual values, researchers have found that a parent’s
dating activities may also affect a child’s sense of well-being. One consequence of
single parents who engage in non-marital sexual intercourse when dating others is
that their children were often an audience to the accusations and counteraccusations
of the other parent’s promiscuity (Wallerstein & Kelly, 1982). Children who
experienced these events often felt dismayed by their parent’s behaviors and began
to respond to this situation in a variety of ways such as becoming depressed, having
a desire to grow up fast, having suicidal thoughts, and engaging in sexual activity in
order to cope with the situation at home.
Parental dating is also associated with the children adopting liberalized
sexual attitudes. Thornton & Camburn (1987) found that daughters of divorced
36
mothers are more likely than the daughters of married mothers to have liberal views
regarding sex outside of a marriage. Furthermore, daughters of remarried mothers
are even more likely than daughters of divorced mothers to have liberal sexual
views. They concluded that the mother’s post-divorce dating behavior, rather than
the marital disruption itself, was responsible for liberalizing their daughter’s sexual
attitudes.
Other studies have looked further into the association between parental dating
behaviors and adolescent sexual activity. Inazu & Fox (1980) found that the
mother’s behavior in her previous relationships is associated with the daughter’s
sexual status. Mothers who cohabitated during their daughter’s lifetime were more
likely than those who did not to have daughters that engaged in sexual activity
during puberty. Furthermore, mothers who were unmarried during their first child’s
birth were more likely to have a daughter who had experienced a sexual episode
before reaching young adulthood than those who were married during this period
(Inazu & Fox, 1980; Thornton & Camburn, 1987).
In another study, Whitbeck, Simons, & Kao (1994) found that mother’s
dating influenced both of her son’s and daughter’s sexual behaviors but through
different pathways. They found that mother’s dating was directly associated with her
son’s sexual behaviors, but indirectly associated with her daughter’s sexual
behaviors. The reason for the indirect relationship between mother’s dating and her
daughter’s sexual behaviors was due to the mother’s attitude regarding sexual
activity. In another words, sons were more likely to engage in sexual behaviors if
they were able to observe their mother’s dating behaviors while daughters are more
37
likely to engage in sexual behaviors if their mothers adopted more liberalized sexual
views.
Parental Supervision Explanation
The social control theory, as posited by Hirschi (1969), is the basis for the
parental supervision explanation in regards to adolescent sexual activity. Social
control and Merton’s (1957) strain theory have contributed to explaining why
adolescent sexual activity occurs. However, these theories differ from each other in
several ways (Lauritsen, 1994). First, social control theory emphasizes the factors
that prevent individuals from engaging in non-normative behaviors rather than the
disparity between culturally accepted values and structurally available means to
achieve those values. Second, social control theory does not need to take into
account the variation of the individual’s frustration level or motivation to deviate
from societal norms because it could be argued that deviant behaviors, such as sexual
activity or drug use, are inherently attractive to a majority of the population and
especially for adolescents. Since there is always adequate motivation for any
individual to engage in deviant activities regardless of the individual’s frustration or
motivation level, social control theorists try to identify the factors that prevent people
from engaging in deviant activities.
Social control theorists, such as (Hirschi, 1969), have argued that deviance is
most likely to occur when the bond between the individual and society is weak or
non-existent. This bond consists of four elements: attachment, commitment,
38
involvement, and belief. Attachment, defined as the affectionate ties that one person
has with another person or a group of people, occurs when one cares about the
wishes and expectations of others. Someone who develops attachment to others is
less likely to engage in deviant activities than those who do not care about what
people think about them when they offend others. Commitment, defined as the stake
that a person has in society, occurs when someone invests lots of time and energy
into activities such as getting an education, building a business, or simply acquiring a
good reputation. People who have worked hard to obtain the things that are in their
possession are not likely to risk that investment by engaging in deviant activities.
The third control on social behavior is involvement in conventional activities.
A person who is involved in many activities, such as having to attend meetings or
meeting deadlines for a project, is too busy to think about engaging in deviant
activities. Finally, belief is the idea that one should conform to societal rules. The
extent to which people conform to societal rules depends on the how much they
believe in obeying these rules. If someone firmly believes that certain deviant acts
are wrong then participation in those acts would be unthinkable to them.
Parental supervision is generally used within the family literature as a proxy
for the social control explanation. This explanation is composed of several
components – parental monitoring and parental involvement. Parental monitoring
generally represents the level of parental presence around the household (Jacobson &
Crockett, 2000). Parental involvement, which generally represents the frequency of
the parent-child interaction, may be a proxy of the parent-child relationship quality
(Moore & Chase-Lansdale, 2001). This component consists of measures such as the
39
amount of parent-children’s activities and the quality of the parent-child
communication (Miller, Forehand, & Kotchick, 1999).
Even though the parental monitoring and parental involvement variables
represent two different dynamics, Wu & Thomson (2001) have argued that these
variables are analogous because parents cannot constantly supervise their children
and that effective parental social control depends on other basis such as the parent-
child attachment. In addition, studies by Newcomer & Udry (1983) and Thornton
(1991) have found that parental involvement is an important mediating influence for
parents to maintain social control over their children.
In general, researchers have found that parents who are constantly present at
home are more likely to monitor and discipline their children than those who are
rarely present (Whitbeck, Simons, & Goldberg, 1996). Accordingly, the amount of
time parents spend to supervise their children seems to be a function of family
structure where children who reside with two parents receive more supervision than
those who reside only with a single parent (Thomson, McLanahan, & Curtin, 1992).
Some researchers have found that lower parental supervision levels are associated
with early adolescent sexual activity (Small & Luster, 1994; Jacobson & Crockett,
2000). However, other researchers have found that this relationship may be
curvilinear with adolescents who engage in early sexual activity being associated
only with parents who exercise very strict supervision (Miller et al., 1986).
Newcomer & Udry (1987) and Thornton (1991) have theorized several
reasons, other than the number of parents at home, for differences between one and
two parent households regarding frequency of monitoring of the children’s activity
40
or the disciplining of the children’s behaviors. One reason is the parent’s gender. In
two parent households, children may find that it is more intimidating to confront
their father than their mother regarding issues such as sexual behaviors since he
could have greater parental authority in the family (Newcomer & Udry, 1987). Since
mothers are usually the custodial parent for the single parent family (Thomson,
McLanahan, & Curtin, 1992), children who live in these households may find their
mothers to be less intimidating and become more willing to challenge her parental
authority. Consequently, this action could lead to a break down in the household’s
hierarchically structure since mothers would have more problems with disciplining
and influencing her children’s behaviors.
Children who reside with a single parent may also receive less parental
discipline and monitoring of their activities by their parents than those who reside in
two-parent households because of the parent’s employment status. These parents
may have to work full time at one job or hold multiple part-time jobs to provide
adequate income for the household since they may be the only financial provider for
the family (Thornton, 1991). In contrast, a parent who is from a two parent
households may have the luxury to either work part time or become a stay at home
parent because there is another adult figure in the household that can help to provide
income for the family. The more time the parents spend at work, the less time they
have to monitor, discipline, and build a relationship with their children. With the
parents being away at work, an empty house may provide adolescents the
opportunity to engage in sexual activity (Zelnick & Kantner, 1980).
41
Researchers have found some support regarding the relationship between
maternal employment and adolescent sexual activity. In a study of unmarried
college students, Hansson et al. (1981) found that females whose mother were
employed outside the home during the high school years had a greater tendency to
begin sexual relations before age 19. Besides being more likely to engage in sexual
behaviors, they were also less concerned about being pregnant. Furthermore,
Jacobson & Crockett (2000) found for adolescents of both genders that full-time
maternal employment was associated with higher likelihood of sexual activity.
In addition, researchers have found some evidence that parents of stepparent
households monitor their children less closely than parents of two biological parent
households (Dornbusch et al., 1985; Thomson, McLanahan, & Curtin, 1992). This
finding may be due to ambiguities concerning the stepparent’s role, duties, and
responsibilities and results in ineffective parenting (Cherlin, 1978; Furstenberg,
1987). Stepchildren also could have difficulty with accepting their stepparent as an
authority figure because the household roles and duties may have been established
before the new parent entered the family. When a new parent enters the household,
new rules would need to be established and family members would have to adjust to
the new family configuration and interactions. Stepchildren may have a difficult
time with these processes which could lead to considerable conflict within the
household and the stepparents feeling that their role as a parent in the household is
ambiguous when attempting to discipline their stepchildren (Cherlin, 1978; Amato,
1987; Furstenberg, 1987). Thus, stepparents may discipline their stepchildren less
strictly in order to maintain peace within the household.
42
Parental Attitude Explanation
Besides being a role model, parents also socialize their children by conveying
to them their own personal attitudes and values. Unlike role modeling where
children learn by observing their parent’s behaviors such as through parental dating,
as described in the family transition section, this explanation contends that children’s
personal attitudes and values are shaped by daily interactions with their parents.
These interactions help in the conveyance of attitudes and values that parents would
want their children to follow so they are able to use them as a guide to navigate
through life when they grow older.
In general, a parent’s attitude in regards to adolescent sexual activity does
differ between the single parent and the two-parent households where unmarried and
divorced parents are more liberal in their attitudes compared to parents who bear and
raise children together in a marriage (Newcomer & Udry, 1984; Thronton &
Camburn, 1987; Thornton, 1991). There are several reasons for these findings.
First, divorced parents frequently encounter intimacy issues within a non-marital
relationship during the course of dating other adults (Thornton, 1991). Although
these pre-marital issues are similar to those that were previously resolved, divorced
parents may have a different view about sex because of their experiences with these
activities as being part of a normative process within a marriage. In addition, they
are aware that the trend in today’s society has become more accepting of non-marital
sexual and cohabitating relationships. The consequence of these events is that one
begins to perceive marriage to be a less relevant means for intimate relations
43
(Thronton, 1989). These events may then lead divorced parents to be more
approving of sexual activity for unmarried people and to convey these attitudes to
their children.
Second, a selection bias may account for unwedded mothers maintaining
more liberal views about sex than parents from two-parent households (Thronton &
Camburn, 1987; Thronton, 1991). According to this view, unwedded mothers may
possess sexual values that are fundamentally different from parents who are married
because of their own personal upbringing. The environment that the unwedded
mother grew up in may strongly influence how she socializes her children, especially
in regards to sexual values. For example, unwedded mothers may be more likely
than mothers who were married before having children to have grown up in families,
cultures, and or neighborhoods that are more accepting of liberal sexual values.
Consequently, their children may be more likely than those who were born within a
marriage to have greater exposure to liberal sexual values while growing up. Thus,
unwedded parents may find it more acceptable for children to engage in sexual
activity during puberty than those who grew up in sexually conservative
environments.
Religion also plays an influential role in the forming a parent’s attitude by
helping to establish an individual’s morality and values (Murry, 1994). Sexual
behaviors are constrained by religion through the fear of imposing religious
sanctions for violating the formal rules as established by religious teachings
(DeLamater, 1981). These religious teachings may help to reinforce a parent’s
attitude and values regarding sexual behaviors or to help establish one if he or she
44
has not yet established an opinion. Consequently, parents may use their religion as
the basis for their own sexual beliefs and values. Since religion has been shown to
be associated with advocating conservative sexual values (Bock, Beeghley, &
Mixon, 1983), very religious parents may be more likely to transmit these attitudes
and values to their children.
In addition to religious values, a parent’s attitude may be associated with his
or her child’s sexual behaviors. A parent’s religion may not necessarily be
associated with his or her attitudes regarding this issue because he or she may
disapprove these types of activities for their children from a practical perspective
such as feeling that they are not yet ready to take on parenthood responsibilities. In
general, the children’s perception of his or her mother’s attitudes in regards to
adolescent sexual activity is predictive of sexual behaviors during puberty (Jaccard et
al., 1996; Dittus & Jaccard, 2000). However, these same researchers also found that
a small number of parents also do not mind at all if their children became sexually
experienced during puberty (Jaccard & Dittus, 1991).
Parent-child sex discussions may also represent a parent’s feeling regarding
adolescent sexual activity. In a manner that is similar to the parent’s attitude
regarding this topic, the parent’s religion may not necessarily influence the decision
to discuss sex topics with his or her children. Besides personal values and beliefs,
these discussions may also depend on whether the parent believes it would prevent or
encourage his or her children’s decision to engage in sexual behaviors during
adolescence. Although many parents believe that it is desirable for them to talk to
children about sex (Fox & Inazu, 1980), they often find that discussing sexual topics
45
with their children during puberty to be uncomfortable in most circumstances
(Brooks-Gunn & Furstenberg, 1989). However, an exception to this finding is
during the time when mothers and their daughters discuss menarche (Brooks-Gunn
& Ruble, 1982).
The results from the literature regarding the relationship between parent-child
sex discussions and adolescent sexual activity has been mixed. Some researchers
have not found a relationship between parental-child sex discussions and adolescent
sexual activity (Hofferth, Kahn, & Baldwin, 1987; McNeely et al, 2002). In contrast,
others have found that children were less likely to be sexually active before marriage
and had fewer sexual partners if the main source of sex education came from their
parents (Fox & Inazu, 1980; Casper, 1990; Holtzman & Rubinson, 1995). Yet,
another study found that parent-child sex discussions are positively associated with
sexual activity during puberty (Thornburg, 1981). One reason for these inconsistent
findings could be that parents and their children may disagree about the truth
regarding the sex discussion content (Newcomer & Udry, 1985). They found that
the child’s and the parent’s recollection about what was said during these discussions
were at odds with each other. Thus, not knowing what really happened during these
discussions may have resulted in ambiguous or conflicting associations for this
relationship.
46
Family Support Explanation
The family support explanation suggests that a child’s decision to engage in
sexual activity during puberty is influenced by feelings of support from other family
members. Adolescence can be a confusing period for children because they
experience numerous simultaneous biological, social, and psychological changes in
their lives during this time (Havighurst, 1967). Parents and other immediate family
members who provide emotional support to children may ease some of the
challenges that they may face. These gestures could help children to develop
feelings of support by people who are close to them. Children who feel supported by
their parent and other immediate family members may have a lower desire to seek
intimate relationships with people outside of the family. However, children who do
not have these experiences may be more likely to seek early independence from the
family due to the unsatisfactory family relationships and to regard sexuality as part
of a rebellious expression (Moore & Rosenthal, 1993).
This explanation is somewhat different from the parental supervision
explanation. The family involvement component of the parental supervision
explanation measures the frequency of the parent-child interaction. However, the
family support explanation measures the quality of the parent-child interaction.
Frequency and quality of the parent-child interaction are not necessarily analogous
because households where frequent parent-child interactions take place may either be
friendly or hostile. In addition, the family support explanation is more of a
subjective measure where an individual’s personal feelings are involved.
47
Meanwhile, the family involvement variables may slightly be a more objective
measure because other sources could confirm the response to these questions.
Empirical studies have not confirmed the popular and clinically influenced
stereotype that the adolescent’s family is a crucible filled with intrafamilial tension
and hostility (Steinberg, 1990). Several large-scale studies that investigated the
relationship between adolescents and their parents found that approximately 75% of
families enjoy warm and pleasant relations during adolescent years (Offer, 1969). In
addition, the vast majority of teenagers are likely to report that they admire, are
willing to listen to, and feel loved and appreciated by their parents (Offer, Ostrov, &
Howard, 1981). Furthermore, another study found that 19-year-old college students
reported to be just as close to their mothers and fathers as they were when they were
fourth graders (Hunter & Youniss, 1982).
While a majority of parents and their adolescent children are getting along
relatively well with each other, empirical studies have shown a mixed relationship
between an adolescent’s attachment to the family and sexual activity. Some
researchers have found that closer parent-child relationships are associated with
delayed sexual debut. For example, Inazu & Fox, (1980) found that adolescents who
felt that the communication between themselves and their parents was good were
less likely to be sexually experienced. In addition, Zelnick, Kantner, & Ford (1981)
found that adolescents who rated their families as “ideal” were likely to delay sexual
initiation by 5 months compared to those who rated their families as “least ideal.”
However, other researchers found inconclusive evidence regarding the relationship
between parent-child relationship quality and virginity status (Ooms, 1981; Miller &
48
Moore, 1990). For example, Newcomer & Udry (1983) found that closeness and
communication between parent and child have little or no effect on adolescent sexual
activity. Furthermore, Weinstein & Thornton (1989) did not find a relationship
between parental communication and parental values. They also found that warm
and supportive parent-child relationships do not transmit these values.
Yet, other researchers have found that conflicts between parent and
adolescent do not necessarily lead to increasing risk for sexual activity during
puberty (Taris & Semin, 1997). This finding may be explained by viewing
intrafamily conflict as an opportunity for the parent and their children to exchange
views regarding the proper conduct in a particular situation rather than being an
indicator of increased family instability (Cooper, 1988; Steinberg, 1990). From this
perspective, parent-child conflicts may be a forum where each person is able to
understand the other person’s point of view regarding sexual behaviors by conveying
what and how strongly they each feel about these issues. Thus, these conflicts may
be viewed as functional rather than destructive.
Researchers have also theorized that family conflict levels may differ among
the various households. For single parent households, the absence of another adult to
help regulate the household, maintain authority, and uphold conventional standards
makes it difficult for these parents to establish order within the family (Smetana,
1989) and may result in greater parent-child conflicts (Weiss, 1979; Hetherington,
1989). In addition, studies have shown that adolescents who reside with a single
parent are more autonomous and independent than those who reside with two
parents. For example, Dornbusch et al. (1985) found that adolescents who reside
49
with a divorced parent have greater control over decisions regarding clothing,
curfew, friend choice, and spending money than those who reside with two parents.
Furthermore, adolescents who reside with a single parent are more likely than those
who reside with two biological parents to have responsibility for household tasks
(Hetherington & Camara, 1984). The absence of parents to reinforce the family
hierarchy along with the greater independence experienced by adolescents who
reside in single parent households may result in greater parent-child conflict
whenever parents try to assert their authority. These factors may result in children
feeling that their parent does not support them and could compel them to seek
intimate relationships with others outside of the family whenever conflicts with their
parent occur.
Researchers have found differences in parental support levels for two parent
households as well. Thomson, McLanahan, & Curtin (1992) found that stepparents
are less likely than married parent households to report frequent positive responses
with their children, e.g. hugging and praising. As mentioned previously, this finding
may be due to ambiguities concerning the stepparent’s roles, duties, and
responsibilities (Cherlin, 1978; Furstenberg, 1987). These factors may result in
ineffective parenting, greater conflict with the stepchildren, and feeling of less
support by family members whenever the stepparent tries to assert his or her
authority.
50
Peer Influence Explanation
The peer influence explanation may be viewed from both functional and
conflict perspectives. The functional perspective views the peer group as a vibrant
organization that provides a source for esteem, acceptance, status, and self-identity to
adolescents that is not provided by immediate family members. Since all adolescents
are involved with the peer culture to some degree, the peer group may also exert an
independent influence on their socialization regardless of the parent-child
relationship quality. This independent influence is due to adolescents spending an
average of 22 hours a week with their friends and other peers which is more than the
time they spend with their families or by themselves (Csikszentmihalyi & Larson,
1984). As a result, adolescents tend to choose friends who share similar interests,
values, beliefs, and attitudes because they believe that these individuals are more
likely to be supportive and understanding (Youniss & Smollar, 1985; McCord,
1990).
The conflict perspective views peers as the conduit through which children
are able to obtain more power within the household (Collins, 1975). Between age
six and puberty, children are big enough to help adults, but they are also small
enough so that the adults are still able to be in control of their lives. During this
time, children are not able to fend for themselves from adults by running away, but
they are able to spend time away from their parents when they grow older. Spending
time away from the parents allows the possibility for childhood peer groups to
appear. These groups help children to reduce the psychological dependence on
51
adults and begin the process of acquiring more power within the family. In addition,
schools play a role in the children’s emancipation from their parents by serving as
the collective caretaking institution that frees the parents to do other activities while
greatly facilitating the formation of peer groups.
One question that has been investigated in the family literature is whether
adolescents are “pulled” away from parental influence by peers who engage in
deviant activities because they find these behaviors attractive or whether they are
“pushed” toward being influenced by these peers due to an absence of emotional
support by, or being in constant conflict with, immediate family members.
Devereaux (1970) found that the deviant crowd does not pull adolescents away from
their parents. Instead, adolescents are attracted to this crowd by their parents’
ineffective child-rearing practice in cases where peer group norms are in obvious
opposition to prevailing adult norms such as those found in delinquent gangs. In a
longitudinal study, Patterson & Dishion (1985) found that the inability of parents to
teach their children effective social skills contributed significantly to their
involvement with delinquent adolescent peer groups. The adolescent’s perception
that the relationship with their parent is negative or is deficient in support and
guidance magnifies the friend and peer’s ability to influence the child’s attitude and
behaviors. Thus, parents may help to create an environment that is more conducive
for their children to seek solace somewhere else (Savin-Williams & Berndt, 1990).
Adolescents also can be “pulled” by peer influence due to the need to
evaluate one’s own experiences and to create a sense of identity and belonging.
These needs may cause adolescents to feel the pressure to conform to a particular
52
peer group or risk being rejected by them. Studies have shown that young people are
sensitive to peer pressure during early adolescence, but their willingness to go along
with the group depends on the particular issues and the peer type in question
(Landsbaum & Willis, 1971; Savin-Williams & Berndt, 1990). However,
researchers have also found that an adolescent’s susceptibility to peer pressure tends
to decline with age (Brown, Clasen, & Eicher, 1986; Gibson & Kempf, 1990).
Regardless of whether the reason for the peer influence is due to family
conflict or to the need for establishing a self-identity, studies generally have found
that adolescents who have acquaintance with friends or peers who engage in deviant
activities are more likely to engaging in sexual activity during puberty. This finding
applies to white (Metzler et al., 1994; Whitbeck, Simons, & Goldberg, 1996), black
(Furstenberg et al., 1987), Hispanic (Gibson & Kempf, 1990), and Asian (Schuster et
al., 1998) adolescents. Chapter 5 will present a more in-depth discussion of how
peer influences may vary by race.
53
Chapter 3: Methodology
Description of the Dataset
Both of the investigations for this study were conducted on the National
Longitudinal Study of Adolescent Health (Add Health) Wave I dataset, which is a 1-
year longitudinal study of 7
th
to 12
th
grade students in the United States. The dataset,
which was compiled between August 1994 and April 1995, consists of a sampling
frame based on all schools in the United States with an 11
th
grade with an enrollment
of at least 30 students. The compilation began with an initial sample of 80 schools
that is selected from a database of 26,666 schools stratified by geographic region,
community size, urbanicity, school type, white student percentage, black student
percentage, and grade span. The initial sample also included a set of “feeder”
schools for those without a 7
th
grade. The final dataset sample consisted of 134
schools with a total enrollment of 119,234 students.
The interviewers for the survey asked over 90,000 students to answer an in-
school questionnaire regarding topics such as family and peer relationships, sports,
academic, and student activities, and adolescent risk behaviors. A random sample of
15,243 students was chosen from the initial sample to complete a 90-minute in-home
interview. The final sample consisted of 12,105 students. The Add Health dataset
also included an oversample of 1,038 black adolescents whose parents have a college
degree, 334 Chinese adolescents, 450 Cuban adolescents, and 437 Puerto Rican
adolescents.
54
The responses from both the child and his or her parent during the in-home
interview were recorded into a laptop computer. During the less sensitive portions of
the child questionnaire, the interviewer asked the question and recorded the child
respondent’s answer. However, during the more sensitive portion of the
questionnaire, the respondent listened to the prerecorded questions through
earphones and entered his or her answers directly into the computer. The residential
parental figure was not present in the room during the interview with the child. He
or she then answered the parental questionnaire after the child interview was
completed.
The Carolina Population Center, which is affiliated with the University of
North Carolina, distributes two versions of the Add Health dataset. The restricted
version contains all the responses from the child and their parents who completed the
in-home interview. However, the public version, which consists of approximately
6,500 adolescent and 5,485 parent respondents, contains half of the respondents
chosen at random and half of the black adolescent oversample from the restricted
version. The data analyses for both of the investigations are conducted from the
public version.
Study Sample
The sample for this study consisted of white, black, Hispanic, and Asian
adolescents of both genders and their biological mothers. This sample consisted of
approximately 65% white adolescents, 22% black adolescents, 11% Hispanic
55
adolescents, and 2% Asian adolescents. The children’s age in this study ranged from
12 to 21 years old.
This study excluded the responses from married adolescents, widowed
mothers, and households with mothers who have a female lover. In addition, this
study excluded the responses from biological fathers or father figures who completed
the parental questionnaire.
Study Measures: Dependent Variable
The dependent variable for this study is the sexual activity status of the child
respondent. This variable came from the question asking the child, “Have you ever
had sexual intercourse?” The responses to this question were converted into a
multiple categorical variable where 0 indicates that he or she was a virgin and 1
indicates that he or she was sexually experienced at the time of the interview.
Study Measures: Independent Variables
The independent variables for this study were grouped into demographic,
family transitions, parental social control, parental attitudes, family support, and peer
influence measures.
56
Demographic Variables
The demographic variables for in this study were the child’s age, child’s
gender, child’s race, household configuration, mother’s education level, and
household income. Child’s age was determined by subtracting the year which the
respondent completed the questionnaire from his or her birth year. Child’s gender
came from the question that asked the interviewer to confirm the respondent’s sex.
This variable was converted into a multiple categorical variable where females were
used as the referent group.
Child’s race came from the question that asked whether the respondent’s race
is white, black, American Indian, Asian, or other. The Hispanic sample came from
the question that asked the child whether he or she is of Hispanic or Latino origin.
For this study, those who answered affirmative were grouped into the Hispanic
category even though they are also able to choose white, black, American Indian,
Asian, and other racial categories later on in the questionnaire.
Respondents who considered themselves to be from more than one racial
group were able to choose more than once racial category. A follow-up question
then asked the child to choose the one racial category that best described his or her
background for those who chose more than one racial category. Mixed race
respondents who considered themselves to be mainly from white, black, or Asian
backgrounds were added to the tally of their respective single race peers. This
method was implemented to prevent the double counting of respondents. In
addition, respondents who chose “other” or do not know their racial background
57
were group into the study sample according to the interviewer’s observation of the
child’s race. This study excluded respondents and interviewers who gave American
Indian, other, and missing responses.
The family structure variables for this study consisted of four household
types: two- biological parents, unwedded mother, divorced/separated parent, and
stepparent households. The household type is determined from the answers to the
questions asking (1) about the biological mother’s marital status, (2) whether the
biological father resides in the household, and (3) about the spouse’s gender for
married household. This study excluded children respondents who reside with an
unwedded, separated, and divorced mother if the biological father was still living in
the household. For the two parent households, parents who did not mention the sex
of the spouse or who mentioned that the sex of the spouse was a female were also
excluded from this study. Two biological parent households were the referent group
for this study.
This study used several socioeconomic status measures. The first measure is
mother’s education. Mothers who answered this question were grouped into five
educational attainment categories: less than high school education, high school
graduate, some college, college graduate, and post-graduate. This question ranged
from 1 to 6 where 1 indicates that the mother has less than high school education and
6 indicates that the mother has at least some post-graduate education.
The second socioeconomic status measure is the total 1994 household
income. This measure included the mother’s income, other household member’s
income, welfare benefits, stock dividends, and all other income sources. The natural
58
logarithm of the raw household income was used for the data analyses. Respondents
who did not report a household income were assigned the mean household income
for their particular racial subgroup. A multiple categorical variable was then used to
indicate respondents with an imputed household income.
Family Transition Measure
The family transition measure for this study came from the question asking
the biological mother “In the past 18 years, how many marriage and marriage-like
relationships have you had?” The response to this question ranged from 0 to 6 where
0 indicates that she had no and 6 indicates that she had 6 or more marriages or
marriage-like relationships in the past 18 years. This variable was not manipulated
in any way throughout this study.
There are several reasons why one needs to be cautious when interpreting the
results of this variable. First, the number of different relationships the mother
experienced in the past 18 years may not necessarily be the same amount of family
transition events that the child experienced during his or her lifetime. Since the ages
of children in this study ranged from 12 to 21 years, there is the possibility that the
younger children in this study are less likely to experience changes in their families
during their lifetime because some of the family transition events may have occurred
before they were born. In addition, some of the younger children may have lived
through multiple family transition events but these events could have occurred when
they were very young so that their impact on their lives may be minimal.
59
Consequently, there may be a question as to whether this measure accurately
captured the stresses that children may have experienced when their household
transformed from one form to another since there is a possibility that these events
may not have taken place within the child’s lifetime or that the impact of these
events may be minimal, especially for younger adolescents.
Second, the extent as to which this variable portrays the family transition
event is limited because it only measures the frequency and not the quality or degree
of the disruption. For example, a family transition event may have more impact on
children for those who reside with parents that only have one or two marriage-like
relations where the break up of the relationship were tumultuous and conflict ridden
compared to the process where the parent tries to establish a stable household during
these times even though they were married several times. Thus, this measure is at
best a very crude approximation for the effect that family transition events may have
on adolescent sexual activity.
Parental Supervision Measure
The parental supervision measure for this study is constituted by eight
variables, seven of which came from the child questionnaire. The variables that
represent parental involvement are mother-child activities, mother-child discussions,
parental strictness, and parental meal presence. The variables that represent parental
monitoring are mother’s employment status and mother’s presence around the house
during morning, afternoon, and evenings.
60
The mother-child activity variable came from the child’s responses to the
questions asking whether he or she did certain activities with his or her mother
during the past four weeks. These activities are (1) shopping, (2) playing a sport, (3)
attending religious service or church related event, (4) going to a movie, play,
museum, concert, or sports event, and (5) working on a project for school. The
mother-child discussion variable came from the child’s responses to the questions
asking whether he or she talked about certain discussion topics with his or her
mother during the past four weeks. These discussions are (1) talking about someone
the adolescent was dating or about a party that he or she went to, (2) talking about a
personal problem the adolescent was having, (3) arguing about the adolescent’s
behavior, (4) talking about schoolwork or grades, and (5) talking about other things
the adolescent was doing in school. The responses to these questions ranged from 0
to 5 where 0 indicates that none of these and 5 indicates that all five of these
activities or discussion topics took place between the respondent and his or her
mother during the past four weeks.
The parental strictness variable came from the child’s responses to the
questions asking whether he or she is allowed to make decisions about (1) the people
to hang around with, (2) what to wear, (3) how much television to watch, (4) which
television programs to watch, (5) what time he or she goes to bed on weeknights, and
(6) what to eat. The responses to these questions ranged from 0 to 6 where 0
indicates that none of these and 6 indicates that the respondent made all of these
decisions. This variable was reverse coded so that higher scores indicate stricter
household rules.
61
The parental presence during meals variable came from the child’s response
to the question asking about the number of days in the past week when at least one
parent was in the room during an evening meal. The response to this question
ranged from 0 to 7 where 0 indicates that the respondent ate alone and 7 indicates
that the respondent ate his or her dinner with a parent every day in the past week.
The mother’s employment status variable came from the question asking
whether she is currently employed full time. The response to this question were
coded into a multiple categorical variable where 0 indicates that she is not working
or is part-time employed and 1 indicates that she is currently full-time employed.
The mother’s presence at home variables came from the child response to the
questions asking about his or her mother’s household presence when he or she (1)
leaves for school, (2) returns from school, and (3) goes to bed. The responses to
these questions ranged from 1 to 5 where 1 indicates that she was always at home
and 5 indicates that she was never at home during these times. These variables were
reverse coded so that higher scores indicate greater mother’s household presence.
Parental Attitude Measure
The parental attitude measure for this study is constituted by three variables -
mother’s religiosity, sex topic discussion with her children, and her attitudes in
regards to adolescent sexual activity. The mother’s religiosity variable is represented
by two questions. The first question asked the mothers about her attendance at
religious services. The response to this question ranged from 1 to 4 where 1
62
indicates that she attends religious service at least once a week and 4 indicates that
she has not attended religious services in the past year. The second question asked
the mothers about the importance of religion in her life. The response to this
question also ranged from 1 to 4 where 1 indicates that religion is very important and
4 indicates that religion is not important to her. The responses to these questions
were reverse coded so that higher scores indicate greater mother’s religiosity. These
responses were added to form a mother’s religiosity scale with a 0.74 Chronbach
alpha.
The mother-child sex discussion variable is represented by the question
asking about how often she discusses sex topics with her child. The response to this
questions ranged from 1 to 4 where 1 indicates that she did not talk about these
subjects at all and 4 indicates that she talked a great deal about these subjects with
her child.
The mother’s attitude in regards to adolescent sexual activity variable is
represented by the question asking about how much she disapproved of her child
currently being sexually active at this time. The response to this question ranged
from 1 to 5 where 1 indicates that she strongly disapproves and 5 indicates that she
strongly approves of her child being sexual active at this time. This question was
reverse coded so that higher scores indicate greater mother’s disapproval regarding
this topic.
63
Family Support Measure
The family support measure is represented by the child’s responses to four
questions. These questions asked whether the child whether or not he felt that (1)
their parents cared about them, (2) the people in the family understood them, (3) they
had fun together with their family, and (4) their family paid attention to them. The
responses to these questions ranged from 1 to 5 where 1 indicates that the respondent
felt no support and 5 indicates that the respondent felt lots of support by immediate
family members. These questions were added to form a family support scale with a
0.78 Chronbach alpha.
Peer Influence Measure
The peer influence measure is represented by the child’s responses to the
questions asking about the number of friends or acquaintances that had engaged in
such activities as 1) smoking cigarettes once a day, 2) drinking alcohol at least once
a month, and 3) using marijuana at least once a month. The responses to these
questions ranged from 0 to 3 where 0 indicates that the respondent does not know
anyone and 3 indicates the respondent have acquaintances with three friends who
have engaged in these behaviors. These responses were added to create a peer
influence variable with a 0.75 Chronbach alpha.
64
Data Analysis
The data was analyzed using SPSS 11.0 for Microsoft Windows. Several
statistical methods were used to analyze the data. First, t-test analyses were used to
examine how the adolescent’s and household’s characteristics may differ in various
contexts. Second, logistic regression analyses were used to determine both the best
explanation for the family structure effect among the three non-traditional
households and which factors were directly and indirectly associated with adolescent
sexual activity.
Statistical Explanations
This study uses three statistical calculations that would require an extra
explanation: the Nagelkerke R
2
(Nagelkerke, 1991), the Bayes Information Criterion
(BIC) statistic (Raferty, 1995), and t-test between slopes (Jaccard, Turrisi, & Wan,
1990). Logistic regression does not have a measure of model variance that is similar
to those found for multiple ordinary least square regressions. However, Nagelkerke
R
2
is one of a number of methods that researchers have proposed to determine the
model variance for logistic regressions. Consequently, this study used Nagelkerke
R
2
as a proxy for the model variance for the logistic regression analyses.
This study also used the BIC statistic to determine the most parsimonious
explanation for the regression models. The model’s variance, or R
2
, is not a reliable
indicator for determining which explanation(s) may have the strongest association
65
regarding adolescent sexual activity because each measure could be comprised of a
different number of independent variables. Since model variance tends to increase
when one adds more independent variables to the regression model, a regression
model with a larger variance may not necessarily be a better explanation for
adolescent sexual activity than a regression model with a smaller variance if it uses
more independent variables.
Model parsimony, or how efficiently the regression model explains the
dependent variable, will be the method used to determine which explanation has a
stronger relationship with adolescent sexual activity for this study. This study used
the BIC statistic to determine model parsimony for the logistic regression analyses.
The BIC statistic formula for logistic regression is -
2
+ p ln N, where
2
is the
model chi-square, p is the number of independent variables, and N is the total sample
size. According to this formula, the most parsimonious logistic regression model is
the one with the lowest BIC statistic.
This study also used the t-test between slope analyses (Jaccard, Turrisi, &
Wan, 1990) in chapters 6 through 9 to determine whether there are significant
differences for the relationship between the independent variable and adolescent
sexual activity between two different subgroups. The formula to determine
significant statistical differences between the independent variable betas of two
subgroups is as follows:
b
1
– b
2
t =
2
2
2
1
) . . ( ) . . ( e s e s +
where t is the t-score for difference between the slopes, b
1
is the variable’s beta in
66
group 1, b
2
is the variable’s beta in group 2, s.e.
1
is the variable’s standard error in
group 1, and s.e.
2
is the variable’s standard error in group 2. The level of significant
difference is set at a t-score of greater than 1.65 (p < .05, one-tailed).
In addition, the terms statistical and partial mediation are used throughout
this study to note whether a particular explanation fully or partially respectively
account for a particular non-traditional household effect. Statistically mediated is
used in this study to signify that a particular non-traditional household effect went
from a statistical significance to non-significance after accounting for a particular
influence. In another words, the term statistical mediation is used to indicate that a
particular influence fully accounts for the reason why adolescents who reside in that
particular non-traditional households are more likely to engage in sexual activity
than those who reside in two biological parent households.
The term partial mediation is also intended to help provide a better
understanding of the non-traditional household effect. Partial mediation is used in
this study to signify that the particular non-traditional household effect did not go
from a statistical significance to statistical non-significance after controlling for a
particular influence. Although these explanations are not as strong as those that
statistically mediate these effects, there still may be some social significance for this
finding. The threshold for partial mediation that is used throughout this study is
arbitrarily set at least a one-quarter decrease of a particular non-traditional
household’s beta after controlling for the family influence or peer influence measure
from this particular non-traditional household’s beta in the baseline model.
67
Chapter 4: Total Sample Analyses
There are two objectives for this investigation. The first objective is to
determine which of the five explanations best explains the family structure and
adolescent sexual activity relationship for three non-traditional households on the
total sample. The second objective is to determine which variables that comprise the
five explanations have a direct association with adolescent sexual activity. This
chapter contains the hypotheses, the analyses of the data using t-tests and logistic
regressions, and a discussion of the findings.
Hypotheses
The following hypotheses are proposed for the relationship between the five
explanations and adolescent sexual activity.
1. Family transition will be positively associated with sexual activity.
2. All of the parental supervision variables will be negatively associated with
sexual activity.
3. All of the parental attitude variables will be negatively associated with sexual
activity.
4. The family support measure will be negatively associated with sexual
activity.
5. The peer influence measure will be positively associated with sexual activity.
68
Descriptive Results
Table 4-1 presents the mean scores for the demographic, family influence,
and peer influence variables by adolescent sexual status. The results show numerous
demographic differences between virgin and sexually experienced adolescents.
Sexually experienced adolescents are more likely than virgin adolescents to be
males, to be black, and to reside in all the non-traditional households. In contrast,
virgin adolescents are more likely than sexually experienced adolescents to be white,
to be Asian, to reside in households with greater family incomes, and to have
mothers who achieved higher education levels.
Among the family and peer influence variables, mothers of virgins are more
likely than mothers of sexually experienced adolescents to be present around the
home throughout the day and night, to be religious, and to disapprove of adolescent
sexual activity. Mothers of virgins are less likely than mothers of sexually
experienced adolescents to have multiple serious relationships or marriages in the
past 18 years. Virgin adolescents are more likely than sexually experienced
adolescents to eat meals with at least one parent present, to have stricter parents, and
to feel more family support. Conversely, sexually experienced adolescents are more
likely than virgin adolescents to have both general and sex discussions with their
mothers and to have greater acquaintances with peers who engage in deviant
activities.
Table 4-2 presents the mean scores for the adolescent sexual activity,
demographic, family influence, and peer influence variables by household status.
69
Table 4-1: Mean Scores for the Demographic, Family Influence, and Peer Influence
Variables By Adolescent Sexual Status
No Sexual Experience Sexually Experienced
Child's Age 15.39 16.80 ***
Male 0.45 0.50 **
White 0.69 0.58 ***
Black 0.17 0.30 ***
Hispanic 0.11 0.10
Asian 0.03 0.02 ***
Two Biological Parents 0.67 0.48 ***
Unwedded Mothers 0.04 0.08 ***
Divorced/Separated Mothers 0.17 0.27 ***
Stepparent 0.12 0.17 ***
Mother's Education 2.89 2.66 ***
Household Income 52.75 41.59 ***
Missing Income 0.10 0.10
Family Transitions 1.25 1.40 ***
Mother-Child Activities 1.83 1.38 ***
Mother-Child Discussions 2.27 2.69 ***
Parent Strictness 1.36 0.08 ***
Parent Meal Presence 5.18 3.98 ***
Mother's Employment Status 0.56 0.63 ***
Mother's Morning Presence 4.25 4.10 ***
Mother's Afternoon Presence 3.34 3.15 ***
Mother's Evening Presence 4.69 4.62 **
Mother's Religiosity 6.37 6.09 **
Mother-Child Sex Discussions 2.88 3.24 ***
Mother Disapprove Sex 4.55 4.11 ***
Family's Support 16.60 15.62 ***
Deviant Peers 1.61 3.89 ***
N 2554 1462
Notes: * = p < .05; ** = p < .01; *** = p < .001
70
Table 4-2: Mean Scores for the Adolescent Sexual Activity, Demographic, Family Influence, and Peer Influence Variables By
Household Status
Two Biological Unwedded Divorced Stepparents Unwedded/Divorced Difference
Sexual Experience 0.29 0.51 *** 0.48 *** 0.44 ***
Child's Age 15.89 15.76 16.01 15.88
Male 0.47 0.42 0.47 0.49
White 0.74 0.18 *** 0.55 *** 0.63 *** ***
Black 0.13 0.68 *** 0.32 *** 0.23 *** ***
Hispanic 0.09 0.13 0.12 0.12
Asian 0.03 0.01 *** 0.01 *** 0.02 *
Mother's Education 2.91 2.39 *** 2.72 *** 2.65 *** ***
Household Income 58.08 21.93 *** 29.75 *** 47.07 *** ***
Missing Income 0.11 0.11 0.09 0.10
Family Transitions 1.03 1.30 ** 1.63 *** 2.02 *** ***
Mother-Child Activities 1.72 1.63 1.60 ** 1.58 **
Mother-Child Discussions 2.29 2.55 ** 2.65 *** 2.60 ***
Parent Strictness 1.17 1.43 * 1.05 * 1.18 **
Parent Meal Presence 5.04 4.36 *** 4.10 *** 4.55 ***
N 2402 230 818 566
Notes: * = p < .05; ** = p < .01; *** = p < .001
71
Table 4-2: Continued
Two Biological Unwedded Divorced Stepparents Unwedded/Divorced Difference
Mother's Employment Status 0.54 0.59 0.68 *** 0.62 *** *
Mother's Morning Presence 4.24 4.16 4.05 ** 4.21
Mother's Afternoon Presence 3.33 3.46 3.07 *** 3.21***
Mother's Evening Presence 4.69 4.57 * 4.61 * 4.66
Mother's Religiosity 6.45 6.07 ** 5.96 *** 6.03 ***
Mother-Child Sex Discussions 2.90 3.30 *** 3.20 *** 3.12 ***
Mother Disapprove Sex 4.48 4.10 *** 4.27 *** 4.31 **
Family's Support 16.41 16.48 15.94 *** 15.88 *** *
Deviant Peers 2.23 2.36 2.96 *** 2.65 *** **
N 2402 230 818 566
Notes: * = p < .05; ** = p < .01; *** = p < .001
72
Adolescents who reside in non-traditional households are more likely than those who
reside with two biological parents to have engaged in sexual activity. White and
Asian adolescents are more likely to reside in a two biological parent households
than any of the non-traditional household. The opposite is true for black adolescents.
Non-traditional household mothers are more likely than biological mothers of two
parent households to report lower educational achievement and household income.
The table also shows some general family influence differences between non-
traditional and the two biological parent households. Adolescents who reside in non-
traditional households are more likely than those who reside with two biological
parents to have general discussions with their mothers. However, adolescents who
reside in non-traditional households are less likely than those who reside with two
biological parents to eat meals with a parent. Non-traditional household mothers are
more likely than biological mothers of two parent households to have had more
serious relationships over the past 18 years, to have sex discussions with their
children, and to be more tolerant of adolescent sexual activity. Conversely, non-
traditional household mothers are less likely than biological mothers of two parent
households to be religious.
There are also family and deviant peer influence differences between specific
non-traditional households and two biological parent households. Unwedded
mothers are more likely than biological mothers of two parent households to
bestricter with their children. However, unwedded mothers are less likely than
biological mothers of two parent households to be present at home during evenings.
73
Adolescents who reside with divorced/separated mothers are less likely than
those who reside with two biological parents to experience doing activities with their
mother, to have a stricter parent, to have the mother present at home throughout the
day, and to feel supported by family members. However, adolescents who reside
with divorced/separated mothers are more likely than those who reside with two
biological parents to be acquainted with friends who smoke, drink alcohol, and do
drugs. Divorced/separated mothers are also more likely than biological mothers of
two parent households to be full-time employed.
Finally, adolescents who reside with stepparents are less likely than those
who reside with two biological parents to have engaged in activities with their
mothers and to feel supported by their immediate family members. However,
adolescents who reside with stepparents are more likely than those who reside with
two biological parents to be acquainted with friends who smoke, drink alcohol, and
do drugs. Mothers of stepparent households are also more likely than biological
mothers of two parent households to be full-time employed.
This table also shows demographic differences between unwedded mother
and divorced/separated mother households. White adolescents are more likely to
live with divorced/separated mothers than with unwedded mothers. The opposite
was found for black adolescents. In addition, divorced/separated mothers are more
likely than unwedded mothers to achieve higher educational levels. On average,
divorced/separated mothers are also wealthier than unwedded mothers.
Among the family influence and peer influence variables, adolescents who
reside with unwedded mothers are more likely than those who reside with
74
divorced/separated mothers to have a stricter mother, to have the mother be present
at home during afternoons, and to feel supported by family members. However,
adolescents who reside with unwedded mothers are less likely than those who reside
with divorced/separated mothers to have friends who drink, smoke, and do drugs.
Divorced/separated mothers are more likely than unwedded mothers to have a
serious relationship over the past 18 years and to be employed full time.
Test of Hypotheses Results
Table 4-3 presents the sexual activity regressions on the demographic, family
influence, and peer influence variables for the total sample. Among the demographic
variables, the baseline model (model 1) shows that being black and adolescents who
reside in non-traditional households are positively associated with sexual activity.
Being Hispanic, mother’s educational attainment, and household income are
negatively associated with sexual activity. The analysis also shows that both the
black x gender and Hispanic x gender variables are positively associated with sexual
activity.
Among the family and peer influence variables, mother-child activity,
parental strictness, parental meal presence, mother’s religiosity, mother’s disapproval
of adolescent sexual activity, and family support are negatively associated with
sexual activity. However, mother’s serious relationships over the past 18-years,
mother-child discussions, mother-child sex discussions, and acquaintance with
deviant peers are positively associated with sexual activity.
75
Table 4-3: Logistic Regressions of Sexual Activity on the Demographic, Family
Influence, and Peer Influence Variables for the Total Sample (N = 4016)
Model 1 Model 2
Beta Exp (B) Beta Exp (B)
Intercept -9.19 *** -9.59 ***
Child's Age 0.57 *** 1.76 0.57 *** 1.77
Male
a
-0.06 0.94 -0.06 0.94
Black
b
0.35 ** 1.42 0.40 ** 1.50
Hispanic
b
-0.52 ** 0.59 -0.51 ** 0.60
Asian
b
-0.66 0.51 -0.65 0.52
Unwedded Mothers
c
0.68 *** 1.97 0.60 ** 1.81
Divorced/Separated Mothers
c
0.68 *** 1.98 0.54 *** 1.72
Stepparents
c
0.65 *** 1.92 0.42 *** 1.53
Mother's Education -0.13 *** 0.88 -0.13 *** 0.88
Household Income (ln) -0.13 * 0.88 -0.13 * 0.88
Missing Household Income -0.08 0.92 -0.06 0.94
Black*Gender 0.67 *** 1.95 0.66 *** 1.94
Hispanic * Gender 0.67 ** 1.96 0.68 *** 1.97
Asian * Gender 0.09 1.09 0.09 1.09
Family Transitions 0.23 *** 1.26
-2 LL 4340.07 4324.34
Chi-Square 926.59 *** 942.32 ***
Nagelkerke R
2
0.28 0.29
BIC -810.42 -817.85
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= females as the reference group;
b
= whites as the reference group;
c
= two biological parents as the reference
group
76
Table 4-3: Continued
Model 3 Model 4
Beta Exp (B) Beta Exp (B)
Intercept -6.24 *** -8.35 ***
Child's Age 0.46 *** 1.59 0.54 *** 1.72
Male
a
0.04 1.04 0.03 1.03
Black
b
0.39 ** 1.48 0.42 ** 1.52
Hispanic
b
-0.50 ** 0.61 -0.42 * 0.66
Asian
b
-0.53 0.59 -0.45 0.64
Unwedded Mothers
c
0.51 ** 1.66 0.43 * 1.54
Divorced/Separated Mothers
c
0.43 *** 1.53 0.50 *** 1.65
Stepparents
c
0.51 *** 1.67 0.51 *** 1.67
Mother's Education -0.15 *** 0.86 -0.09 * 0.91
Household Income (ln) -0.21 *** 0.81 -0.14 * 0.87
Missing Household Income -0.05 0.95 -0.10 0.91
Black*Gender 0.70 *** 2.02 0.65 *** 1.92
Hispanic * Gender 0.68 ** 1.98 0.67 ** 1.95
Asian * Gender -0.11 0.89 0.10 1.11
Mother-Child Activities -0.29 *** 0.74
Mother-Child Discussions 0.23 *** 1.26
Parent Strictness -0.16 *** 0.85
Parent Meal Presence -0.09 *** 0.92
Mother's Employment Status 0.09 1.09
Mother's Morning Presence -0.06 0.95
Mother's Afternoon Presence -0.04 0.96
Mother's Evening Presence -0.04 0.96
Mother's Religiosity -0.12 *** 0.88
Mother-Child Sex Discussions 0.36 *** 1.43
Mother Disapprove Sex -0.21 *** 0.81
-2 LL 4134.17 4202.23
Chi-Square 1132.48 *** 1064.43 ***
Nagelkerke R
2
0.34 0.32
BIC -949.92 -923.36
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= females as the reference group;
b
= whites as the reference group;
c
= two biological parents as the reference
group
77
Table 4-3: Continued
Model 5 Model 6
Beta Exp (B) Beta Exp (B)
Intercept -6.97 *** -9.11 ***
Child's Age 0.56 *** 1.75 0.49 *** 1.64
Male
a
-0.05 0.95 -0.07 0.93
Black
b
0.37 ** 1.44 0.83 *** 2.30
Hispanic
b
-0.57 ** 0.57 -0.32 0.73
Asian
b
-0.72 0.49 -0.64 0.53
Unwedded Mothers
c
0.70 *** 2.00 0.57 ** 1.78
Divorced/Separated Mothers
c
0.64 *** 1.89 0.50 *** 1.66
Stepparents
c
0.59 *** 1.80 0.57 *** 1.76
Mother's Education -0.13 *** 0.88 -0.11 ** 0.89
Household Income (ln) -0.13 * 0.88 -0.13 * 0.88
Missing Household Income -0.05 0.95 -0.07 0.93
Black*Gender 0.74 *** 2.09 0.56 ** 1.76
Hispanic * Gender 0.75 ** 2.12 0.57 * 1.76
Asian * Gender 0.03 1.03 0.15 1.16
Family Support -0.13 *** 0.88
Deviant Peers 0.35 *** 1.42
-2 LL 4254.92 3837.89
Chi-Square 1011.74 *** 1428.77 ***
Nagelkerke R
2
0.31 0.41
BIC -887.27 -1304.30
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= females as the reference group;
b
= whites as the reference group;
c
= two biological parents as the reference
group
78
Peer influence is the most parsimonious adolescent sexual activity
explanation according to the BIC statistic. Among the family influence measures,
the BIC statistic shows that parental supervision is the most parsimonious while
family transition is the least parsimonious adolescent sexual activity explanation.
None of the explanations in this investigation was able to fully explain the
non-traditional household effects. However, some measures were able to partially
explain this relationship. Recall that the definition of partial explanation of a
particular non-traditional household effect for this study occurs when a particular
measure decreases the non-traditional household beta by at least 25% from the
baseline model. When using this standard, the results show that the parental attitude
is the best reason for the unwedded mother household, the parental supervision is the
best reason for the divorced/separated mother household, and family transition
explanation is the best reason for the stepparent household effects regarding
adolescent sexual activity.
Discussion
The main reason for this investigation was to determine the best reason for
the family structure and adolescent sexual activity relationship among five
explanations across three non-traditional households for the total sample. Despite
the subtle characteristic differences among the non-traditional households, the results
of this investigation revealed that none of the explanations was able to fully account
for any of the non-traditional household effect. However, these influences were able
79
to partially explain some of the non-traditional household effect. From one
perspective, this finding seems to suggest that the non-traditional household effect
may have multiple causes and that more than one family influence would be required
to fully explain this effect. If this were the situation, simply adding more of the
influence explanations to the regression models would eventually account for the
non-traditional household effect regarding adolescent sexual activity.
Another reason why the five explanations individually were not able to fully
explain the non-traditional household effects is because the focus of this
investigation is to analyze the total sample instead of analyzing the data within the
adolescent’s race and/or gender contexts. Contextualizing the analysis would imply
that the non-traditional household effect varies by the adolescent’s culture and/or
gender. For example, these non-traditional household effects may differ by race
because the non-traditional household may be institutionalized within a certain
culture. Cultural institutionalization of some non-traditional households, such as a
single mother household, would suggest that these households could be more
prevalent in the neighborhoods where the children live. If children who reside in
such households see that their friends and neighbors live in households that are
similar to them, they may not feel as disadvantaged compared to the adolescents who
reside in non-traditional households where the community is predominately
composed of two biological parent families (Moore, 2001). This phenomenon may
also encourage extended family members to assist the immediate family members of
non-traditional households since they may need more help compared to two
biological parent households (Hines & Boyd-Franklin, 1996). Both of these reasons
80
may blunt the impact of the household environment in regard to adolescent sexual
activity.
Contextualizing the analyses within the adolescent’s gender and/or race
categories would also help to determine whether there are differences for non-
traditional household effect by these subgroups. Unlike the results that are found in
this investigation, these analyses may also reveal that a certain explanation by itself
may fully mediate the non-traditional family structure effect. Because of these
reasons, the contexualized analyses of the data by the adolescent’s gender,
adolescent’s race, and the adolescent’s race-gender categories will be conducted in
chapters 6, 7, and 8.
Unwedded Mother Household Discussion
Mother’s attitude was found to be the best reason for the unwedded mother
household effect. This finding is somewhat supported by Thronton (1991). One
probable reason why mother’s attitudes is a better explanation for the unwedded
mother household effect than parental supervision is because there is no debate as to
who has the parental authority over the children for those who reside in these
households. Unlike children who live with divorced/separated mothers, children
who reside with unwedded mothers may have lived in these households for their
whole life. This allows the resident parent (in this case the mother) to be the only
parental figure in their children’s lives. Thus, the mother’s attitude may play a much
81
larger role in the socialization of these children since the parental authority issue is
not a main concern for unwedded mothers.
This premise may be one reason why it was surprising to find the similar
levels of parental strictness and some of the parental monitoring variables between
unwedded mother and two biological parent households. It seems that unwedded
mothers are very much involved in their children’s lives despite the absence of
another parental figure at home. This finding is contrary to studies that find that
single parents do not discipline their children as strictly as the parents of two parent
households because they a lack of another adult figure to help discipline the children
(Smetana, 1989). In addition, these findings also support the premise that the one-
parent household environment may not be the same even though they share similar
family organization and that they, therefore, should be analyzed separately (Moore,
2001).
Divorced/Separated Mother Household Discussion
The results from this study revealed that parental supervision is the best
reason for the divorced/separated mother effect. The finding seems to partially
support the study by Thomson, McLanahan, & Curtin (1992). It seems that lower
levels of parent-child interaction and parental monitoring of children’s activities may
both play a role in the divorced/separated mother household effect.
The findings from the divorced/separated mother household investigation do
support this premise. Divorced/separated mothers were less likely than the parents
82
of two parent households to be present around the home. This finding may be either
due to the need to work longer in order to maintain the household income (Thronton,
1991) or to the need to do activities that could have been done by the absent parent
such as running household errands (Biblarz & Raftery, 1999). Divorced/separated
mothers were also less strict with their children than mothers of two biological parent
households. This result may be either due to these mothers struggling with the
parental authority issue because there still may be conflict between parents about
their roles in taking care of the children (Ahrons & Rogers, 1987) or that these
mothers are still enacting the traditional maternal role (Thomson, McLanahan, &
Curtin, 1992) despite having the sole responsibility to rear their children.
In addition, the divorce process is stressful for children since they would need
to adapt to a new family environment (Holmes & Rahe, 1967; Wu & Martinson,
1993). Children may have difficulty adjusting to the new family environment and
this stress is further compounded when a parental figure is not at home as often as
before the divorce. The lack of a parental presence for children who reside with
divorced/separated mothers may result in them feeling unsupported by family
members. These children may then turn to peers for support when their immediate
family members are not emotionally available to them during these times (Coleman,
1981). The presence of peers in the child’s life may soften the impact of the divorce
process (Hetherington, 1979), but also hastens a child’s entry into sexual activity
(Hetherington & Clingempeel, 1992). In essence, the divorce process may make it
more difficult for a divorced/separated mother to take care of her children and the
consequences may then lead to an increasing peer influence on a child’s behaviors.
83
Stepparent Household Discussion
Stepparent households seem to be similar to the two biological parent
households regarding levels of parental monitoring and parental strictness. This
finding would be consistent with the parental social control hypothesis which posits
that having two parents at home would equate to greater monitoring of the children’s
activities and greater disciplining of the children’s behaviors (Thomson, McLanahan,
& Curtin, 1992). However, adolescents who reside with a stepparent do not seem to
have a closer bond with their mothers compared to those who reside with two
biological parents as evidenced by their relatively lower parent-child activities and
parental meal presence levels. These findings, which are consistent with studies by
Dornbusch et al. (1985) and Thomson, McLanahan, & Curtin (1992), may be due to
ambiguities concerning the stepparent’s role, duties, and responsibilities. Not
resolving these issues may lead to ineffective parenting (Cherlin, 1978; Furstenberg,
1987). In addition, mothers of these households seemed to be similar to mothers of
other non-traditional households in terms of being more tolerant of adolescent sexual
activity. Thus, the results of this investigation revealed that stepparent households
could be an exception to the rule that marriage tends to be a conservatizing force
regarding parental attitudes and values (Newcomer & Udry, 1984; Thronton &
Camburn, 1987; Thornton, 1991).
Family transition was the best reason for the stepparent household effect.
This finding suggests that family members of these households may be struggling
with issues related to accommodating a new person into the family (Cherlin, 1978;
84
Amato, 1987; Furstenberg, 1987). However, the parental supervision and parental
attitude explanations also play a role in accounting the stepparent effect, but they
were not as strong as the family transition explanation.
Relative Importance of the Five Explanations Discussion
Sexually experienced adolescents were significantly different from virgin
adolescents in many ways. In addition, many of the demographic, family influence,
and peer influence variables were also directly associated with adolescent sexual
activity. However, some of the variables that comprise the family influence
measures either were not associated with sexual activity or were the reverse of the
predicted direction. For example, the parental monitoring variables were not
associated with adolescent sexual activity even though mother of virgin adolescents
are more likely than mothers of sexually experienced adolescents to be present
around the home. This finding disagrees with studies that found parental monitoring
of the children’s activities being negatively associated with sexual activity (Hansson
et al., 1981; Small & Luster, 1994; Jacobson & Crockett, 2000). It is possible that
parent-child interactions have more influence in affecting an adolescent’s decision to
engage in sexual activity than parental monitoring of the children’s activities.
Peer influence was the most parsimonious adolescent sexual activity
explanation according to the BIC criterion. The hypothesis that acquaintance with
peers who engage in deviant activities would be positively associated with sexual
activity was supported by the results of this investigation. This finding is in
85
agreement with studies by Metzler et al. (1994) and Whitbeck, Simons, & Goldberg
(1996). From one perspective, this finding would seem to suggest that a peer’s
influence over an adolescent’s decision to engage in sexual activity may be much
stronger compared to a parent’s influence during puberty. A plausible reason for this
finding may be that adolescents feel pressure to conform to their friend’s and their
fellow peer’s values and expectations by engaging in similar activities so they do not
feel isolated from them. Thus, adolescents who associate themselves with peers who
engage in deviant activities may be more likely to perform in these types of activity
themselves.
From another perspective, peer influence being the most parsimonious
adolescent sexual activity explanation when compared to the others may not be
completely surprising. This measure, which asked children the number of
acquaintances or friends they know who have engaged in deviant activities, may
have a built in bias in its association with adolescent sexual activity because those
who do drugs or drink alcohol are also more likely to engage in sexual activity
(Flewelling & Bauman, 1990). Using another peer influence measure for the
analysis, such as the time one spends with friends, may produce a different result.
Consequently, the type of questions that one uses to measure peer influence may
drastically affect its outcome regarding adolescent sexual activity.
Parental supervision was found in this investigation to be the second most
parsimonious adolescent sexual activity explanation. However, this explanation was
the most parsimonious family influence explanation. The hypothesis that all the
variables that constitute the parental supervision measure would be negatively
86
associated with sexual activity was not supported by the analysis. Only mother-child
activities, mother-child discussions, parent’s presence during meals, and parental
strictness variables were negatively associated with sexual activity which seems to
be congruent with results by Whitbeck, Conger, & Kao (1993). These findings seem
to suggest that the everyday interactions between the parent and the child may play a
much more significant factor in preventing sexual behaviors than the constant
monitoring of the child’s activities. It may be that parents who interact more with
their children are building a stronger relationship with them. This bonding process
may result in the child feeling supported by the family and not having the need to
engage in sexual activity during puberty in order to feel accepted as an individual.
Mother-child discussions were hypothesized to be negatively associated with
sexual activity because it was assumed that these events are times when parents and
their children have the opportunity to strengthen their relationship. Instead, the
results revealed that mother-child discussions were positively associated with sexual
activity. These discussions are apparently a proxy for how dissatisfied their mothers
are about their children’s behaviors. Since the levels of mother-child discussions
were significantly greater for non-traditional households than for two biological
parent households, it is possible that adolescents who reside in non-traditional
households are more likely to get into troublesome behaviors. Some of these
behaviors may be due to the children having difficulty adjusting to a new family
environment as in the case for those who reside with either divorced/separated
mothers or stepparents, or the lack of another adult figure at home who can help
87
parent the children, as in the case of those who reside with either divorced/separated
or unwedded mothers.
Parental attitude was found to be the third most parsimonious adolescent
sexual activity explanation. The results did not support the hypothesis that all of the
parental attitude variables would be negatively associated with sexual activity.
Among the parental attitude measures, only mother’s religiosity and her disapproval
of adolescent sexual activity were negatively associated with sexual activity.
Mother’s religiosity being negatively associated with sexual activity is congruent
with studies by Thornton & Camburn (1987), Jaccard et al. (1996), and Dittus &
Jaccard (2000). Mother’s disapproval of adolescent sexual activity being negatively
associated with sexual activity also parallels the literature by Jaccard et al. (1996)
and Dittus & Jaccard (2000). Thus, these findings seem to suggest that children who
grow up in sexually conservative households are less likely to engage in sexual
activity than those who grow up in sexually liberal households.
Similar to mother-child general discussions, mother-child sex discussions
were positively associated with sexual activity. This is a surprising finding since the
proposed hypothesis was that frequent mother-child sex discussions would be
negatively associated with sexual activity since these conversations would be a way
for parents to explain the possible dangers of engaging in sexual activity during
puberty to their children. However, if this variable’s scale were reversed, the result
would have been that not talking about sex would be negatively associated with
sexual activity. When taking both mother’s religiosity and mother’s disapproval of
adolescent sexual activity into account, the result seems to suggest that parents who
88
do not discuss sex with their children are more conservative in their sexual values or
that parents who discuss sex with their children are more liberal about sex. This
premise is supported by findings that the parents of two biological parent households
are more religious, more disapproving of adolescent sexual activity, and have fewer
discussions about sex with their children than parents of non-traditional households.
There are several plausible explanations why mother-child sex discussions
were positively associated with sexual activity. However, it is not known at this time
as to which reason is primarily responsible for this positive association. The first
plausible explanation is that mothers may have frequent discussions with their
children about sex when they become aware of their sexual activity. This is a
reverse causality explanation where mothers would talk to her children about sex as a
reaction to their behavior in the hopes of preventing them from engaging in future
sexual activity during puberty.
The second plausible explanation for this positive relationship is that mothers
who discuss sex openly may be more liberal about sexual expression and pass on this
philosophy to their children. The content of these discussions may involve a lot of
talk about protections such as the use of condoms and the birth control pill.
Although such mothers may not be espousing early sexual experience for their
children, they may be subtly normalizing sexual activity by sending the message to
them that these behaviors are acceptable behavior as long as he or she is well
protected. Further research would be needed to better illuminate why parent-child
sex discussions are positively related to adolescent sexual activity.
89
The results of this investigation revealed family support to be the fourth most
parsimonious adolescent sexual activity explanation. The hypothesis that family
support would be negatively associated with sexual activity is supported by this
investigation. This finding is also congruent with studies such as those by Metzler et
al. (1994). It seems that adolescents who do not feel supported by their parents are
more likely to seek alternate intimate relationships in order to replace the lack of
emotional support by family members.
Family transition was found to be the least parsimonious adolescent sexual
activity explanation for this investigation. There are two possible explanations for
this finding. First, this finding may be due to other family influence or the peer
influence explanation accounting for the effect of this measure. In another words,
family transition may be indirectly associated with adolescent sexual behaviors.
Second, this finding may be due to the measure representing only the number of
marital or family transition events a child may potentially experience in his or her
life rather than the quality or degree of disruption and emotional chaos that is being
experienced by him or her during these occasions. Nevertheless, the findings
supported the hypothesis that family transition would be positively associated with
sexual activity. Furthermore, this finding is in agreement with studies by Chase-
Lansdale, Cherlin, & Kiernan (1995) and Peterson & Zill (1986).
Finally, an important finding of this study is that the black x gender and
Hispanic x gender variables remained statistically significant for adolescent sexual
activity throughout the analysis. This finding suggests that the factors for adolescent
sexual activity may differ within a particular subgroup culture. A subgroup culture
90
may strongly influence the way parents raise their children and could result in
adolescents experiencing different family environments even though they live in
similar household configurations. According to the ecological perspective as
proposed by Bronfenbrenner (1979), each individual brings to the family certain
personal attributes and characteristics. Since family processes always occur within a
specific context, these processes may have different effects for different groups of
individuals in distinct family settings. This premise also coincides with the
discussion earlier that the non-traditional family structure effect may not be similar
across race and/or gender subgroups. Consequently, chapters 6 through 8 will
further investigate these issues.
Chapter Summary
The results of these analyses revealed that the family and peer influences
were not able to fully explain the family structure and adolescent sexual activity
relationship among the three non-traditional households. However, this finding is
not conclusive because this investigation analyzed the data by the total sample rather
than by the adolescent’s race and/or gender subgroups. It is possible that the non-
traditional household effect does differ by these race and/or gender contexts. In
addition, contextualizing the analyses within these subgroups may lead to the
discovery that some of the non-traditional household effects could be mediated by
several of the explanations that are used in this study.
91
The results of this investigation revealed that the best reasons for the
unwedded mother and the divorced/separated mother household effects were parental
attitude and parental supervision explanations respectively when using the partial
mediation definition. Meanwhile, family transition was the best reason for the
stepparent household effect.
This investigation also revealed significant differences between sexually
experienced and virgin adolescents. For example, sexually experienced adolescents
were more likely to reside in non-traditional households, to reside in poorer
households, and to have greater acquaintances with deviant peers than virgin
adolescents. Compared to virgin adolescents, sexually experienced adolescents
experienced lower levels of mother-child activities, parental presence during meals,
parental reinforcement of rules, and support by family members. However, sexually
experienced adolescents experienced greater general and sex discussions with their
parents than virgin adolescents. Mothers of sexually experienced adolescents were
more likely to have gone through multiple relationships and were more liberal in
their attitudes regarding sex than mother of virgin adolescents. Not surprisingly,
most of these characteristics were also significantly associated with adolescent
sexual activity.
The results from this investigation supported most of the proposed
hypotheses. However, the analysis revealed some exceptions such as mother-child
general and sex discussions being associated with a positive rather than a negative
association with sexual activity. A number of reasons were presented for these
findings. Another exception to the proposed hypotheses is the non-significance of
92
the parental monitoring variables (mother’s employment and mother’s presence at
home throughout the day and night) and adolescent sexual activity relationship.
Peer influence was the most parsimonious adolescent sexual activity
explanation according to the BIC statistic. However, as noted in the discussion, this
measure may have some degree of built in correlation with sexual behaviors. In
another words, adolescents who engage in deviant behaviors may also engage in
sexual activity. The second most parsimonious adolescent sexual activity
explanation was parental supervision. Family transition was the least parsimonious
adolescent sexual activity explanation according to the BIC statistic. This finding
may be due to the measure only representing the frequency of marital or family
transitions the child may potentially experience in his or her life rather than
representing the quality or degree of disruption and emotional chaos experienced by
children during these events.
93
Chapter 5: Adolescent Sexual Activity by Gender and Race: Theories and Rationale
The main objective for this chapter is to discuss how gender and race may
influence an adolescent’s decision to engage in sexual activity. This chapter begins
with a presentation of several theories that attempt to explain the reasons for gender
differences regarding adolescent sexual activity. The chapter then continues with a
presentation of several theories that attempt to explain the reasons for racial
differences regarding this issue for black, Hispanic, and Asian adolescents. Finally,
this chapter concludes with a literature review of how the relationship between each
of these explanations and adolescent sexual activity may differ within gender and
race contexts.
Race and gender are discussed together for this chapter because these factors
are generally intermingled in the family literature. It is difficult to discuss gender
without talking about race since the racial effects on adolescent sexual activity do
vary by gender and vice versa. Consequently, generalizing the findings by either
gender or race may distort the relationship between the five explanations and
adolescent sexual activity.
Gender Differences: Rationales
Gender is an important factor for adolescent sexual activity because during
childhood, males and females are exposed to very different cultural scripts and
expectations regarding sexual conduct. In general, society defines the appropriate
94
gender roles for each sex (Moore & Rosenthal, 1993). Many cultures and societies
have defined the male role in terms of men as primary breadwinners, head of the
households, and leaders in the community. Men are encouraged to develop
personality traits such as independence, assertiveness, confidence, and bravery to
help them achieve these objectives. In the sexual arena, they learn to initiate sexual
activity, to be the one with the more powerful sex drive, and to be the more powerful
figure in their relationship with women. In contrast, many cultures and societies
have defined the female as the one who gives, sustains, and protects life by bearing
and nurturing children and being responsible for household duties. Women learn to
achieve these tasks by developing personality traits such as warmth, expressiveness,
nurturance, dependence, and cooperation. Furthermore, in the sexual arena, the
expectation for women is to be pleasant, cooperative, placative, flirtatious, and be
available for the pleasure of the male while retaining a lady-like demeanor in public.
Obviously, recent trends such as the wide scale employment of women in the
workplace and the feminist movement have challenged and redefined many of these
stereotypic female roles. Still, many of these traditional roles persist and are likely
to have a strong impact on female adolescents.
These prescribed gender roles may help to establish a double standard
regarding how parents socialize their children about sexual issues. Parents may be
more likely to accept or encourage premarital sexual activity for their sons while
discouraging it for their daughters since males experience less severe social and
moral consequences of engaging in premarital sex than females (Nathanson, 1991).
95
Consequently, parents may socialize their sons and daughters differently so that they
will conform to expected societal norms and mores as adults.
Although researchers have found significant differences between male and
female sexuality rates during adolescence (Laumann et al., 1994), this disparity at
any given age has been diminishing in recent cohorts (Davis & Friel, 2001).
Nevertheless, studies have found that males are more likely to engage in sexual
activity during puberty than females (Dryfoos, 1990; Laumann et al., 1994,
Lauritsen, 1994). There are several reasons for these findings.
First, adolescent sexual activity rates may differ by gender because of
biological differences. This argument contends that the release of androgenic
hormones, such as testosterone, by endocrine glands increases an individual’s libido
(Udry, 1988). During pre-puberty, these androgenic hormone levels are equivalent
for all individuals regardless of their gender. However, during puberty, this level
increases to approximately twenty times for males while it only doubles for females.
Since these hormones are associated with higher libido levels, males are more likely
than females to be predisposed to engaging in sexual activity during puberty.
Some studies have found support for this argument by finding that the
androgenic hormonal levels during puberty strongly influence a male’s, but weakly
influence a female’s, sexual behaviors (Udry, 1988; Udry & Billy, 1987; Udry,
Talbert, & Morris, 1986). Udry & Billy (1987) found that a male’s decision to
engage in sexual activity during early adolescence is due to a combination of
hormonal effects along with popularity with the opposite sex. For females, increased
testosterone levels are found to be associated with sexual interests but not sexual
96
behaviors (Udry, Talbert, & Morris; 1986). Although the biological argument may
explain some gender differences in regards to adolescent sexual activity rates, this
finding suggests that social factors may play an even larger role for females than for
males regarding this issue.
Second, adolescent sexual activity rates may differ by gender because of
differences in the way males and females perceive the consequences of being
sexually involved with another person. According to this argument, females are
more likely than males to perceive the consequences of sexual activity as being more
severe for them because they have to take into account factors such as obtaining birth
control and becoming pregnant (Small, Silverberg, & Kerns, 1993). For females, the
cost of pregnancy may also include a diminished chance of achieving higher
educational and occupational status (Card & Wise, 1978; Hofferth & Moore, 1979).
In contrast, males may have a greater desire to engage in sexual activity because they
do not have to bear the cost of pregnancy. This reasoning is somewhat supported by
Day (1992) who found that older male adolescents with higher educational
aspirations are more likely to seek a sexual partner while female adolescents with
higher occupational desire were less likely to have a sexual partner.
Race Differences: Black Adolescents
A number of researchers have found that black adolescents were more likely
than white adolescents to have engaged in sexual activity during puberty (Hayward,
Grady, & Billy, 1992; Stanton et al., 1994; Warren et al., 1998). In addition,
97
researchers have found that Hispanic (Slonim-Nevo, 1992; Aneshensel, Fielder, &
Becerra, 1989) and Asian adolescents (Strunin, 1991; Schuster et al., 1998) were less
likely to have engaged in sexual activity during puberty than white adolescents.
There are several reasons for the racial differences regarding sexual activity
rates between white and black adolescents. First, this racial difference may be due to
biological differences. Some researchers have theorized that black adolescents are
more likely to engage in sexual activity than white adolescents because they are
more likely to enter puberty at a younger age (Harlan, Harlan, & Grillo, 1980).
However, the support for this argument has been mixed. Some researchers
have found that early physically maturing black children of both genders are more
likely to have engaged in sexual activity at an earlier age than those who matured
later (Murry, 1992; Zabin & Hayward, 1993). In contrast, other researchers have
found little or no association between physical maturation and earlier sexual activity
during puberty (Moore, Simms, & Betsey, 1986; Hofferth, Kahn, & Baldwin, 1987).
Yet, some researchers are not convinced that the biological factors can account for
these race differences. Furstenberg et al. (1987) argued that the biological argument
only partially explain the differences in sexuality levels between white and black
females because virtually all of them reach puberty by age 15. They also pointed out
that the biological argument could not explain the differences in sexuality levels
between white and black males because other studies, such as those by Clark, Zabin,
& Hardy (1984), have found persistent racial differences for adolescent sexual
activity even though they are of similar gender.
98
Socioeconomic difference is the second reason why sexual activity rates may
differ between black and white adolescents. According to this explanation, sexuality
rates are greater for black adolescents because they are more likely to reside in
poverty (Allen & Mitchell, 1996; Sells & Blum, 1996; Taylor & Katz, 1996). Being
poor limits the opportunities for social and economic advancement and makes the
opportunity cost of early parenthood minimal (Furstenberg et al., 1987).
Wilson (1987) speculated that the opportunities for social and economic
advancement for blacks are limited because of the absence of middle class blacks in
large inner city areas due to their migration to the more desirable suburbs during the
1970s and the 1980s. The consequences of the flight to suburbia by middle class
blacks from large inner city areas included an absence of successful role models for
black youths and these areas being turned into places of isolation, poverty, and a
higher concentration of lower socioeconomic status African American families
(Wilson, 1987; Ransford, 1994). People who live in these areas were then
encouraged to incorporate a new set of norms and values that are inconsistent with
the middle class values. The adoption of these norms and values by these residents
further contributed to the weakening of the social control mechanisms such as those
found in family, peers, and schools, and pro-social activities that are sponsored by
the schools, churches, and communities. These weakened social control mechanisms
would then promote a socialization pattern that could lead to an increase risk for
adolescent sexual activity.
Despite this reasoning, empirical studies have found some gender difference
for black adolescents regarding the relationship between socioeconomic status and
99
sexual behaviors. Hogan & Kitagawa (1985) found that black females who reside in
economically lower class families were more likely than those who reside in middle
and upper class families to have engaged in sexual activity. In contrast, Ramirez-
Valles, Zimmerman, & Juarez (2002) did not find a relationship between family
socioeconomic status and sexual activity for 14-16 year old black adolescents of both
genders. Instead, they found that neighborhood poverty was positively associated
with sexual activity only for males. This finding was attributed to the belief that
young men were seeking sexual activity as a means to achieve status in response to
the absence of educational and well-paid job opportunities.
Cultural differences are the third reason why sexual activity rates may differ
between white and black adolescents. According to this view, the African American
culture is more likely to tolerate liberal sexual norms and values including the
acceptance of out-of-wedlock childbearing (Staples, 1978). While the cultural
differences between blacks and whites were initially due to differences in
socioeconomic status, the African American culture has gradually adopted tolerant
attitudes regarding early childbearing over time (Moore, Simms, & Betsey, 1986).
Consequently, children who grew up in subcultures that are more tolerant of early
sexuality and non-marital childbearing may find it more acceptable to engage in
sexual activity before reaching young adulthood.
Several studies that investigated white and black adolescents have found
cultural differences regarding childbearing issues. Zelnick, Kantner, & Ford (1981)
found that black females were less likely than white females to perceive
condemnation of single mother childbearing. Abrahamse, Morrison, & Waite (1988)
100
found that black females who are high school sophomores were more likely than
their white female counterpart to consider having a child outside of marriage. In
addition, Moore, Peterson, & Furstenberg (1986) found that black adolescents were
more likely than white adolescents to favor having their first birth before 20, to
report their ideal marital age older than the age at which they want their first child,
and to perceive less unfavorable reaction to non-marital sex from peers and male
partners.
Race Differences: Hispanic Adolescents
Cultural factors are generally cited as the reasons for differences in
adolescent sexual activity rates between Hispanic and white adolescents. Compared
to the black and white populations, Hispanics more conservative in their attitudes
regarding sexuality (Davis & Harris, 1982; Padilla & O’Grady, 1997) since family
members tend to be more familistic, where the family ties and family loyalty are
placed above all else including individual needs (Mirande, 1977). In another words,
Hispanic adolescents may have a lower sexuality risk than both black and white
adolescents because they usually reside in households that uphold traditional values
that discourage sexual behaviors during puberty (Darabi & Ortiz, 1987; East, 1998).
In addition, Hispanic families are predominately Catholics (Bach-y-Rita, 1982),
which could strongly influence both a parent’s view regarding sexuality and the type
of sexual teachings that is conveyed from the parent to their children.
101
Another cultural factor that may also affect the sexuality risk for Hispanic
adolescents is the emphasis of distinct gender roles for males and females (Staples &
Mirande, 1989). The emphasis of this belief throughout the Hispanic culture may
lead parents to have different expectations for their sons and daughters when they
become adults. These expectations help to guide parents in teaching the appropriate
behaviors to their children and help establish the acceptable pathways for them to
choose as they enter adulthood (Driscoll et al., 2001). As in many other cultures, the
Latino culture values the concept of marianismo for women, which emphasizes the
female’s role as a mother and celebrates the mothers’ self-sacrifice and suffering for
their children (Ginorio et al., 1995). Based on the Catholic ideal of the Virgin Mary,
this view promotes chastity for young women and values sexual abstinence until
marriage or at least until a serious, long-term relationship takes place. However, the
Latino culture values the concept of machismo for men which emphasizes the man’s
role as a provider, protector, and representative of the family with the duty to uphold
the family honor and to maintain its integrity (Mikawa et al., 1992). This role also
provides greater sexual freedom and public social interaction for men than for
women.
Race Differences: Asian Adolescents
Cultural and family factors are found to strongly influence the sexual
behaviors of Asian adolescents (Connor, 1976; Hirayama & Hirayama, 1986).
Children are discouraged from displaying strong emotions since social order and
102
control of emotions are highly valued attributes in traditional Asian cultures
(Hirayama & Hirayama, 1986). In addition, the Asian culture also places a very high
priority on family unity (Hirayama & Hirayama, 1986), which teaches children to
have the utmost respect for their parents and to depend on the family for their needs.
The emphasis on family unity provides parents an opportunity to maintain greater
control over their children’s lives and may strongly influence the sexual expressions
of individuals who grow up in these households.
These sections have presented several theories and rationales why adolescent
sexual activity rates may differ by gender and race. Since the child’s gender and
race contexts have a huge influence in shaping his or her life experiences, it is also
possible that these contexts also affect the levels of family influence that children are
exposed to during their lifetime. These differences may result in some of the family
influences having a stronger association with adolescent sexual behaviors than other
family influences. The next few sections present a review of empirical studies that
examined the relationship between the five explanations and adolescent sexual
behaviors within the gender and race contexts.
Family Transitions: Gender & Race Differences
Studies have found gender differences, mostly among white adolescents, for
the relationship between family transitions and adolescent sexual activity with these
events having a greater impact for males than for females. One reason for this
finding may be due to differences in time parents spend with their sons and
103
daughters during the family transition period (Emery, Hetherington, & Dilalla, 1985;
Gove & Crutchfield, 1982). For example, Wallerstein & Kelly (1980) found that
mothers who were in the divorce process are more willing to spend more time with
their daughters than with their sons. In addition, Newcomer & Udry (1987) found
for white males that the transitional event of going from a two-parent to a single
parent household is associated with adolescent sexual activity. However, for white
females, those who resided with an unwedded mother were at a greater risk for
adolescent sexual activity. They concluded that this gender difference was due to
mothers being more likely to neglect their sons than their daughters during the
divorce process, which would then lead sons to seek alternate intimate relationships
outside of the immediate family.
There may be racial differences between blacks and whites for the
relationship between family transition and adolescent sexual activity. For single
parent households, researchers, such as Wu & Thomson (2001) and Moore (2001)
have found that the transition event has less of an impact on the lives of black
adolescents when compared to white adolescents. There are several plausible
reasons for this finding. The first reason may be due to the role of extended family
members in black households (Moore, 2001). One characteristic of black households
is for the extended family members to help support families in times of need (Hines
& Boyd-Franklin, 1996). This action may mitigate the stresses that family members
of non-traditional households usually experience during the family transition event
so its impact on children would be minimal for those who have to live through
during this difficult period.
104
The second reason is because of differences in parental social control levels
that both of these adolescents may experience after these events has taken place. Wu
& Thomson (2001) found that family transition is directly associated with adolescent
sexual activity only for white females. They speculated that black females were not
significantly affect by the family transition event because parents were able to
mitigate the stresses by either implementing stricter household rules or by closer
monitoring of their children’s activities during these periods.
The family transition and sexual activity relationship may also differ between
white and black adolescents for those who reside with two parents due to the
perception of these events by the culture and community. Researchers have found
that white adolescents who reside with a stepparent are more likely to have
adjustment problems (Hetherington & Clingempeel (1992) and be sexually active
(Moore, 2001) than those who reside with two biological parents. However, several
researchers have found no greater likelihood of sexual activity for black adolescents
who reside with a stepparent compared to those who reside with two biological
parents (McLanahan & Bumpass, 1988; McLanahan & Sandefur, 1994; Moore,
2001).
One reason for this finding may be that both the cultural and community
context strongly influences the adolescent’s perception of stepparent households
(Moore, 2001). Since white families are more likely to reside in an advantaged
environment, e.g. two biological parents, there may be a negative perception for
those who reside with a stepparent or who live in other non-traditional households.
This negative perception may result in increased problem behaviors for white
105
adolescents. In contrast, black families are more likely to reside in a disadvantaged
context where the community has significantly higher rates of unemployment,
welfare receipt, non-marital childbearing, and adolescent delinquency. In this
context, adding another adult to the household may help stabilize the household by
being another source of financial and psychological support for the family. Children
from these households may experience less “relative deprivation” compared to their
friends or peers around the neighborhood. These adolescents would be less likely to
engage in problem behaviors because they may perceive themselves as being better
off economically and socially than friends or peers who reside with a single parent.
Another reason may be due to type of marriage entered into by the parent.
Since non-marital birth rates tend to be higher for low-income individuals and
African American families compared to the general population, the mother’s
marriage to the stepparent may feel more like a first marriage than a remarriage
(Taylor et al., 1997). Thus, the black stepparent home environment may be more
similar to those of two biological parents because these family members could have
avoided the need to deal with the issues that usually accompany divorces.
Parental Supervision: Gender & Race Differences
While some studies have not found significant gender differences for the
parental presence and adolescent sexual activity relationship (Smith, 1997; Small &
Luster, 1994), those that have found gender differences for this relationship find that
parental monitoring of the children’s activities and parent-child interaction frequency
106
have a greater influence on a female’s than a male’s sexual behaviors. For example,
Jacobson & Crockett (2000) found for white female adolescents that they were less
likely to engage in sexual activity when their mothers did not work. Furthermore,
Ramirez-Valles, Zimmerman, & Juarez (2002) found among black adolescents that
spending time with their mother reduced the risk of sexual behaviors for females but
not for males. They attribute this finding to mothers encouraging their daughters to
delay sexual intercourse in order to attain higher levels of education and
socioeconomic status.
Researchers also found some gender differences for the association between
parent-child relationship quality and adolescent sexual activity. Upchurch et al.
(1999) found that mother’s strictness was positively associated with sexual activity
for males, but not for females. They attributed this finding to males being less likely
to abide by their parent’s wishes to postpone sexual activity and becoming more
likely to be susceptible to peer influence if they felt that their parents were
controlling. In addition, Benda & Corwyn (1998) found that white females were less
likely than white males or black adolescents of both genders to be sexually
experienced because they are more likely to form close emotional bonds with their
mothers. These emotional bonds would allow mothers to have more influence over
their daughters than their sons, which resulted in daughters being more accepting of
their parent’s beliefs.
The literature for Hispanic adolescents presents a mixed relationship between
parental supervision levels and adolescent sexual activity. Miller, Forehand, &
Kotchick (1999) found for Puerto Rican adolescents that mother’s monitoring of the
107
children’s activities was negatively associated with sexual activity. However, they
did not find a relationship between mother’s monitoring of the children’s activities
and sexual activity for other Hispanic adolescents.
A literature search has not revealed a published study that examined the
parental supervision and adolescent sexual activity relationship for Asian
adolescents. However, a study that used the number of parents in the household as a
proxy for parental supervision found that the parental supervision effect on an
Asian’s sexual experience was similar to that of white adolescents. Schuster et al.
(1998) found for Asian adolescents that those who were who reside with two parents
were more likely to be virgins than those who were who reside with a single parent.
Parental Attitude: Gender & Race Differences
The relationship between a parent’s religious values and adolescent sexual
activity may differ by both gender and/or race contexts. Thornton & Camburn
(1987) found for white adolescents that mother’s church attendance was associated
with virginity status only for males. However, a literature search did not reveal a
published study for black adolescents that investigated the parent’s religiosity and
adolescent sexual activity relationship. Nevertheless, researchers have found a
gender differences for black adolescents when adolescent religiosity is used as a
proxy for parental religiosity. Ramirez-Valles, Zimmerman, & Juarez (2002) found
that black females who participated in church activities were less likely to initiate
108
sexual activity. In contrast, they did not find for black males an association between
participation in church activities and adolescent sexual activity.
A literature search has not revealed a published study for Hispanic
adolescents that examined the parent’s religiosity and adolescent sexuality
relationship. However, the literature has revealed mixed results for this relationship
when adolescent’s religiosity is used as a proxy for parental religiosity. Liebowitz,
Castellano, & Cuellar (1999) found for Hispanic adolescents of both genders that
religious importance and church attendance were negatively associated with sexual
activity. In contrast, Aneshensel, Fielder, & Becerra (1989) did not find for Latinas
an association between adolescent’s religiosity and their sexual activity.
Parent-child sex discussions were generally associated with adolescent sexual
activity, but this effect may differ by race and gender. Fox & Inazu (1980) found for
white female adolescents that mothers who had discussions with them about sex are
less likely to be sexually experienced. In addition, Moore, Peterson, & Furstenberg
(1986) found for white females that not only were parent-child sex discussions
negatively associated sexual activity, but that these conversations more likely took
place when their parents held traditional values. For white males, these discussions
were not negatively associated with sexual activity regardless of whether their
parents held traditional values. Although the parent’s reason for discussing sex
topics with their children was to prevent sexual behaviors, the timing of these
discussions may be responsible for differences in this finding. While parents were
more likely to discuss sex topics with their daughters as a way to protect them from
potential life-altering consequences of premarital sexual activity such as pregnancy,
109
they were more likely to discuss sex topics with their sons as a reaction to the
possibility that sexual activity had taken place.
The relationship between parent-child sex discussion and adolescent sexual
activity for black adolescents has revealed mixed results. Some researchers found
that the discussion about sex is associated with delayed sexual activity if their
parents talk to them about such topics as menstruation, pregnancy, contraception,
and sexually transmitted diseases (Scott-Jones & Turner, 1988; Murry, 1992).
However, other studies have not supported this finding. For example, Tucker (1989)
did not find a relationship between inter-generational patterns of mother-daughter
communication about sex and delayed sexual activity. Murry (1994) also found for
black females that parents discussing sex topics with their children neither fostered
nor inhibited coital timing. In addition, Miller, Forehand, & Kotchick (1999) found
for black adolescents of both genders that sex communication was associated with
greater sexual intercourse frequency and sexual partners.
Studies also have found mixed results for the parent-child sex discussions and
sexual activity relationship among Hispanic adolescents. Adolph et al. (1995) and
Baumeister, Flores, & Marin (1995) found for Latinas that those who talked to their
parents about sex were less likely to become pregnant. In addition, Pick & Palos
(1995) found for females who reside in Mexico City that those who talked frequently
with their mothers about sex were less likely to be sexually active than other teens.
However, O’Sullivan et al. (1999) did not find a significant association between
mother-child sex discussions and adolescent sexual activity for Latinas.
110
In contrast to the studies on white, black, and Hispanic adolescents, only a
few researchers examined the parent-child sex discussions and sexual activity
relationship for Asian adolescents. One study found that Asian adolescents
communicated much less with their parents about AIDS and HIV compared to white
adolescents (Hou & Basen-Engquist, 1997). They attributed the parent’s reluctance
to talk with their children about sex to sexual conservatism within the Asian culture.
A gender effect may exist among white adolescents for the mother’s
disapproval of adolescent sexual activity and subsequent adolescent sexual activity
relationship. Whitbeck, Simons, & Goldberg (1996) found for white females that
mother’s sexually permissive attitudes were associated with adolescent sexual
activity. In addition, Thornton & Camburn (1987) found for white families an
association between mother’s attitudes regarding sex and sexual initiation for
daughters but not for sons.
Studies have found that mother’s disapproval of adolescent sexual activity
may be negatively associated with sexual activity for non-white adolescents as well.
In one study, black adolescents were less likely to engage in sexual activity if they
perceived their mothers to be disapproving of sex even though they have a greater
likelihood of exposure to a more liberal attitude regarding this issue (Jaccard, Dittus,
& Gordon, 1996). In addition, researchers have also found for Hispanic adolescents
of both genders that mother’s disapproval of adolescent sexual activity is negatively
associated with sexual activity (Slonim-Nevo, 1992; Hovell et al., 1994; Miller,
Forehand, & Kotchick, 1999). However, for Asian adolescents, a literature search
111
had not revealed a published study that examined the variables that comprises the
parental attitude measures and adolescent sexual activity relationships.
The Family Support: Gender & Race Differences
The frequency in which parents communicate with their children may differ
by gender. Noller & Callan (1991) found that parents were more likely to
communicate more frequently and talk about a wider range of issues with daughters
than with sons. They also found that daughters viewed the communication with their
mothers more positively than the communication with their fathers despite reporting
more conflict with their mothers. Thus, mothers who have more frequent and
meaningful communication with their daughters may have a greater opportunity to
form a closer relationship with her.
Some studies have found gender differences among white adolescents
regarding the family support and adolescent sexual activity relationship. Although
daughters may be closer to their mothers, Lauritsen (1994) found that white males
who feel close to the family were less likely to engage in sexual activity. However,
this relationship was not found to be significant for white females. In contrast,
Whitbeck et al. (1992) reported that the adolescent’s perception of parental support
absence was more strongly associated with the onset of adolescent sexual activity for
white females than for white males.
Several studies have found mixed findings for gender differences among
black adolescents regarding the family support and adolescent sexual activity
112
relationship. Lauritsen (1994) did not find an association between family attachment
and sexual initiation for black adolescents of both genders. However, Ramirez-
Valles, Zimmerman, & Juarez (2002) found that an adolescent’s participation in
family decision making, as measured by the extent to which family members made
decisions together, was negatively associated with sexual activity for males, but not
for females. In contrast, Moore & Chase-Lansdale (2001) found for black females
that the quality of the parent daughter relationship, as measured by degree of mutual
trust, quality of communication, and extent of anger and alienation felt by the
daughter regarding her parent, was negatively associated with sexual activity. One
possible explanation for this finding could be these adolescents may experience
“poverty in the relationship” where the adolescent feels alienated from family
members (Kaplan, 1997). Consequently, these adolescents may believe that getting
pregnant and giving birth to a baby may provide the love and security to replace the
lack of closeness and bonding with their mothers.
However, studies have not found a relationship between family support and a
decreased risk for sexual activity for Hispanic adolescents. Christopher, Johnson, &
Roosa (1993) found no relationship between mother’s warmth, open communication,
and communication problems and sexual involvement for Hispanic males and
females. In contrast, Miller, Forehand, & Kotchick (1999) found for Puerto Rican
adolescents that good general communication with the mother was negatively
associated with sexual activity.
113
Peer Influence: Gender & Race Differences
The results from studies that examined the peer influence and adolescent
sexual activity relationship by gender are mixed. Many researchers found that
spending time with deviant peers was positively associated with sexual activity for
both males and females (Billy & Udry, 1985; Whitbeck, Conger, & Kao, 1993;
Whitbeck, Simons, & Goldberg, 1996). However, some researchers found gender
differences for this relationship. Metzler et al. (1994) found that females who were
less involved with their families were more likely to be susceptible to their peers and
become sexually experienced than males. Because of their deeper emotional
involvement with their parents, Noller & Callan (1991) found that friends might have
more of an influence on a female’s decision to engage in sexual behaviors than a
male’s decision to engage in these activities for those who reside in unsupportive
family environments.
A review of the literature revealed mixed results by race regarding the
relationship between peer’s influence and adolescent sexual activity. Furstenberg et
al. (1987) found for black adolescents that those who are associated with sexually
experienced peers are more likely to initiate sex. In contrast, Billy & Udry (1985)
found for black adolescents no significant association between the sexual behavior of
friends and their decision to engage in sexual activity.
Researchers have found a strong association between peer influence and
adolescent sexual activity for Hispanic adolescents. Christopher, Johnson, & Roosa
(1993) found among Hispanic adolescents of both genders that the best friend
114
perceived behavior was significantly associated with sexual involvement. However,
the effect that peer influence may have on adolescent sexual behavior may depend on
the adolescent’s age. Gibson & Kempf (1990) found that the virginity status of
young Latinas was highly correlated with perception of friends’ attitudes and
behaviors regarding sexual activity. For older Latinas, the adolescent’s own attitude,
rather than the behaviors of friends, was a much stronger predictor of whether they
would remain virgins.
In contrast, researchers found that peers could discourage sexual activity for
Asian adolescents. Schuster et al. (1998) found that Asian adolescents are more
likely to think that their friends disapproved of them having sex and that it would be
acceptable for an adolescent male not to be sexually experienced when compared to
white, black, and Hispanic adolescents. The percentage of Asian females who think
that it is acceptable for their friends to be sexually inexperienced were similar to the
percentage of other races but significantly lower than black females. Although the
peers of Asian adolescents may tend to discourage sexual activity, it is unknown
whether these peers have engaged in deviant activities.
115
Chapter 6: Gender Context Analyses
Before examining the best reason for the family structure and adolescent
sexual activity relationship by the joint impact of race and gender, it seemed
important, as a baseline, to analyze this relationship within the adolescent’s gender
and adolescent’s race contexts taken separately. Thus, the main objective for this
chapter is to determine the best reason for the family structure and adolescent sexual
activity relationship among the three non-traditional households within the gender
context. This chapter also includes the t-test for difference between slopes analysis,
as explained by Jaccard, Turrisi, & Wan (1990), in order to determine whether there
are significant gender differences for the relationship between the variables that
comprise the five explanations and adolescent sexual activity. Finally, the chapter
concludes with a discussion of the findings.
Hypotheses
Despite the literature showing gender differences in the relationship between
the five measures and adolescent sexual activity, the hypotheses for this investigation
are based on the results from chapter 4. This is because the literature review
regarding gender differences may not apply to all adolescents since many of them
examined only white adolescents. However, this gender analyses are conducted on
adolescents of various racial backgrounds. In addition, the results in the family
literature are generally mixed for the five explanations and adolescent sexual activity
116
relationship within the gender context. Consequently, it would be very difficult to
hypothesize how the relationship between these influences and adolescent sexual
activity truly differ between males and females. Establishing the hypotheses based
on the results from chapter 4 and discussing the ones that do not support these
hypotheses would be one way to resolve this situation.
The hypotheses for this investigation are as follows:
1. Family transition, mother-child discussions, mother-child sex discussions,
and acquaintance with deviant peers will be positively associated with sexual
activity for all adolescents regardless of their gender.
2. Mother-child activities, parental strictness, and parental meal presence,
mother’s religiosity, mother’s disapproval of adolescent sexual activity, and
family support will be negatively associated with sexual activity for all
adolescents regardless of their gender.
3. None of the parental monitoring variables (mother’s full time employment
and mother’s morning, afternoon, and evening presence) will be associated
with sexual activity for all adolescents regardless of their gender.
Descriptive Results
Table 6-1 presents the means for the sexual activity, the demographic
variables, and the variables that constitute the five explanations by adolescent’s
gender. Females are less likely than males to be sexually experienced. However,
they are also more likely than males to experience doing activities with and to have
117
Table 6-1: Mean Scores for the Sexual Activity, Demographic, Family Influence,
and Peer Influence Variables By Adolescent’s Gender
Males Females
Sexually Experienced 0.39 0.34**
Child's Age 16.01 15.81 ***
Whites 0.66 0.64
Blacks 0.21 0.22
Hispanics 0.10 0.11
Asians 0.03 0.02
Two Biological Parents 0.60 0.60
Unwedded Mothers 0.05 0.06
Divorced/Separated Mothers 0.20 0.20
Stepparent 0.15 0.14
Mother's Education 2.81 2.80
Household Income 48.15 49.16
Missing Income 0.10 0.10
Family Transitions 1.31 1.31
Mother-Child Activities 1.54 1.78***
Mother-Child Discussions 2.18 2.64***
Parent Strictness 1.23 1.10**
Parent Meal Presence 4.79 4.69
Full Time Employment 0.58 0.59
Presence During Morning 4.18 4.21
Presence During Afternoon 3.24 3.30
Presence During Night 4.66 4.67
Mother's Religiosity 6.31 6.24
Sex Communication 2.90 3.11**
Sex Attitudes 4.38 4.40
Family's Support 16.35 16.15 *
Deviant Peers 2.58 2.33 **
N 1886 2130
Notes: * = p < .05; ** = p < .01; *** = p < .001
118
both general and sex discussions with their mothers. In addition, males are more
likely than females to have stricter parents, to experience greater family support, and
to have acquaintances with greater number of friends who engage in deviant
activities.
Test of Hypotheses Results
Tables 6-2 and 6-3 display the adolescent sexual activity regressions on the
demographic, family influence, and peer influence variables by gender. The baseline
models show that males and females who are black and children who reside in non-
traditional households are positively associated with sexual activity. However,
mother’s education and being Hispanic are negatively associated with sexual activity
for males and females respectively. There is a significant gender difference
regarding the adolescent sexual activity risk for black, Hispanic, and stepparent
households. Black females and Latinas have a significantly decreased risk for sexual
activity compared to black males and Latinos respectively. Furthermore, females
who reside with a stepparent have a significantly increased risk for sexual activity
compared to males who reside in these households.
The relationship between the variables that constitute the five explanations
and adolescent sexual activity are similar to those found in Chapter 4 for males and
females. Family transitions, mother-child discussions, mother-child sex discussions,
and acquaintance with deviant peers are positively associated with sexual activity for
adolescents of both genders. In addition, mother-child activity, parental strictness,
119
Table 6-2: Logistic Regression of Adolescent Sexual Activity on the Demographic, Family Transition, and Parental Supervision
Variables By Gender (N = 1886 for males and 2130 for females)
All Males All Females All Males All Females
Beta Exp (B) Beta Exp (B) Beta Exp (B) Beta Exp (B)
Intercept -8.17 *** -10.27 *** -8.68 *** -10.60 ***
Child's Age 0.51 *** 1.66 0.62 *** ^ 1.86 0.52 *** 1.68 0.63 *** ^ 1.87
Black
a
1.00 *** 2.73 0.38 ** ^ 1.46 1.07 *** 2.90 0.42 ** ^ 1.52
Hispanic
a
0.11 1.12 -0.50 ** ^ 0.61 0.13 1.13 -0.49 ** ^ 0.61
Asian
a
-0.56 0.57 -0.71 0.49 -0.54 0.59 -0.70 0.50
Unwedded Mothers
b
0.67 * 1.95 0.67 ** 1.95 0.59 * 1.81 0.59 * 1.80
Divorced/Separated Mothers
b
0.65 *** 1.92 0.71 *** 2.04 0.47 ** 1.60 0.60 *** 1.83
Stepparents
b
0.43 ** 1.53 0.88 *** ^ 2.41 0.13 1.14 0.69 *** ^ 2.00
Mother's Education -0.18 *** 0.84 -0.09 0.92 -0.18 ** 0.84 -0.08 0.92
Household Income (ln) -0.12 0.88 -0.14 0.87 -0.12 0.89 -0.14 0.87
Missing Household Income -0.04 0.96 -0.08 0.93 -0.03 0.97 -0.06 0.94
Family Transitions 0.29 ** 1.33 0.19 * 1.21
-2 LL 2106.21 2222.45 2095.09 2216.73
Chi-Square 415.82 *** 512.06 *** 426.95 *** 517.78 ***
Nagelkerke R
2
0.27 0.30 0.28 0.30
BIC -340.40 -435.42 -343.99 -433.48
Notes: * = p < .05; ** = p < .01; *** = p < .001; ^ = gender difference;
a
= whites as the reference group;
b
= two biological
parents as the reference group
120
Table 6-2: Continued
All Males All Females
Beta Exp (B) Beta Exp (B)
Intercept -5.51 *** -6.95 ***
Child's Age 0.41 *** 1.50 0.52 *** ^ 1.68
Black
a
1.09 *** 2.97 0.42 ** ^ 1.52
Hispanic
a
0.14 1.15 -0.48 * ^ 0.62
Asian
a
-0.57 0.57 -0.60 0.55
Unwedded Mothers
b
0.50 1.64 0.48 * 1.61
Divorced/Separated Mothers
b
0.43 ** 1.53 0.41 ** 1.50
Stepparents
b
0.28 1.32 0.74 *** ^ 2.10
Mother's Education -0.20 *** 0.82 -0.10 0.91
Household Income (ln) -0.20 * 0.82 -0.22 ** 0.80
Missing Household Income -0.01 0.99 -0.06 0.94
Mother-Child Activities -0.27 *** 0.76 -0.32 *** 0.73
Mother-Child Discussions 0.23 *** 1.26 0.24 *** 1.27
Parent Strictness -0.14 ** 0.87 -0.20 *** 0.82
Parent Meal Presence -0.06 * 0.94 -0.11 *** 0.90
Mother's Employment Status 0.05 1.05 0.13 1.14
Mother's Morning Presence -0.03 0.97 -0.08 0.92
Mother's Afternoon Presence -0.06 0.95 -0.02 0.98
Mother's Evening Presence -0.02 0.98 -0.07 0.93
-2 LL 2028.44 2084.88
Chi-Square 493.59 *** 649.64 ***
Nagelkerke R
2
0.31 0.36
BIC -357.83 -511.69
Notes: * = p < .05; ** = p < .01; *** = p < .001; ^ = gender difference;
a
= whites
as the reference group;
b
= two biological parents as the reference group
121
Table 6-3: Logistic Regression of Adolescent Sexual Activity on the Demographic,
Parental Attitudes, Family Support, and Peer Influence Variables By
Gender (N = 1886 for males and 2130 for females)
All Males All Females
Beta Exp (B) Beta Exp (B)
Intercept -6.99 *** -9.58 ***
Child's Age 0.47 *** 1.60 0.61 *** ^ 1.85
Black
a
1.04 *** 2.82 0.46 ** ^ 1.59
Hispanic
a
0.20 1.22 -0.39 * ^ 0.68
Asian
a
-0.33 0.72 -0.50 0.61
Unwedded Mothers
b
0.41 1.51 0.41 1.51
Divorced/Separated Mothers
b
0.46 ** 1.58 0.54 *** 1.72
Stepparents
b
0.28 1.32 0.73 *** ^ 2.08
Mother's Education -0.15 ** 0.86 -0.04 0.96
Household Income -0.12 0.88 -0.15 0.86
Missing Household Income -0.07 0.93 -0.09 0.92
Mother's Religiosity -0.10 ** 0.90 -0.14 *** 0.87
Mother-Child Sex Discussions 0.35 *** 1.42 0.36 *** 1.44
Mother Disapprove Sex -0.24 *** 0.79 -0.20 *** 0.82
-2 LL 2041.37 2146.78
Chi-Square 480.67 *** 587.73 ***
Nagelkerke R
2
0.31 0.33
BIC -382.62 -488.10
Notes: * = p < .05; ** = p < .01; *** = p < .001; ^ = gender difference;
a
= whites
as the reference group;
b
= two biological parents as the reference group
122
Table 6-3: Continued
All Males All Females All Males All Females
Beta Exp (B) Beta Exp (B) Beta Exp (B) Beta Exp (B)
Intercept -6.20 *** -7.85 *** -7.28 *** -10.88 ***
Child's Age 0.49 *** 1.64 0.63 *** ^ 1.87 0.39 *** 1.48 0.59 *** ^ 1.80
Black
a
1.06 *** 2.88 0.41 ** ^ 1.51 1.41 *** 4.08 0.86 *** ^ 2.37
Hispanic
a
0.13 1.14 -0.54 ** ^ 0.58 0.18 1.20 -0.26 0.77
Asian
a
-0.65 0.52 -0.80 * 0.45 -0.46 0.63 -0.72 0.49
Unwedded Mothers
b
0.72 ** 2.06 0.64 ** 1.89 0.58 * 1.78 0.53 * 1.69
Divorced/Separated Mothers
b
0.63 *** 1.87 0.64 *** 1.90 0.48 ** 1.62 0.54 *** 1.71
Stepparents
b
0.38 * 1.46 0.79 *** ^ 2.21 0.35 * 1.41 0.78 *** ^ 2.19
Mother's Education -0.19 *** 0.83 -0.08 0.92 -0.20 *** 0.82 -0.04 ^ 0.96
Household Income -0.12 0.88 -0.13 0.87 -0.14 0.87 -0.13 0.88
Missing Household Income -0.04 0.96 -0.01 0.99 -0.05 0.95 -0.06 0.94
Family Support -0.11 *** 0.90 -0.16 *** ^ 0.85
Deviant Peers 0.38 *** 1.46 0.33 *** 1.39
-2 LL 2080.48 2157.30 1825.48 1987.40
Chi-Square 441.55 *** 577.21 *** 696.56 *** 747.11 ***
Nagelkerke R
2
0.28 0.33 0.42 0.41
BIC -358.59 -492.91 -613.60 -662.81
Notes: * = p < .05; ** = p < .01; *** = p < .001; ^ = gender difference;
a
= whites as the reference group;
b
= two biological
parents as the reference group
123
parental meal presence, mother’s religiosity, mother’s disapproval of adolescent
sexual activity, and family support are negatively associated with sexual activity for
both males and females. There is a significant gender difference for the family
support variable where this effect is significantly stronger for females than for males
(t-test for difference between slopes with p < .05).
The deviant peer measure is the most parsimonious adolescent sexual activity
explanation for both genders according to the BIC statistic. Mother’s attitude and
parental supervision are the most parsimonious adolescent sexual activity
explanations for males and females respectively. Family transition is the least
parsimonious sexuality explanation among adolescents of both genders.
Table 6-4 summarizes the family structure and adolescent sexual activity
relationship for the three non-traditional households by gender. The first column
displays the non-traditional household configurations by the adolescent’s gender.
The second column indicates whether the relationship between the non-traditional
household and adolescent sexual activity is statistically significant. The third
column indicates whether there is a significant non-traditional household effect by
gender when accounting only for the demographic variables. The fourth and fifth
columns indicate which explanation partially and statistically respectively explains
the non-traditional household effect. An asterisk indicates the explanation that
produces the largest drop in beta (otherwise known as the best explanation that best
mediates the family structure effect) for that particular non-traditional household.
The parenthesis surrounding an explanation indicates that this particular explanation
would have been the best one to mediate that particular non-traditional household
124
Table 6-4: Summary of the Family Structure and Adolescent Sexual Activity Relationship By Gender
Significant (Yes/No) Gender Difference Partially Explained By Fully Explained By
Males
Unwedded Mothers Yes PS, PA *
Divorced/Separated Mothers Yes FT, PS *, PA, DP
Stepparent Yes FT *, PS, PA
Females
Unwedded Mothers Yes PS PA
Divorced/Separated Mothers Yes PS
Stepparent Yes Yes (FT)
Notes: FT = Family Transitions; PS = Parental Supervision; PA = Parental Attitudes; FS = Family Support; DP = Deviant Peers
125
effect in cases where none of the explanations were able to meet the level of partial
mediation.
In essence, the best reason for the family structure and adolescent sexual
activity relationship for the three non-traditional households for both males and
females is similar to the ones found in chapter 4. Mother’s attitude is the best reason
for the unwedded mother household effect for both males and females. Meanwhile,
the best reason for the divorced/separated mother household effect for adolescents of
both genders is parental supervision. Furthermore, family transition is the best
reason the stepparent household effect for males. Although none of the explanations
met the level for partial mediation of the stepparent household effect for females,
family transition, as for males, is the best reason for this relationship.
Discussion
The main reason for these analyses was to determine the best reasons for the
family structure and adolescent sexual activity relationship among the three non-
traditional households within the gender context using four family influence and a
deviant peer explanations. This investigation initially found that all of the non-
traditional households were positively associated with sexual activity within the
gender context. In another words, adolescents who reside in non-traditional
households are more likely than those who reside with two biological parents to
engage in sexual activity regardless of their gender. The initial investigation also
126
found that there was a gender difference for the stepparent effect where this
relationship is stronger for females than for males.
However, while the family transitions, parental supervision, and parental
attitudes explanations were able to fully mediate the stepparent household effect for
males, none of these explanations was able to at least meet the criteria of partial
mediation for females. For both males and females, family transition provided the
best reason for this effect. The results from this study would seem to suggest that
there might be multiple reasons why females who reside with a stepparent have an
increased risk for sexual activity compared to those from two biological parent
households since none of the explanations were able to individually fully account for
the stepparent household effect. Meanwhile, males who reside with a stepparent
seem to have an increased risk for sexual activity perhaps because they may view
their home as being turbulent due to the stresses of integrating a new parental figure.
Other explanations for this stepparent effect may also include lower levels of parent-
child interactions or mothers being more tolerant of adolescent sexual activity. Thus,
the reasons for the stepparent household effect for females seem to be multifaceted,
while for males, this effect may be due to any number of issues with the primary one
being the stress of living with a new father.
The best explanation for the divorced/separated mother household effect did
not differ between males and females. Lower levels of parent-child interactions
seem to be the best reason why adolescents who reside with a divorced/separated
mother have an increased risk for sexual activity. However, for males, there are
other reasons that could partially explain the divorced/separated mother household
127
effect including struggling with living in chaotic household, having mothers who
possess more tolerant sexual attitudes, and being influenced by their friends and
peers. In a situation that is similar to the stepparent household effect, the
divorced/separated household effect for males may be due to any number of reasons
with the primary one being the lower levels of interaction with his mother. By
contrast, low level of parent-child interactions was the best reason for the increased
sexual activity risk among females who reside with divorced/separated mothers.
This finding suggests that the time that mothers spend with their daughters is
important because the lack of spending time with her may directly lead to sexual
activity during adolescence possibly due to the need to filling the void due to the lack
of a mother-daughter relationship.
The best reason for the unwedded mother household effect for both males
and females was mother attitudes. This effect is probably due to unwedded mothers
being more tolerant of adolescent sexual activity than mothers of two biological
parent households. These findings also suggest that unwedded mother households
are environmentally different compared to divorced/separated mother households
which support Moore’s (2001) argument that both of these households do not share
similar family environments and that they should be examined separately whenever
possible.
Although there are some gender differences regarding the amount of
socialization that children receive from their parents, the results of this investigation
revealed that for both males and females, the relationship between both family and
peer influences were similar to those found in chapter 4. These findings may be due
128
the diminishing gender disparity regarding adolescent sexual activity at any given
age in recent cohorts (Davis & Friel, 2001). Consequently, it may be presumed that
the factors having an association with sexual behaviors will not differ significantly
between males and females.
The family support effect was found to be significantly stronger for females
than for males. Compared to males, it seems that females who feel supported by
their family members are less likely seek this support from individuals outside of the
family. Family support may be more important for females than for males because
of their need to learn how to nurture others (Moore & Rosenthal, 1993). Thus, the
development of this skill may be further encouraged through the experience of being
supported by immediate family members while growing up.
Peer influence was the most parsimonious adolescent sexual activity
explanation for both males and females. However, there was a gender difference for
the most parsimonious family influence explanation. Parental attitude and parental
supervision were the most parsimonious adolescent sexual activity explanations for
males and females respectively. This gender difference may be due to the relatively
higher levels of interaction between mothers and their daughters compared to
mothers and their sons. According to Chodorow (1974), the mother’s interaction
with her daughters is important because this is where female identity formation takes
place. Since daughters tend to identify with their mothers since they are similar in
gender, the interactions that they have with each other may help daughters to develop
their own gender identity.
129
The interactions that mothers have with their sons are somewhat less
important compared to those that she would have with her daughters because sons
are more likely to separate themselves from their mother’s identity (Chodorow,
1974). This separation process may be the reason why males were less likely than
females to experience doing activities and having discussions with their mothers. In
addition, this premise also suggests that the interactions with the fathers would be
important in the formation of a son’s identity. However, the results of this study
found that mothers still have significant influence on their son’s decision to engage
in sexual activity through the transmission of sexual values.
Family transition was the least parsimonious adolescent sexual activity
explanation for both males and females. As mentioned in chapter 4, this finding may
due to family transition being indirectly associated with adolescent sexual activity
because other explanation(s) could be explaining this effect. This finding may also
be due to the family transition measure only representing the number of marital or
family transition events that children may have potentially experienced during their
lifetime rather than the quality or degree of disruption and emotional chaos of these
occurrences.
Chapter Summary
The results of this investigation revealed that there were many similarities
regarding the best explanation for family structure and adolescent sexual activity
relationship among three non-traditional households for both males and females.
130
Family transition, either due to either adjustment to the new family configuration or
to the stresses of having gone through multiple transitions, was the best explanation
for the stepparent household effect among males and females. For males, other
influences were able to fully account for this effect including lower levels of parent-
child interactions and mothers possessing more tolerant sexual attitudes. In contrast,
for females, none of the influences was able to reach the level of partial mediation of
the stepparent effect. This finding suggests that the reasons for the stepparent
household effect for females seem to be multifaceted, while for males, this effect
may be due to any number of issues with the primary one being the stress of living
with a new father.
Either lower levels of parent-child interaction or a weakened parental
authority may be the best reason for the divorced/separated mother effect for both
males and females. However, for males, family turbulence, mothers possessing more
tolerant sexual attitudes, and being influenced by friends and peers were also other
explanations that partially mediated this effect. By contrast, for females, parental
supervision was the only explanation that reached the level of partial mediation for
the divorced/separated parent effect. Thus, for males, the divorced/separated
household effect may be due to any number of reasons with the primary one being
the lower levels of interaction with his mother, while for females, the time that
mothers spend with their daughters is important because the lack of spending time
with her may directly lead to sexual activity during adolescence possibly due to the
need to filling the void due to the lack of a mother-daughter relationship.
131
Mothers who are more tolerant of adolescent sexual activity may be the best
reason for the unwedded mother household effect for both males and females. This
household seems to be somewhat different from divorced/separated mother
households because they may not experience the family turbulence unlike those who
reside in divorced/separated households.
Peer influence was the most parsimonious adolescent sexual activity
explanation for both genders. However, mother’s attitude was the most
parsimonious family influence explanation for males while parental supervision was
the most parsimonious family influence explanation for females. This gender
difference may be due to the mother-child interaction being a greater importance for
personal identity formation for females than for males. Although mother-child
interaction levels were lower for males than for females, mothers still have an
influence over their son’s decision to engage in sexual activity through the
transmission of attitudes and values.
Meanwhile, family transition was the least parsimonious adolescent sexual
activity explanation for male and females. One reason for this finding is that this
measure has an indirect relationship with adolescent sexual activity and that other
explanation(s) could explain this effect. Another reason for this finding may be that
this measure represents only the number of marital or family transition events
children may potentially experience during their lifetime rather than the quality or
degree of disruption and emotional chaos of these events.
The results of this investigation revealed that the significant family influence
and peer influence variables regarding adolescent sexual activity were similar to
132
those found in chapter 4 for both males and females. The exception to this finding is
that the family support effect was stronger for females than for males suggesting the
importance of family support for females in developing the skills to nurture others.
This investigation also revealed that a conflict measure, in this case peer
influence, was a better explanation for adolescent sexual activity than the functional
measures for males and females. However, functional measures were a better
explanation for the single parent household effect than conflict measures.
Meanwhile, a conflict measure, in this case family transition, is the one that best
explained the stepparent effect. It seems that mothers have more influence on their
children when they are the only ones in charge of the household. In addition,
mothers seem to have less of an influence over their children when there is another
parental figure in the household, which in this case, is the stepfather. But, it is
difficult to pinpoint whether mothers of stepparent households have less influence
over their children is due to the stresses of adjusting to a different household
configuration once the mother remarry or to mothers reenacting the traditional
mother role by deferring a part of their parental authority to the stepfather.
133
Chapter 7: Racial Context Analyses
The analysis for this investigation is similar to that of the previous chapter
except that the analysis is conducted within three racial categories. This
investigation also includes a t-test for difference between slopes analysis, as
explained by Jaccard, Turrisi, & Wan (1990), in order to determine whether there are
significant racial differences for the variables that comprise the five explanations and
adolescent sexual activity relationship
1
. The chapter concludes with a discussion of
the results.
Asian adolescents will not be discussed in this investigation because the
results from this group may not be statistically valid due to small sample size. A
number of sources, such as those by Cohen (2001) and Newton & Rudestam (1999),
have provided several guidelines for determining the minimal number of respondents
one needs in order for the multivariable regression analyses results to be statistically
valid
2
. According to these various guidelines, the minimal number of respondents
that is needed for this investigation in order for the results to be considered
statistically valid is 160. However, a preliminary investigation revealed that the
Asian adolescent subgroup do not meet this criteria. Although the results may not be
1
White adolescents are the comparison group for the race difference analyses.
2
According to Cohen (2001) there are three rules of thumb that one can use to
determine the minimal sample size needed to generate valid statistical results
regarding multiple variable regressions analyses. The first rule is the number of
predictors plus 41. The second rule is the number of predictors times 10. The third
rule is eight times the number of predictors plus 50. For stepwise regressions, the
rule of thumb to generate valid statistical results is 30 to 40 subjects for each
predictor.
134
statistically valid, the results of Asian adolescents will be presented but not discuss
in order to provide a complete the examination of the data.
Hypotheses
Even though there may be some race differences regarding some of the
relationships between the five explanations and adolescent sexual activity as pointed
out by the family literature, the hypotheses of this investigation will be the based on
the results that were found in chapter 4. The preponderance of the literature that was
presented in chapter 5 seems to suggest no racial differences regarding these
relationships. However, there were some exceptions to this premise, such as racial
differences found for family transition and adolescent sexual activity relationship
between whites and blacks (Wu & Thomson, 2001). Nevertheless, this investigation
will only discuss significant racial differences findings for the five explanations and
adolescent sexual activity relationship in order to highlight the noteworthy results.
The hypotheses for this investigation are as follows:
1. Family transition, mother-child discussions, mother-child sex discussions,
and acquaintance with deviant peers will be positively associated with sexual
activity for all adolescents regardless of their race.
2. Mother-child activities, parental strictness, parental meal presence, mother’s
religiosity, mother disapproval of adolescent sex, and family support will be
negatively associated with a sexual activity for all adolescents regardless of
their race.
135
3. None of the parental monitoring variables (mother’s full time employment
and mother’s morning, afternoon, and evening presence) will be associated
with sexual activity for all adolescents regardless of their race.
Descriptive Results
Table 7-1 presents the means for the adolescent sexual activity, the
demographic variables and the variables that constitute the five explanations within
race categories. Black adolescents are more likely than white adolescents to have
engaged in sexual activity. Both black and Hispanic adolescents are more likely than
white adolescents to reside in single parent households. These adolescents are also
less likely than white adolescents to reside with two biological parents. The
households of black and Hispanic adolescents are also poorer than the households of
white adolescents. Both black and Hispanic adolescents are less likely than white
adolescents to have a parent eat meals with them. Black adolescents are less likely
than white adolescents to have acquaintance with friends who engage in deviant
activities. Meanwhile, Hispanic adolescents are less likely than white adolescents to
engage in activities with their mothers.
Mothers of both black and Hispanic adolescents are more likely than mothers
of white adolescents to be religious, to have greater conversations their children
about sex, and to be tolerant of adolescent sexual activity. Mothers of black
adolescents are more likely than mothers of white adolescents to be employed full
time. Mothers of Hispanic adolescents are less likely than mother of white
136
Table 7-1: Mean Scores for the Adolescent Sexual Activity, Demographic, Family
Influence, and Peer Influence Variables Within Race Categories
Whites Blacks Hispanics Asians
Sexual Initiation 0.33 0.51 *** 0.34 0.21 **
Child's Age 15.89 15.93 15.92 15.96
Male 0.48 0.45 0.46 0.52
Two Biological Parents 0.68 0.37 *** 0.53 *** 0.80 **
Unwedded Mothers 0.02 0.18 *** 0.07 *** 0.02
Divorced/Separated Mothers 0.17 0.30 *** 0.23 ** 0.08 **
Stepparent 0.14 0.15 0.16 0.11
Mother's Education 2.89 2.86 2.04 *** 3.31 **
Household Income 55.05 35.97 *** 34.01 *** 55.42
Missing Income 0.09 0.11 0.14 * 0.15
Family Transitions 1.31 1.27 1.40 * 1.14 **
Mother-Child Activities 1.66 1.74 1.54 * 1.74
Mother-Child Discussions 2.44 2.45 2.36 2.13 *
Parent Strictness 1.07 1.35 *** 1.37 *** 1.18
Parent Meal Presence 5.01 3.96 *** 4.67 ** 4.68
Full Time Employment 0.55 0.70 *** 0.52 0.68 **
Presence During Morning 4.23 4.14 4.20 3.85 *
Presence During Afternoon 3.22 3.30 3.51 *** 3.23
Presence During Night 4.65 4.66 4.77 ** 4.59
Mother's Religiosity 6.01 6.96 *** 6.42 *** 6.38
Sex Communication 2.97 3.31 *** 2.85 * 2.41 ***
Disapprove Sex 4.46 4.26 *** 4.17 *** 4.55
Family's Support 16.24 16.39 16.11 15.80
Deviant Peers 2.60 2.02 *** 2.46 2.06 *
N 2614 872 426 104
Notes: * = p < .05; ** = p < .01; *** = p < .001
137
adolescents to receive higher educational attainment. However, mothers of Hispanic
adolescents are more likely than mothers of white adolescents to have serious
relationships over the past 18 years and be present at home during afternoons and
evenings.
Test of Hypotheses Results
Table 7-2 displays the adolescent sexual activity regressions on the
demographic variables within race categories. The results show that residing in non-
traditional households is positively associated while mother’s education and
household income levels are negatively associated with sexual activity for white
adolescents. Being a male and children who reside in stepparent households are
positively associated with sexual activity for black adolescents. Being a male and
children who reside in non-traditional households are positively associated with
sexual activity for Hispanic adolescents.
A significant race difference is found between white and black adolescents
for gender, the unwedded mother, and the divorced/separated mother household
variables (t-test for difference between slopes with p < .05). Black males have a
significantly increased risk for sexual activity compared to white males. The single
parent effect is significantly stronger for white than for black adolescents.
Another significant race difference is found between white and Hispanic
adolescents for both the gender and the stepparent household variables. Latinos have
a significantly increased risk for sexual activity compared to white males. In
138
Table 7-2: Logistic Regression of Adolescent Sexual Activity on the Demographic Variables Within Race Categories
Whites Blacks Hispanics Asians
Beta Exp (B) Beta Exp (B) Beta Exp (B) Beta Exp (B)
Intercept -10.61 *** -6.26 *** -10.19 *** -6.79 *
Child's Age 0.67 *** 1.94 0.41 *** ^ 1.51 0.53 *** ^ 1.70 0.32 * 1.38
Male
a
-0.07 0.93 0.58 *** ^ 1.79 0.57 * ^ 1.77 0.10 1.10
Unwedded Mothers
b
1.60 *** 4.93 0.31 ^ 1.37 1.02 * 2.76 -6.42 0.00
Divorced/Separated Mothers
b
0.82 *** 2.27 0.31 ^ 1.37 1.06 ** 2.89 -0.49 0.61
Step Parent
b
0.57 *** 1.77 0.69 ** 1.99 1.21 *** ^ 3.34 -0.63 0.53
Mother's Education -0.18 *** 0.84 -0.10 0.91 -0.03 0.97 -0.35 0.71
Household Income (ln) -0.17 * 0.85 -0.14 0.87 0.09 1.09 0.42 1.51
Missing Household Income 0.06 1.06 -0.18 0.83 -0.19 0.83 -1.78 0.17
-2 LL 2665.26 1066.67 457.57 94.92
Chi-Square 634.99 *** 141.95 *** 88.80 *** 12.40
Nagelkerke R
2
0.30 0.20 0.26 0.18
N 2614 872 426 104
BIC -572.04 -87.78 -40.36 24.76
Notes: * = p < .05; ** = p < .01; *** = p < .001; ^ = race difference;
a
= female as the reference group;
b
= two biological parents
as the reference group
139
addition, the stepparent effect is significantly stronger for Hispanic than for white
adolescents.
Table 7-3 displays the adolescent sexual activity regressions on the
demographic and family transition variables within race categories. The results
show that mother’s serious relationships over the past 18 years are positively
associated with sexual activity only for white and black adolescents. A racial
difference is found for the family transition effect between whites and Hispanic
adolescents. There is an unusual non-significant finding regarding the family
transition effect for Hispanic adolescents where family transitions is negatively
associated with sexual activity even though these mothers report going through more
serious or marriage like relationships over the past 18 years compared to mothers of
white adolescents. A cross tabulation analysis (results not shown) revealed that
Hispanic mothers who reported greater number of serious relationships are more
likely to have a child who is a virgin.
Table 7-4 displays the adolescent sexual activity regressions on the
demographic and parental supervision variables within race categories. The results
show that parent-child activities, parental strictness, parental meal presence, and
mother’s night presence are negatively associated while parent-child discussions are
positively associated with sexual activity for white adolescents. There is a racial
difference for mother’s night presence between white and both black and Hispanic
adolescents. The effect of mother’s night presence is stronger for white than for
either black or Hispanic adolescents.
140
Table 7-3: Logistic Regression of Adolescent Sexual Activity on the Demographic and Family Transition Variables Within Race
Categories
Whites Blacks Hispanics Asians
Beta Exp (B) Beta Exp (B) Beta Exp (B) Beta Exp (B)
Intercept -11.22 *** -6.60 *** -9.84 *** -5.78
Child's Age 0.68 *** 1.97 0.42 *** ^ 1.52 0.52 *** ^ 1.69 0.31 1.36
Male
a
-0.07 0.93 0.58 *** ^ 1.79 0.57 * ^ 1.76 0.11 1.11
Unwedded Mothers
b
1.40 *** 4.05 0.29 ^ 1.34 1.22 * 3.40 -6.03 0.00
Divorced/Separated Mothers
b
0.58 *** 1.79 0.23 1.26 1.22 *** ^ 3.38 -0.19 0.83
Stepparent
b
0.18 1.20 0.53 * 1.70 1.45 *** ^ 4.26 -0.34 0.71
Mother's Education -0.16 ** 0.85 -0.10 0.90 -0.02 0.98 -0.35 0.71
Household Income -0.17 * 0.85 -0.13 0.88 0.08 1.09 0.40 1.50
Missing Household Income 0.06 1.06 -0.15 0.86 -0.21 0.81 -1.77 0.17
Family Transitions 0.36 *** 1.43 0.22 * 1.24 -0.26 ^ 0.77 -0.68 0.51
-2 LL 2647.73 1061.79 454.97 94.51
Chi-Square 652.52 *** 146.84 *** 91.40 *** 12.81
Nagelkerke R
2
0.31 0.21 0.27 0.18
N 2614 872 426 104
BIC -581.70 -85.90 -36.91 28.99
Notes: * = p < .05; ** = p < .01; *** = p < .001; ^ = race difference;
a
= female as the reference group;
b
= two biological parents
as the reference group
141
Table 7-4: Logistic Regression of Adolescent Sexual Activity on the Parental
Supervision Variables By Race Subgroups
Whites Blacks
Beta Exp (B) Beta Exp (B)
Intercept -6.90 *** -4.09 ***
Child's Age 0.55 *** 1.73 0.32 *** ^ 1.38
Male
a
0.02 1.02 0.73 *** ^ 2.08
Unwedded Mothers
b
1.22 ** 3.40 0.18 ^ 1.19
Divorced/Separated Mothers
b
0.54 *** 1.72 0.06 ^ 1.06
Stepparent
b
0.42 ** 1.53 0.51 * 1.66
Mother's Education -0.19 *** 0.83 -0.12 0.88
Household Income -0.25 ** 0.78 -0.22 0.80
Missing Household Income 0.09 1.10 -0.21 0.81
Mother-Child Activities -0.33 *** 0.72 -0.24 ** 0.79
Mother-Child Discussions 0.25 *** 1.28 0.26 *** 1.30
Parent Strictness -0.16 ** 0.85 -0.20 *** 0.82
Parent Meal Presence -0.08 *** 0.92 -0.06 0.94
Full Time Employment 0.02 1.03 0.17 1.19
Mother's Morning Presence -0.04 0.96 -0.09 0.91
Mother's Afternoon Presence -0.05 0.95 -0.02 0.98
Mother's Night Presence -0.16 * 0.85 0.05 ^ 1.05
-2 LL 2527.60 1011.42
Chi-Square 772.65 *** 197.20 ***
Nagelkerke R
2
0.36 0.27
N 2614 872
BIC -646.75 -88.87
Notes: * = p < .05; ** = p < .01; *** = p < .001; ^ = race difference;
a
= female as
the reference group;
b
= two biological parents as the reference group
142
Table 7-4: Continued
Hispanics Asians
Beta Exp (B) Beta Exp (B)
Intercept -10.17 *** -3.50
Child's Age 0.48 *** 1.61 0.34 1.41
Male
a
0.70 ** ^ 2.01 -0.48 0.62
Unwedded Mothers
b
1.03 * 2.79 -7.75 0.00
Divorced/Separated Mothers
b
0.91 ** 2.49 -1.02 0.36
Stepparent
b
1.15 *** ^ 3.16 -0.82 0.44
Mother's Education -0.01 0.99 -0.17 0.85
Household Income 0.07 1.08 0.09 1.09
Missing Household Income -0.28 0.76 -1.68 0.19
Mother-Child Activities -0.25 * 0.78 -0.13 0.88
Mother-Child Discussions 0.20 * 1.22 -0.02 0.98
Parent Strictness -0.08 0.92 0.12 1.13
Parent Meal Presence -0.11 * 0.90 -0.28 * 0.76
Full Time Employment 0.17 1.19 -1.63 0.20
Mother's Morning Presence -0.03 0.97 -0.35 0.70
Mother's Afternoon Presence 0.03 1.03 -0.44 0.65
Mother's Night Presence 0.25 ^ 1.28 0.52 1.69
-2 LL 438.51 83.51
Chi-Square 107.86 *** 23.81
Nagelkerke R
2
0.31 0.32
N 426 104
BIC -10.99 50.50
Notes: * = p < .05; ** = p < .01; *** = p < .001; ^ = race difference;
a
= female as
the reference group;
b
= two biological parents as the reference group
143
Table 7-5 displays the adolescent sexual activity regressions on the
demographic and parental attitude variables within race categories. The results show
that mother’s religiosity and mother’s disapproval of adolescent sexual activity are
negatively associated while mother’s sex discussion is positively associated with
sexual activity for white adolescents. However, there is a racial difference between
whites and blacks for mother’s religiosity and mother’s disapproval of adolescent
sexual activity where both of these variables are not associated with sexual activity
for black adolescents. In addition, there is a racial difference between whites and
Hispanics for mother’s disapproval of adolescent sexual activity in which this
variable is not associated with sexual activity for Hispanic adolescents.
Tables 7-6 and 7-7 show respectively the adolescent sexual activity
regressions on the demographic and both the family support and peer influence
variables within race categories. Family support is negatively associated, while
having greater acquaintance with deviant peers is positively associated with sexual
activity for white, black, and Hispanic adolescents. In addition, this effect is
significantly stronger for white than for black adolescents.
The peer influence measure is also the most parsimonious sexual activity
explanation for white, black, and Hispanic adolescents according to the BIC statistic.
Among the family influence measures, the BIC statistic shows that parental attitude
is the most parsimonious sexual activity explanation for white and Hispanic
adolescents. Meanwhile, family support is the most parsimonious sexual activity
explanation among black adolescents. This statistic also shows that family
transition, parental attitude, and parental supervision are the least parsimonious
144
Table 7-5: Logistic Regression of Adolescent Sexual Activity on the Demographic and Parental Attitude Variables Within Race
Categories
Whites Blacks Hispanics Asians
Beta Exp (B) Beta Exp (B) Beta Exp (B) Beta Exp (B)
Intercept -9.30 *** -6.69 *** -9.24 *** -5.57
Child's Age 0.64 *** 1.90 0.39 *** ^ 1.48 0.54 *** 1.72 0.33 1.39
Male
a
0.04 1.04 0.65 *** ^ 1.91 0.69 ** ^ 1.98 0.01 1.01
Unwedded Mothers
b
1.33 ** 3.78 0.27 ^ 1.31 0.60 1.83 -7.09 0.00
Divorced/Separated Mothers
b
0.59 *** 1.80 0.30 1.35 0.80 * 2.23 -0.44 0.64
Stepparent
b
0.36 * 1.44 0.66 ** 1.93 1.09 ** ^ 2.98 -0.33 0.72
Mother's Education -0.11 * 0.89 -0.10 0.91 -0.06 0.95 -0.37 0.69
Household Income -0.18 * 0.84 -0.13 0.88 0.04 1.04 0.56 1.76
Missing Household Income -0.01 0.99 -0.19 0.82 -0.22 0.81 -1.87 0.15
Mother's Religiosity -0.14 *** 0.87 -0.01 ^ 0.99 -0.24 ** 0.79 0.11 1.11
Sex Communication 0.39 *** 1.48 0.29 ** 1.34 0.31 * 1.37 -0.07 0.94
Disapprove Sex -0.32 *** 0.73 -0.05 ^ 0.95 -0.08 ^ 0.92 -0.52 0.59
-2 LL 2538.34 1055.92 438.14 91.25
Chi-Square 761.91 *** 152.71 *** 108.24 *** 16.07
Nagelkerke R
2
0.35 0.21 0.31 0.22
N 2614 872 426 104
BIC -675.35 -78.23 -41.64 35.02
Notes: * = p < .05; ** = p < .01; *** = p < .001; ^ = race difference;
a
= female as the reference group;
b
= two biological parents as
the reference group
145
Table 7-6: Logistic Regression of Adolescent Sexual Activity on the Demographic and Family Support Variables Within Race
Categories
Whites Blacks Hispanics Asians
Beta Exp (B) Beta Exp (B) Beta Exp (B) Beta Exp (B)
Intercept -8.48 *** -3.57 *** -8.66 *** -3.60
Child's Age 0.66 *** 1.94 0.39 *** ^ 1.48 0.52 *** ^ 1.68 0.27 1.32
Male
a
-0.06 0.94 0.68 *** ^ 1.97 0.64 ** ^ 1.91 0.11 1.12
Single Family
b
1.52 *** 4.59 0.31 ^ 1.36 1.07 * 2.92 -5.94 0.00
Divorced/Separated Family
b
0.77 *** 2.17 0.26 ^ 1.30 1.04 ** 2.83 -0.77 0.46
Step Family
b
0.51 *** 1.67 0.58 * 1.78 1.19 *** ^ 3.30 -0.72 0.49
Mother's Education -0.17 *** 0.84 -0.11 0.90 -0.05 0.95 -0.35 0.71
Household Income -0.16 * 0.85 -0.17 0.85 0.10 1.10 0.50 1.65
Missing Household Income 0.05 1.05 -0.10 0.91 -0.19 0.83 -1.38 0.25
Family Support -0.13 *** 0.88 -0.14 *** 0.87 -0.09 * 0.91 -0.18 0.84
-2 LL 2613.90 1041.57 452.40 92.72
Chi-Square 686.34 *** 167.06 *** 93.97 *** 14.60
Nagelkerke R
2
0.32 0.23 0.27 0.20
N 2614 872 426 104
BIC -615.52 -106.12 -39.48 27.20
Notes: * = p < .05; ** = p < .01; *** = p < .001; ^ = race difference;
a
= female as the reference group;
b
= two biological parents as
the reference group
146
Table 7-7: Logistic Regression of Adolescent Sexual Activity on the Demographic and Peer Influence Variables Within Race
Categories
Whites Blacks Hispanics Asians
Beta Exp (B) Beta Exp (B) Beta Exp (B) Beta Exp (B)
Intercept -10.65 *** -5.72 *** -10.44 *** -5.61
Child's Age 0.60 *** 1.82 0.34 *** ^ 1.40 0.50 *** 1.66 0.14 1.15
Male
a
-0.09 0.91 0.48 ** ^ 1.62 0.49 * ^ 1.63 0.16 1.18
Single Family
b
1.15 ** 3.14 0.36 1.43 1.15 * 3.17 -7.55 0.00
Divorced/Separated Family
b
0.60 *** 1.83 0.32 1.38 0.86 ** 2.36 -1.88 0.15
Step Family
b
0.42 ** 1.52 0.77 ** 2.15 1.07 ** ^ 2.93 -0.95 0.39
Mother's Education -0.15 ** 0.86 -0.09 0.91 -0.05 0.95 -0.34 0.71
Household Income -0.18 * 0.84 -0.12 0.89 0.06 1.07 0.66 1.93
Missing Household Income 0.07 1.07 -0.19 0.83 -0.23 0.79 -1.24 0.29
Deviant Peers 0.38 *** 1.47 0.28 *** ^ 1.32 0.31 *** 1.36 0.36 ** 1.43
-2 LL 2279.04 1002.57 414.00 84.79
Chi-Square 1021.21 *** 206.06 *** 132.38 *** 22.54 **
Nagelkerke R
2
0.45 0.28 0.37 0.30
N 2614 872 426 104
BIC -950.39 -145.12 -77.89 19.26
Notes: * = p < .05; ** = p < .01; *** = p < .001; ^ = race difference;
a
= female as the reference group;
b
= two biological parents as
the reference group
147
sexual activity explanations for sexuality for white, black, and Hispanic adolescents
respectively.
Table 7-8 summarizes the family structure and adolescent sexual activity
relationship for the three non-traditional households within the adolescent’s race
categories. The first column displays the non-traditional household configurations
by the adolescent’s race. The second column indicates whether the relationship
between the non-traditional household and adolescent sexual activity is statistically
significant. The third column indicates whether there is a significant non-traditional
household effect by race when accounting only for the demographic variables. The
fourth and fifth columns indicate which explanation partially and statistically
respectively accounts for the non-traditional household effect. An asterisk indicates
the explanation that produces the largest drop in beta (otherwise known as the
explanation that best mediates the family structure effect) for that particular non-
traditional household. The parenthesis surrounding an explanation indicates that this
particular explanation would have been the best one to mediate that particular non-
traditional household effect in cases where none of the explanations was able to meet
the level of partial mediation.
Peer influence, parental supervision, and family transition are the best
reasons for the unwedded mother, the divorced/separated mother, and the stepparent
household effects for white adolescents. Parental supervision is the best reason for
the stepparent household effect for black adolescents. Parental attitude is the best
reason for both the unwedded mother and divorced/separated mother household
effects for Hispanic adolescents.
148
Table 7-8: Summary of the Family Structure and Adolescent Sexual Activity Relationship Within Race Categories
Significant (Yes/No) Racial Difference Partially Explained By Fully Explained By
Whites
Unwedded Mothers Yes DP
Divorced/Separated Mothers Yes FT, PS *, PA, DP
Stepparent Yes PS, PA, DP FT
Blacks
Unwedded Mothers No Yes
Divorced/Separated Mothers No Yes
Stepparent Yes PS
Hispanics
Unwedded Mothers Yes PA
Divorced/Separated Mothers Yes PA
Stepparent Yes Yes (DP)
Notes: FT = Family Transitions; PS = Parental Supervision; PA = Parental Attitudes; FS = Family Support; DP = Deviant Peers
149
Discussion
These analyses examined the family structure and adolescent sexual activity
relationship for three non-traditional households within race categories using four
family influence explanations and a peer explanation. This section is divided into
three parts: the non-traditional household discussion, the family and peer influence
variables discussion, and the relative importance of the five explanations discussion.
Non-Traditional Household Discussion
This investigation initially found a significant racial difference between white
and black adolescents for both the unwedded mother and divorced/separated mother
household effects. There are several reasons why black adolescents who reside with
a single parent may have the same sexual activity risk as those who reside with two
biological parents. First, this finding may be due to the institutionalization of the
non-traditional households within the black community. The percentage of black
adolescents who reside with a single parent (48%) is similar to those who reside with
two-parents (52%). Although white adolescents who reside with two biological
parents may feel that they are in an advantaged environment compared to those who
reside in other non-traditional households (Moore, 2001), black adolescents who
reside with a single parent may not perceive themselves to be disadvantaged
compared to their peers who with two parents. However, the problem with this
premise is that black adolescents who reside with a stepparent are still at a higher
150
risk for sexual activity than those who reside with two biological parents even
though they may be considered to be living in an advantaged environment according
to the norms of the community.
The second reason for this finding may be due to the family structure
influences being weaker than cultural influences for the black community. The
relatively more tolerant attitudes regarding sexuality espoused by the black culture
(Staples, 1978) may be so pervasive throughout the community that its influence
could overcome the supposedly ideal family environment of who reside with two
biological parents. Essentially, this premise argues that the attitude and values for
black parents of two biological parent households would be similar to those for black
parents of non-traditional households. Consequently, one would not expect any
significant differences for the family structure and adolescent sexual activity
relationship among the three non-traditional households if the influences of family
structure were to be weaker than cultural influences. However, this premise still
does not explain the significant stepparent effect for black adolescents.
The third reason for this finding, and the most plausible explanation, may be
that extended family members are providing help to the family members of single
parent households. According to Hines & Boyd-Franklin (1996), one characteristic
of African-American families is for the extended family members to provide help in
times of need. This help may mitigate some of the stresses of living in a single
parent household for black adolescents.
This explanation may also be the reason why black adolescents who reside
with a stepparent are at a higher risk for sexual activity compared to those who reside
151
with two biological parents. Although the stepparent household environment may be
as turbulent as those in divorced/separated households, extended family members
may not help the family members of stepparent households as much as those of
single parent households. It could be that extended family members perceive family
members of stepparent households do not need as much external help from other
family members in order for the household to function normally since they are seen
to be in an advantaged context (Moore, 2001). In another words, the family
members of stepparent households may have to rely on their own resources to solve
family problems as they encounter them.
There were racial differences regarding the best reasons for the single parent
household effect. Peer influence was the best reason for the unwedded mother
household effect for white adolescents, which suggests a strong influence by friends
and peers. A probable reason for this finding is that unwedded mothers are more
likely than divorced/separated mothers to live in poorer neighborhoods. In addition,
these households may have been in these neighborhoods for a longer period.
Consequently, these adolescents may have grown up in this type of neighborhood
during their whole life which would indicate that they have greater exposure to
values which encourages deviant behaviors since poorer neighborhoods are
associated with antisocial values (Wilson, 1987; Ransford, 1994). If this were the
situation, white adolescents who reside with unwedded mothers may be more likely
to have friendships with peers who engage in deviant activities.
All of the explanations except for family support at least partially explained
the divorced/separated mother household effect for white adolescents. Initially, this
152
finding would seem to suggest that the divorced/separated mother household effect
has multiple dimensions. However, the best reason for this effect is parental
supervision which would suggests that, for white adolescents, the divorced/separated
mother is primarily due to either mothers having lower levels of interaction with her
children or a weakened parental authority.
Having a mother who is relatively tolerant about adolescent sexual activity
was the best reason for both the unwedded mother and divorced/separated mother
household effect for Hispanic adolescents. This finding seems to confirm the thesis
by Thornton (1991) that unwedded mothers are more likely to be more liberal in
their sexual views compared to the mothers of two biological parent households. It
may be possible that familism (Davis & Harris, 1982; Padilla & O’Grady, 1997) is
weaker in a single parent household than in two parent households for Hispanic
families.
Racial differences were also found in this investigation regarding the best
reason for the stepparent household effect. Family transition was the best reason for
this relationship among white adolescents. This effect is probably due to family
members feeling the stresses that accompany marital disruption and subsequent
marital transitions through parental divorce, remarriage, and redivorce (Wu &
Martinson, 1993; Capaldi, Crosby, & Stoolmiller, 1996). These disruptions may
force family members into a period of trying to adjust to the new household, which
contributes to a turbulent family environment and affects the interaction quality
between parents and their children (Wu & Martinson, 1993). Children who
experience these disruptions may either engage in impulsive, rebellious, and
153
aggressive actions or spend less time at home (Wallerstein & Kelly, 1980;
Wallerstein & Blakeslee, 1989).
Parental supervision was the best explanation for the stepparent effect for
black adolescents. Although it may seem that black adolescents who reside with a
stepparent may be relatively more advantaged than those who reside with a single
parent (Moore, 2001), it could still be difficult for them to adjust to the new family
environment because the stepparent may be seen as a “relative stranger” (Beer,
1988). This difficulty could lead to considerable conflict in the household and the
stepparent feeling that their role as a parent is ambiguous during attempts to
discipline their stepchildren (Cherlin, 1978; Amato, 1987; Furstenberg, 1987).
Stepparents may be less strict when it comes to disciplining their stepchildren in
order to maintain peace within the household.
Even though none of the explanations was able to partially mediate the
stepparent effect for Hispanic adolescents, peer influence seemed to be the best
reason for this effect. Unlike black adolescents, the finding suggests that Hispanic
adolescents who reside in these households may handle their difficulty by spending
less time at home (Wallerstein & Kelly, 1980; Wallerstein & Blakeslee, 1989) and
more time with their friends (Coleman, 1981).
Family and Peer Influence Variables Discussion
Family transition was proposed to be positively associated with sexual
activity for all adolescents regardless of their race. This hypothesis was partially
154
supported by the analyses where this effect is positively associated with sexual
activity only for white and black adolescents. These findings do not support the
study by Wu & Thomson (2001), who found that family transition was not
associated with adolescent sexual activity for black adolescents.
The parent-child interaction variables (mother-child activities, parental
strictness, and parental meal presence) were proposed to be negatively associated
with sexual activity while parent-child discussions were proposed to be positively
associated with sexual activity for all adolescents regardless of race. These
hypotheses were supported by the analyses. Parental meal presence and parental
strictness effects are both positively associated with sexual activity only for white
adolescents. However, these effects are not significantly different for blacks and
Hispanics adolescents even though the effects of these variables on adolescent sexual
activity for these two racial groups are not statistically significant. In addition, these
analyses did not fully support the hypothesis that parental monitoring variables
would not have an association sexual activity for all three racial groups. Instead,
mother’s night presence was negatively associated with a sexual activity for white
adolescents.
This investigation proposed that mother’s religiosity and mother’s
disapproval of adolescent sexual activity would be negatively associated with sexual
activity for all adolescents regardless of race. However, mother’s religiosity was not
negatively associated with sexual activity for black adolescents even though these
mothers are more religious compared to mothers of white adolescents. A possible
explanation for this finding is that these mothers may not be using religious values as
155
a guide to discourage her children from engaging in sexual activity. Another reason
may be that the values of the neighborhood, which could be relatively liberal toward
sexuality (Staples, 1978), are much stronger than the mother’s attitudes in
influencing her children’s decision to engage in sexual activity.
Mother’s disapproval of adolescent sexual activity was much stronger for
whites than for both black and Hispanic adolescents. For black adolescents, it may
be that their mothers are relatively more tolerant of adolescent sexual activity
compared to mothers of white adolescents. However, in the Hispanic adolescents
situation, it could be possible that parental disapproval of adolescent sexual activity
may differ by gender because of the emphasis on distinct gender roles found in the
Hispanic culture (Staples & Mirande, 1989). Hispanic mothers may be more likely
to be approving, or at accepting, of her son’s sexual behavior while disapproving her
daughter’s sexual behavior. Yet, the result of this premise is unknown because this
analysis combined both the Latino’s and Latina’s mother responses to this question.
Analyzing Latinos separately from Latinas would provide a better insight as to
whether there are distinct attitudes regarding sexuality within a gender context for
Hispanic adolescents.
The study supported the hypothesis that family support is negatively
associated with sexual activity for white, black, and Hispanic adolescents. These
findings are also congruent with studies that found family support is negatively
associated with sexual activity for white (Lauritsen, 1994; Whitbeck et al., 1992),
black (Ramirez-Valles, Zimmerman, & Juarez, 2002; Moore & Chase-Lansdale,
2001), and Hispanic (Miller, Forehand, & Kotchick, 1999) adolescents.
156
Meanwhile, this investigation supported the hypothesis that contact with
peers engaging in deviant activities would be positively associated with sexual
activity for white, black, and Hispanic adolescents. However, peer influence seems
to have a stronger relationship with sexual activity for white than for black
adolescents. This finding is probably due to white adolescents having more
acquaintance with friends or peers who engage in deviant activities than black
adolescents. This finding is also supported by studies that found acquaintance with
deviant peers to be positively associated with sexual activity for white (Whitbeck,
Simons, & Goldberg, 1996), black (Furstenberg et al., 1987), and Hispanic
(Christopher, Johnson, & Roosa, 1993) adolescents.
Relative Importance of the Five Explanations Discussion
Peer influence was the most parsimonious adolescent sexual activity
explanation for all the racial subgroups. However, the most parsimonious family
influence explanation varied by the adolescent’s race. The most parsimonious
family influence explanation for white and Hispanic adolescents was mother’s
attitudes. This finding seems to suggest that these mothers have a huge influence on
their children’s decision to engage in sexual activity through the transmission of
sexual values from the parent to the child.
Family support was the most parsimonious family influence explanation for
black adolescents. However, this result may be somewhat controversial because
three of the four parent-child interaction variables were significant while none of the
157
parental monitoring variables were significant for adolescent sexual activity. If the
parental monitoring variables were to be excluded from the parental supervision
regression model, it may be possible that the parent-child interaction would be the
most parsimonious adolescent sexual activity explanation.
Nevertheless, finding that parental support was the most parsimonious family
explanation for black adolescents is probably due to the increased likelihood that
these adolescents reside in non-traditional households. These non-traditional
households may need help from extended family members in order for it to function
normally. Since extended family members are usually present during times of need
for black households (Hines & Boyd-Franklin, 1996), it may be that black
adolescents would view them as being part of the family. Consequently, the
definition of family support may be broader for adolescents who consider their
extended family members as being part of the immediate family.
Family transitions, parental attitude, and parental supervision were the least
parsimonious adolescent sexual activity explanations for white, black, and Hispanic
adolescents respectively according to the BIC statistic. The fact that family
transition is the least parsimonious adolescent sexual activity explanation for white
adolescents may be due to other measures explaining this effect. Parental attitude
was the least parsimonious adolescent sexual activity explanation for blacks,
probably because of the relatively more liberal sexual values of the black culture
(Staples, 1978) negating the conservative bias of several parental attitude variables.
Finally, parental supervision being the least parsimonious adolescent sexual activity
explanation for Hispanics adolescents is probably due to the smaller number of
158
respondents for this racial subgroup compared to other subgroups. The smaller
sample size would have resulted in fewer variables having a significant association
with adolescent sexual activity.
Chapter Summary
The findings of this investigation revealed some racial difference regarding
the relationship between family structure and adolescent sexual activity for three
non-traditional households. First, black adolescents who reside in single parent
households were no more likely than those who reside with two biological parents to
engage in sexual activity. Several reasons were proposed for this finding such as the
institutionalization of the non-traditional households in the black culture, cultural
influences being stronger than the family influences, and the most plausible one
which is single parent households getting help from extended family members.
Second, racial differences were found for the best reasons regarding the
various non-traditional household effects. The unwedded mother household effect
was best explained by peer influence for white adolescents while mother’s relatively
tolerant attitudes in regards to adolescent sexual activity best explained this effect for
Hispanic adolescents. Lower levels of parent-child interaction or a weakened
parental authority best explained the divorced/separated mother household effect for
white adolescents while mother’s relatively tolerant attitudes in regards to adolescent
sexual activity best explained this effect for Hispanic adolescents. The stepparent
household effect was best explained by family transitions for white and to some
159
extent Hispanic adolescents. However, lower levels of parent-child interaction or a
weakened parental authority best explained the stepparent household effect for black
adolescents.
Several racial differences were found regarding the relationship between
family influences and adolescent sexual activity. Some of these differences were
parental night presence for white adolescents, mother’s religiosity and mother’s
disapproval of adolescent sex for black adolescents, and family transition and
mother’s disapproval of adolescent sexual activity for Hispanic adolescents.
Mother’s night presence was associated with prevention of adolescent sexual activity
for white but not for black or Hispanic adolescents. In addition, mother’s religiosity
and her disapproval of adolescent sexual activity were not associated with adolescent
sexual activity for black adolescents. This finding is possibly due to the
pervasiveness of tolerant sexual attitudes found in the black culture. Furthermore,
family transition was negatively associated with sexual activity while there is no
association between mother’s disapproval of adolescent sexual activity and sexual
activity for Hispanic adolescents.
Peer influence was the most parsimonious adolescent sexual activity
explanation for white, black, and Hispanic adolescents. However, the most
parsimonious family influence explanation differed by race. Mother’s attitude was
the most parsimonious family influence explanation for white and Hispanic
adolescents. This finding suggests that these mothers have a huge influence in their
children’s decision to engage in sexual activity through the transmission of sexual
values from the parent to the child. In addition, family support was the most
160
parsimonious family influence explanation for black adolescents, which may be due
to black family members being more likely to receive help from extended family
members since they are more likely to live in single parent households.
Consequently, black adolescents may broaden the definition of what constitutes
family support since they may consider these extended family members as part of the
household.
The least parsimonious sexuality explanation also differed by the
respondent’s race. Family transitions, parental attitude, and parental supervision
were the least parsimonious adolescent sexual activity explanation for white, black,
and Hispanic adolescents respectively. The fact that family transition is the least
parsimonious adolescent sexual activity explanation for white adolescents is
probably due to other measures explaining the effects of this measure. The finding
that parental attitude is the least parsimonious adolescent sexual activity explanation
for black adolescents is probably due to the tolerant attitudes of the black culture
negating the conservative bias of several parental attitude variables. Finally, a
reason why parental supervision was the least parsimonious adolescent sexual
activity explanation for Hispanic adolescents is probably due to a small number of
respondents in this racial subgroup, which resulted in fewer variables having a
significant association with sexual activity.
The results of this investigation revealed that a conflict measure, namely peer
influence, was a better adolescent sexual activity explanation than any of the
functional measures for all racial subgroups. In contrast, there was no consistent
theme as to which perspective predominately best explained the non-traditional
161
household effect when analyzing by adolescent’s race. These findings imply that
peers have a strong influence on whether a child decides to engage in sexual activity
while the way in which parents socialize their children has a mixed effect in
explaining the family structure effect depending upon the adolescent’s race.
Despite the interesting findings, the weakness of this investigation was that
one could not determine if the racial differences regarding the non-traditional
household and adolescent sexual activity relationship differ by gender. In another
words, this analysis did not determine whether there were race differences within the
gender subgroups regarding the family influence explanations and adolescent sexual
activity relationship. The next chapter addresses this deficiency by analyzing
adolescents within race-gender subgroup categories.
162
Chapter 8: Race-Gender Context Analyses
The main objective for this chapter is to provide further insight into the
family structure and adolescent sexual activity relationship for the three non-
traditional households within six adolescent race-gender subgroups. This chapter
includes a t-test for difference between slopes analysis, as explained by Jaccard,
Turrisi, & Wan (1990), in order to determine whether there are significant race-
gender differences for relationships between the five explanations and adolescent
sexual activity
3
. The chapter concludes with a discussion of the findings.
The results from this investigation should be taken with some caution
because the number of adolescent respondents that are analyzed within each race-
gender context is much smaller than those that are analyzed within gender or race
categories. This is an important issue because the small sample size may lead to
results that are not statistically valid. The fewest number of respondents that are
necessary for the analysis to have some statistical validity is 150 (Cohen, 2001;
Newton & Rudestam, 1999). Preliminary analyses determined that the number of
male and female Asian adolescents do not meet this criteria. Thus, the analysis for
this investigation will only focus on white, black, and Hispanic adolescents of both
genders. However, this investigation will present, but not discuss, the results of the
Asian adolescents to provide a complete the examination of the data.
3
The gender difference analyses are only between adolescents of similar race. The
race difference analyses are only between adolescents of similar gender with
white adolescents as the comparison group.
163
Hypotheses
Although the results from the previous investigation have revealed some
differences for the five explanations and adolescent sexual activity relationship
within racial categories, the hypotheses for this investigation will be similar to the
ones were proposed for the previous two investigations. In a manner similar to the
other investigations, there will also be a discussion for any significant differences
regarding the relationship between the five explanations and adolescent sexual
activity for the six race-gender subgroup categories.
The hypotheses for this investigation are as follows:
1. Family transition, mother-child discussions, mother-child sex discussions,
and acquaintance with deviant peers will be positively associated with sexual
activity for all adolescents regardless of their race-gender subgroup category.
2. Mother-child activities, parental strictness, parental meal presence, mother’s
religiosity, mother disapproval of adolescent sex, and family support will be
negatively associated with sexual activity for all adolescents regardless of
their race-gender subgroup category.
3. None of the parental monitoring variables (mother’s full time employment
and mother’s morning, afternoon, and evening presence) will be associated
with sexual activity for all adolescents regardless of their race-gender
subgroup category.
164
Descriptive Results
Table 8-1 presents the sexual activity, demographic, and the variables that
constitute the five explanations mean scores for white adolescents. The comparison
between gender shows that white males are more likely than white females to be
older during the time of the interview. White males are less likely than white
females to experience mother-child activities and to have both general and sex
discussions with their mothers. However, white males are more likely than white
females to have parents who are stricter with them.
Table 8-2 presents the sexual activity, demographic, and the variables that
constitute the five explanations mean scores for black adolescents. Black males are
more likely than black females to engage in sexual activity, to be older during the
time of the interview, and to live with a stepparent. Compared to black females,
black males have fewer general and sex discussions with their mothers. However,
black males are more likely than black females to report greater family support and
to have more acquaintances with friends who engage in deviant activities.
Table 8-3 presents the sexual activity, demographic, and the variables that
constitute the five explanations mean scores for both Hispanic and Asian
adolescents. There will not be discussion of the results for the Asian adolescents.
Latinos are more likely than Latinas to be sexually experienced. They are also less
likely than Latinas to experience mother-child discussions. However, Latinos more
likely than Latinas to report greater family support.
165
Table 8-1: Sexual Activity, Demographic, Family Influence, and Peer Influence
Variables Mean Scores For White Adolescents
White Males White Females Gender Difference
Sexually Experienced 0.33 0.32
Child's Age 16.00 15.79 **
Two Biological Parents 0.68 0.67
Unwedded Mothers 0.01 0.02
Divorced/Separated Mothers 0.17 0.17
Stepparents 0.13 0.14
Mother's Education 2.87 2.90
Household Income 54.01 55.99
Missing Income 0.10 0.09
Family Transitions 1.30 1.32
Mother-Child Activities 1.51 1.80 ***
Mother-Child Discussions 2.16 2.69 ***
Parent Strictness 1.17 0.98 ***
Parent Meal Presence 5.08 4.95
Mother's Employment Status 0.55 0.56
Mother's Morning Presence 4.22 4.24
Mother's Afternoon Presence 3.19 3.25
Mother's Evening Presence 4.65 4.65
Mother's Religiosity 6.04 5.98
Mother-Child Sex Discussions 2.86 3.06 ***
Mother Disapprove Sex 4.46 4.47
Family's Support 16.27 16.20
Deviant Peers 2.65 2.55
N 1242 1372
Notes: * = p < .05; ** = p < .01; *** = p < .001
166
Table 8-2: Sexual Activity, Demographic, Family Influence, and Peer Influence
Variables Mean Scores For Black Adolescents
Black Males Black Females Gender Difference
Sexually Experienced 0.60 0.43 ***
Child's Age 16.08 15.80 *
Two Biological Parents 0.34 0.40
Unwedded Mothers 0.17 0.19
Divorced/Separated Mothers 0.32 0.29
Stepparents 0.18 0.12 *
Mother's Education 2.89 2.83
Household Income 35.90 36.02
Missing Income 0.10 0.12
Family Transitions 1.31 1.23
Mother-Child Activities 1.69 1.78
Mother-Child Discussions 2.32 2.57 **
Parent Strictness 1.36 1.34
Parent Meal Presence 3.96 3.96
Mother's Employment Status 0.70 0.71
Mother's Morning Presence 4.09 4.17
Mother's Afternoon Presence 3.32 3.29
Mother's Evening Presence 4.67 4.66
Mother's Religiosity 6.99 6.93
Mother-Child Sex Discussions 3.20 3.40 **
Mother Disapprove Sex 4.19 4.32
Family's Support 16.60 16.21 *
Deviant Peers 2.41 1.70 **
N 396 476
Notes: * = p < .05; ** = p < .01; *** = p < .001
167
Table 8-3: Sexual Status, Demographic, Family Influence, and Peer Influence
Variables Mean Scores For Hispanic and Asian Adolescents
Latinos Latinas Gender Difference
Sexually Experienced 0.40 0.29 *
Child's Age 16.01 15.85
Two Biological Parents 0.55 0.52
Unwedded Mothers 0.08 0.07
Divorced/Separated Mothers 0.20 0.25
Stepparents 0.17 0.16
Mother's Education 2.13 1.98
Household Income 35.88 32.46
Missing Income 0.14 0.13
Family Transitions 1.35 1.44
Mother-Child Activities 1.43 1.64
Mother-Child Discussions 2.12 2.56 **
Parent Strictness 1.37 1.37
Parent Meal Presence 4.82 4.54
Mother's Employment Status 0.52 0.51
Mother's Morning Presence 4.18 4.22
Mother's Afternoon Presence 3.42 3.59
Mother's Evening Presence 4.74 4.80
Mother's Religiosity 6.56 6.31
Mother-Child Sex Discussions 2.77 2.91
Mother Disapprove Sex 4.21 4.13
Family's Support 16.44 15.84 *
Deviant Peers 2.68 2.28
N 194 232
Notes: * = p < .05; ** = p < .01; *** = p < .001
168
Table 8-3: Continued
Asian Males Asian Females Gender Difference
Sexually Experienced 0.20 0.22
Child's Age 15.76 16.18
Two Biological Parents 0.78 0.82
Unwedded Mothers 0.02 0.02
Divorced/Separated Mothers 0.07 0.08
Stepparents 0.13 0.08
Mother's Education 3.19 3.44
Household Income 47.16 64.35
Missing Income 0.17 0.14
Family Transitions 1.19 1.10
Mother-Child Activities 1.50 2.00 **
Mother-Child Discussions 1.78 2.50 **
Parent Strictness 1.41 0.94
Parent Meal Presence 4.20 5.20 *
Mother's Employment Status 0.70 0.66
Mother's Morning Presence 3.93 3.76
Mother's Afternoon Presence 3.15 3.32
Mother's Evening Presence 4.56 4.62
Mother's Religiosity 6.52 6.24
Mother-Child Sex Discussions 2.24 2.60
Mother Disapprove Sex 4.44 4.66
Family's Support 15.85 15.74
Deviant Peers 1.83 2.30
N 54 50
Notes: * = p < .05; ** = p < .01; *** = p < .001
169
Table 8-4 presents the mean scores for the sexual activity, the demographic
variables, and the variables that constitute the five explanations among male
adolescents. The comparison between race shows that black males are more likely
than white males to be sexually experienced. Among the demographic variables,
black males are more likely than white males to reside in non-traditional households
but are less likely to reside with two biological parents. In addition, the black male
households are relatively poorer than white male households.
Among the family and peer influence measures, black males are more likely
than white males to experience mother-child activities, to have stricter parents, and
to report support from family members. However, black males are less likely than
white males to experience parental presence during meals. Mothers of black males
are more likely than mothers of white males to be employed full time, to be religious,
to discuss sex with their children, and to be tolerant of adolescent sexual activity.
The risk for sexual activity is similar between Latinos and white males.
Latinos are less likely than white males to live with two-biological parents and
unwedded mothers. In addition, households that Latinos reside in are significantly
poorer compared to the households of white males. Among the five measures,
Latino mothers are more likely than mothers of white males to be present at home
during afternoons. Mothers of white males are more likely than Latino mothers to
achieve higher educational levels and be disapproving of adolescent sexual activity.
However, Latino mothers are more likely than mothers of white males to be
religious.
170
Table 8-4: Mean Scores for the Sexual Activity, Demographic, Family Influence, and Peer Influence Variables Among Male
Adolescents
White Males Black Males Latinos Asian Males
Sexually Experienced 0.33 0.60 *** 0.40 0.20 *
Child's Age 16.00 16.08 16.01 15.76
Two Biological Parents 0.68 0.34 *** 0.55 ** 0.78
Unwedded Mothers 0.01 0.17 *** 0.08 ** 0.02
Divorced/Separated Mothers 0.17 0.32 *** 0.20 0.07 *
Stepparents 0.13 0.18 * 0.17 0.13
Mother's Education 2.87 2.89 2.13 *** 3.19
Household Income 54.01 35.90 *** 35.88 *** 47.16 *
Missing Income 0.10 0.10 0.14 0.17
Family Transitions 1.30 1.31 1.35 1.19
Mother-Child Activities 1.51 1.69 ** 1.43 1.50
Mother-Child Discussions 2.16 2.32 2.12 1.78 *
Parent Strictness 1.17 1.36 * 1.37 1.41
Parent Meal Presence 5.08 3.96 *** 4.82 4.20 *
Mother's Employment Status 0.55 0.70 *** 0.52 0.70 *
Mother's Morning Presence 4.22 4.09 4.18 3.93
Mother's Afternoon Presence 3.19 3.32 3.42 3.15
Mother's Evening Presence 4.65 4.67 4.74 4.56
N 1242 396 194 54
Notes: * = p < .05; ** = p < .01; *** = p < .001
171
Table 8-4: Continued
White Males Black Males Latinos Asian Males
Mother's Religiosity 6.04 6.99 *** 6.56 *** 6.52
Mother-Child Sex Discussions 2.86 3.20 *** 2.77 2.24 ***
Mother Disapprove Sex 4.46 4.19 *** 4.21 ** 4.44
Family's Support 16.27 16.60 * 16.44 15.85
Deviant Peers 2.65 2.41 2.68 1.83 *
N 1242 396 194 54
Notes: * = p < .05; ** = p < .01; *** = p < .001
172
Table 8-5 presents the mean scores for the sexual activity, the demographic
variables, and the variables that constitute the five explanations among female
adolescents. Black females are more likely than white females to be sexually
experienced. Among the demographic variables, black females are more likely than
white females to reside in a single parent household. However, black females are
less likely than white females to reside with two-biological parents. The black
female households are significantly poorer compared to white female households.
Among the family and peer influence measures, black females are more
likely than white female to experience stricter parents. However, black females are
less likely than white females to have meals with at least one parent and to have
acquaintances with friends or peers who engage in deviant activities. Mothers of
black females are less likely than mothers of white females to have multiple
marriages or serious relationships in the past 18 years. Conversely, mothers of black
females are more likely than mothers of white females to be employed full-time, to
be religious, to be tolerant of adolescent sexual activity, and to discuss sex with her
children.
The risk for sexual activity is not significantly different between Latinas and
white females. Latinas are less likely than white females to live with two-biological
parents but are more likely to live in single parent households. In addition, Latina
households are significantly poorer compared to white female households.
Furthermore, mothers of Latinas achieve lower levels of formal education compared
to mothers of white females.
173
Table 8-5: Mean Scores for the Sexual Activity, Demographic, Family Influence, and Peer Influence Variables Among Females
Adolescents
White Females Black Females Latinas Asian Females
Sexually Experienced 0.32 0.43 *** 0.29 0.22
Child's Age 15.79 15.80 15.85 16.18
Two Biological Parents 0.67 0.40 *** 0.52 *** 0.82 *
Unwedded Mothers 0.02 0.19 *** 0.07 ** 0.02
Divorced/Separated Mothers 0.17 0.29 *** 0.25 ** 0.08 *
Stepparents 0.14 0.12 0.16 0.08
Mother's Education 2.91 2.83 1.99 *** 3.44 **
Household Income 55.99 36.02 *** 32.46 *** 64.35
Missing Income 0.09 0.12 0.13 0.14
Family Transitions 1.32 1.23 * 1.44 1.10 ***
Mother-Child Activities 1.80 1.78 1.64 * 2.00
Mother-Child Discussions 2.69 2.57 2.56 2.50
Parent Strictness 0.98 1.34 *** 1.37 *** 0.94
Parent Meal Presence 4.95 3.96 *** 4.54 * 5.20
Mother's Employment Status 0.56 0.71 *** 0.51 0.66
Mother's Morning Presence 4.24 4.17 4.22 3.76 *
Mother's Afternoon Presence 3.25 3.29 3.59 ** 3.32
Mother's Evening Presence 4.65 4.66 4.80 ** 4.62
N 1372 476 232 50
Notes: * = p < .05; ** = p < .01; *** = p < .001
174
Table 8-5: Continued
White Females Black Females Latinas Asian Females
Mother's Religiosity 5.98 6.93 *** 6.31 ** 6.24
Mother-Child Sex Discussions 3.06 3.40 *** 2.91 * 2.60 **
Mother Disapprove Sex 4.47 4.32 * 4.13 *** 4.66
Family's Support 16.20 16.21 15.84 15.74
Deviant Peers 2.55 1.70 *** 2.28 2.30
N 1372 476 232 50
Notes: * = p < .05; ** = p < .01; *** = p < .001
175
Among the family and peer influence measures, Latinas are more likely than
white females to engage in activities with their mothers, to have stricter parents, and
to have mothers present at home during afternoons and evenings. Latinas are less
likely than white females to experience general discussions with her mothers, to have
meals with at least one parent present, and to report feeling support from family
members. In addition, mothers of Latinas are more likely than mothers of white
adolescents to have multiple marriages or serious relationships during the past 18
years, to be religious, to be tolerant of adolescent sexual activity, and to discuss sex
topics with their daughters.
Test of Hypotheses Results
Table 8-6 displays the regression of adolescent sexual activity on the
demographic variables within the six race-gender categories. Both the unwedded
mother and divorced/separated mother households are positively associated with
sexual activity for white males and females. White females who reside with a
stepparent have an increased risk for sexual activity. Mother’s education is
negatively associated with sexual activity for white males. There is a significant
gender difference among white adolescents for the stepparent household effect where
this effect is stronger for females than for males.
Stepparent households are positively associated with sexual activity for black
adolescents of both genders. Hispanic adolescents of both genders are positively
associated with sexual activity for those who reside with divorced/separated mothers.
176
Table 8-6: Logistic Regression of Adolescent Sexual Activity on the Demographic Variables Within Race-Gender Categories
White Males Black Males Latinos Asian Males
Beta Exp (B) Beta Exp (B) Beta Exp (B) Beta Exp (B)
Intercept -8.98 *** -4.90 *** -9.54 *** -11.07 *
Child's Age 0.58 *** 1.79 0.34 *** 1.40 0.58 *** 1.78 0.49 * 1.63
Unwedded Mothers
a
1.50 * 4.48 0.63 1.87 0.22 1.25 -7.47 0.00
Divorced/Separated Mothers
a
0.69 *** 1.98 0.39 1.47 1.11 * 3.03 0.30 1.35
Stepparents
a
0.31 (g) 1.37 0.71 * 2.03 0.58 (g) 1.79 -1.21 0.30
Mother's Education -0.27 *** 0.76 -0.08 0.92 0.07 1.07 -0.42 0.66
Household Income (ln) -0.18 0.84 -0.07 0.94 -0.20 0.82 0.95 2.60
Missing Household Income 0.18 1.19 -0.04 0.96 -0.31 0.73 -8.36 0.00
-2 LL 1317.59 494.97 215.69 42.09
Chi-Square 257.91 *** 39.33 *** 45.76 *** 12.51
Nagelkerke R
2
0.26 0.13 0.28 0.33
N 1242 396 194 54
BIC -208.04 2.54 -3.89 15.41
Notes: * = p < .05; ** = p < .01; *** = p < .001; (r) = racial difference; (g) = gender difference;
a
= two biological parents as the
reference group
177
Table 8-6: Continued
White Females Black Females Latinas Asian Females
Beta Exp (B) Beta Exp (B) Beta Exp (B) Beta Exp (B)
Intercept -12.32 *** -7.15 *** -9.65 *** -4.63
Child's Age 0.75 *** 2.11 0.48 *** 1.61 0.48 *** 1.61 0.26 1.30
Unwedded Mothers
a
1.68 *** 5.39 0.11 (r) 1.11 1.61 * 4.99 -8.80 0.00
Divorced/Separated Mothers
a
0.96 *** 2.62 0.26 (r) 1.29 1.04 * 2.82 -7.74 0.00
Stepparents
a
0.82 *** 2.27 0.70 * 2.02 1.71 *** (r) 5.51 0.10 1.10
Mother's Education -0.09 0.91 -0.10 0.90 -0.11 0.90 -0.14 0.87
Household Income -0.16 0.85 -0.17 0.84 0.17 1.18 -0.07 0.93
Missing Household Income -0.05 0.95 -0.17 0.84 0.03 1.04 -0.54 0.58
-2 LL 1332.75 567.70 234.26 47.93
Chi-Square 391.81 *** 84.07 *** 44.65 *** 4.76
Nagelkerke R
2
0.35 0.22 0.25 0.14
N 1372 476 232 50
BIC -341.24 -40.91 -6.52 22.62
Notes: * = p < .05; ** = p < .01; *** = p < .001; (r) = racial difference; (g) = gender difference;
a
= two biological parents as the
reference group
178
Latinas are positively associated with sexual activity for those who reside with either
unwedded mothers or stepparents. There is a significant gender difference among
Hispanic adolescents for the stepparent household effect where this effect is stronger
for Latinas than for Latinos.
There are no significant racial differences between white males, black males,
and Latinos regarding the non-traditional household effect. However, there is a
significant racial difference between white and black females for both the unwedded
mother and divorced/separated mother household effects. In addition, the stepparent
effect is much stronger for Latinas than for white females.
Table 8-7 displays the adolescent sexual activity regressions on the family
transition variable within the six race-gender categories. Family transition is
positively associated with sexual activity for white adolescents but this effect is not
significantly different between 1) white and black males and 2) white and black
females. However, family transition is negatively associated with sexual activity for
Latinas. In addition, this effect is significantly different from white females and
Latinos.
Table 8-8 displays the adolescent sexual activity regressions on the parental
supervision variables within the six race-gender categories. An initial overview of
the data seems to show numerous differences between subgroups regarding the
relationship between the parental supervision variables and adolescent sexual
activity. However, there are only a few true race and gender differences for this
relationship. One gender difference is found for parental presence during meals for
white adolescents where this measure is negatively associated with sexual activity
179
Table 8-7: Logistic Regression of Adolescent Sexual Activity on the Family Transition Variable Within Race-Gender Categories
White Males Black Males Latinos Asian Males
Beta Exp (B) Beta Exp (B) Beta Exp (B) Beta Exp (B)
Intercept -9.67 *** -5.22 *** -9.67 *** -10.92 *
Child's Age 0.60 *** 1.82 0.34 *** 1.41 0.58 *** 1.79 0.48 1.62
Unwedded Mothers
a
1.28 * 3.60 0.64 1.89 0.20 1.22 -7.49 0.00
Divorced/Separated Mothers
a
0.38 1.46 0.32 1.38 1.06 * 2.88 0.33 1.40
Stepparents
a
-0.17 0.84 0.58 1.79 0.50 1.65 -1.17 0.31
Mother's Education -0.26 *** 0.77 -0.08 0.92 0.07 1.07 -0.42 0.66
Household Income (ln) -0.19 0.83 -0.05 0.95 -0.19 0.83 0.95 2.58
Missing Household Income 0.18 1.19 -0.03 0.98 -0.31 0.74 -8.35 0.00
Family Transitions 0.43 ** 1.54 0.17 1.18 0.08 1.08 -0.06 0.95
-2 LL 1306.67 493.36 215.62 42.08
Chi-Square 268.83 *** 40.93 *** 45.84 *** 12.51
Nagelkerke R
2
0.27 0.13 0.28 0.33
N 1242 396 194 54
BIC -211.83 6.92 -3.70 19.40
Notes: * = p < .05; ** = p < .01; *** = p < .001; (r) = racial difference; (g) = gender difference;
a
= two biological parents as the
reference group
180
Table 8-7: Continued
White Females Black Females Latinas Asian Females
Beta Exp (B) Beta Exp (B) Beta Exp (B) Beta Exp (B)
Intercept -12.90 *** -7.55 *** -9.17 *** 3.33
Child's Age 0.76 *** 2.14 0.49 *** 1.63 0.48 *** 1.61 0.22 1.24
Unwedded Mothers
a
1.51 ** 4.54 0.06 1.06 2.29 ** 9.91 -1.43 0.24
Divorced/Separated Mothers
a
0.76 *** 2.14 0.15 1.17 1.32 ** 3.76 -6.62 0.00
Stepparents
a
0.48 * 1.61 0.50 1.65 2.18 *** 8.83 0.75 2.11
Mother's Education -0.07 0.93 -0.10 0.90 -0.08 0.93 -0.09 0.91
Household Income -0.15 0.86 -0.18 0.84 0.19 1.20 -0.10 0.90
Missing Household Income -0.04 0.96 -0.11 0.89 -0.04 0.96 -0.35 0.70
Family Transitions 0.32 ** 1.37 0.29 1.34 -0.55 *(r)(g) 0.58 -7.26 0.00
-2 LL 1324.94 563.81 227.98 47.14
Chi-Square 399.63 *** 87.97 *** 50.92 *** 5.55
Nagelkerke R
2
0.35 0.23 0.28 0.16
N 1372 476 232 50
BIC -341.84 -38.65 -7.35 25.75
Notes: * = p < .05; ** = p < .01; *** = p < .001; (r) = racial difference; (g) = gender difference;
a
= two biological parents as the
reference group
181
Table 8-8: Logistic Regression of Adolescent Sexual Activity on the Parental
Supervision Variables Within Race-Gender Categories
White Males Black Males
Beta Exp (B) Beta Exp (B)
Intercept -5.39 *** -3.23
Child's Age 0.48 *** 1.61 0.26 *** 1.29
Unwedded Mothers
a
1.25 * 3.50 0.38 1.46
Divorced/Separated Mothers
a
0.46 * 1.58 0.09 1.10
Stepparents
a
0.17 1.18 0.48 1.62
Mother's Education -0.29 *** 0.75 -0.12 0.89
Household Income -0.25 * 0.78 -0.25 0.78
Missing Household Income 0.20 1.22 -0.13 0.88
Mother-Child Activities -0.34 *** 0.71 -0.22 0.80
Mother-Child Discussions 0.25 *** 1.29 0.30 *** 1.35
Parent Strictness -0.15 * 0.86 -0.19 * 0.83
Parent Meal Presence -0.04 (g) 0.96 -0.02 0.98
Mother's Employment Status -0.08 0.92 0.51 1.67
Mother's Morning Presence 0.01 1.01 -0.07 0.93
Mother's Afternoon Presence -0.08 0.92 -0.01 0.99
Mother's Evening Presence -0.21 * 0.81 0.11 (r) 1.12
-2 LL 1257.17 467.85
Chi-Square 318.33 *** 66.45 ***
Nagelkerke R
2
0.32 0.21
N 1242 396
BIC -211.46 23.27
Notes: * = p < .05; ** = p < .01; *** = p < .001; ^ = race difference;
a
= two
biological parents as the reference group
182
Table 8-8: Continued
Latinos Asian Males
Beta Exp (B) Beta Exp (B)
Intercept -11.58 *** -9.88
Child's Age 0.58 *** 1.79 0.39 1.47
Unwedded Mothers
a
0.11 1.11 -6.25 0.00
Divorced/Separated Mothers
a
1.04 * 2.82 -0.42 0.65
Stepparents
a
0.65 1.91 -1.43 0.24
Mother's Education 0.09 1.09 -0.15 0.86
Household Income -0.18 0.84 1.09 2.96
Missing Household Income -0.37 0.69 -9.61 0.00
Mother-Child Activities -0.07 0.94 -0.36 0.70
Mother-Child Discussions 0.13 1.14 0.25 1.28
Parent Strictness 0.00 1.00 0.01 1.01
Parent Meal Presence -0.15 0.86 -0.38 0.69
Mother's Employment Status -0.15 0.86 -2.45 0.09
Mother's Morning Presence -0.18 0.83 0.05 1.06
Mother's Afternoon Presence -0.07 0.93 -0.34 0.71
Mother's Evening Presence 0.72 * (r)(g) 2.06 0.63 1.87
-2 LL 203.20 34.32
Chi-Square 58.25 *** 20.27
Nagelkerke R
2
0.35 0.49
N 194 54
BIC 20.77 39.56
Notes: * = p < .05; ** = p < .01; *** = p < .001; ^ = race difference;
a
= two
biological parents as the reference group
183
Table 8-8: Continued
White Females Black Females
Beta Exp (B) Beta Exp (B)
Intercept -8.36 *** -4.59 **
Child's Age 0.63 *** 1.89 0.40 *** 1.50
Unwedded Mothers
a
1.19 * 3.27 0.03 1.03
Divorced/Separated Mothers
a
0.61 ** 1.84 0.00 1.00
Stepparents
a
0.67 ** 1.95 0.59 1.80
Mother's Education -0.11 0.90 -0.12 0.89
Household Income -0.26 * 0.77 -0.16 0.85
Missing Household Income 0.02 1.02 -0.18 0.84
Mother-Child Activities -0.34 *** 0.71 -0.27 ** 0.76
Mother-Child Discussions 0.26 *** 1.29 0.22 ** 1.25
Parent Strictness -0.17 * 0.84 -0.23 ** 0.80
Parent Meal Presence -0.12 *** 0.89 -0.09 * 0.91
Mother's Employment Status 0.14 1.15 -0.15 0.86
Mother's Morning Presence -0.09 0.92 -0.11 0.89
Mother's Afternoon Presence -0.04 0.96 -0.03 0.97
Mother's Evening Presence -0.13 0.88 -0.02 0.98
-2 LL 1246.10 533.57
Chi-Square 478.46 *** 118.21 ***
Nagelkerke R
2
0.41 0.30
N 1372 476
BIC -370.10 -25.73
Notes: * = p < .05; ** = p < .01; *** = p < .001; ^ = race difference;
a
= two
biological parents as the reference group
184
Table 8-8: Continued
Latinas Asian Females
Beta Exp (B) Beta Exp (B)
Intercept -9.30 *** -4.24
Child's Age 0.43 *** 1.54 0.50 1.66
Unwedded Mothers
a
1.75 * 5.73 -8.46 0.00
Divorced/Separated Mothers
a
0.82 2.28 -10.47 0.00
Stepparents
a
1.53 ** 4.60 1.02 2.77
Mother's Education -0.07 0.93 0.23 1.25
Household Income 0.08 1.09 -1.55 0.21
Missing Household Income -0.24 0.78 -1.81 0.16
Mother-Child Activities -0.42 * 0.66 -0.14 0.87
Mother-Child Discussions 0.30 * 1.35 -0.51 0.60
Parent Strictness -0.15 0.86 0.61 1.85
Parent Meal Presence -0.07 0.93 -0.43 0.65
Mother's Employment Status 0.42 1.53 0.93 2.54
Mother's Morning Presence 0.14 1.15 -1.05 * 0.35
Mother's Afternoon Presence 0.07 1.07 -0.15 0.86
Mother's Evening Presence 0.00 1.00 1.45 4.26
-2 LL 217.71 31.62
Chi-Square 61.11 *** 21.07
Nagelkerke R
2
0.33 0.53
N 232 50
BIC 20.59 37.61
Notes: * = p < .05; ** = p < .01; *** = p < .001; ^ = race difference;
a
= two
biological parents as the reference group
185
for females. Another gender difference is found for mother’s night presence for
Hispanic adolescents where this measure is positively associated with sexual activity
for Latinos. Finally, a racial difference is found for mother’s night presence between
white males and both black males and Latinos where this measure is negatively
associated with sexual activity for white males than for males of the other two race-
gender subgroups. Thus, mother’s night presence is positively associated with
sexual activity for Latinos and this effect is significantly different from white males
and Latinas.
Table 8-9 displays the adolescent sexual activity regressions on the parental
attitude variables within the six race-gender categories. There is a race difference
mother’s religiosity between white and black males where this effect is negatively
associated with sexual activity for whites. Another race difference is found for
mother’s disapproval of adolescent sexual activity between white females and both
black females and Latinas where this effect is negatively associated with sexual
activity for white females. Finally, a gender difference is found for mother’s
disapproval of adolescent sexuality between Latinos and Latinas where this effect is
negatively associated with sexual activity for Latinos.
Table 8-10 displays the adolescent sexual activity regressions on the family
support variable within the six race-gender categories. Family support is negatively
associated with sexual activity for white and black adolescents of both genders.
However, there are no significant racial differences for the family support effect
between white and Hispanic adolescents of both genders. Thus, one cannot conclude
186
Table 8-9: Logistic Regression of Adolescent Sexual Activity on the Parental Attitude Variables Within Race-Gender Categories
White Males Black Males Latinos Asian Males
Beta Exp (B) Beta Exp (B) Beta Exp (B) Beta Exp (B)
Intercept -7.84 *** -5.13 -5.43 * -10.70
Child's Age 0.55 *** 1.74 0.30 1.35 0.50 *** 1.65 0.55 1.73
Unwedded Mothers
a
1.12 3.07 0.62 1.86 -0.11 0.89 -11.14 0.00
Divorced/Separated Mothers
a
0.42 * 1.53 0.39 1.47 0.86 2.37 -0.93 0.40
Stepparents
a
0.10 1.10 0.66 1.94 0.47 1.61 -1.00 0.37
Mother's Education -0.22 ** 0.80 -0.08 0.92 0.05 1.05 -0.48 0.62
Household Income -0.19 0.83 -0.04 0.96 -0.21 0.81 1.92 6.82
Missing Household Income 0.10 1.11 -0.08 0.92 -0.15 0.86 -10.12 0.00
Mother's Religiosity -0.12 ** 0.89 0.06 (r) 1.07 -0.29 * 0.75 0.24 1.27
Mother-Child Sex Discussions 0.42 *** 1.52 0.28 1.33 0.29 1.33 -0.73 0.48
Mother Disapprove Sex -0.26 *** 0.77 -0.14 0.87 -0.39 * 0.68 -1.05 0.35
-2 LL 1264.70 487.42 202.43 37.42
Chi-Square 310.80 *** 46.88 *** 59.02 *** 17.17
Nagelkerke R
2
0.31 0.15 0.35 0.43
N 1242 396 194 54
BIC -239.56 12.93 -6.34 22.72
Notes: * = p < .05; ** = p < .01; *** = p < .001; (r) = racial difference; (g) = gender difference;
a
= two biological parents as the
reference group
187
Table 8-9: Continued
White Females Black Females Latinas Asian Females
Beta Exp (B) Beta Exp (B) Beta Exp (B) Beta Exp (B)
Intercept -10.78 *** -7.62 -10.69 *** -5.72
Child's Age 0.74 *** 2.09 0.47 1.60 0.54 *** 1.72 0.36 1.43
Unwedded Mothers
a
1.50 ** 4.50 0.01 1.01 1.16 3.17 -9.38 0.00
Divorced/Separated Mothers
a
0.78 *** 2.19 0.20 1.23 0.79 2.20 -7.27 0.00
Stepparents
a
0.64 ** 1.89 0.66 1.93 1.58 *** 4.85 0.38 1.46
Mother's Education -0.02 0.98 -0.10 0.91 -0.11 0.89 -0.20 0.82
Household Income -0.15 0.86 -0.18 0.84 0.11 1.11 -0.21 0.81
Missing Household Income -0.09 0.91 -0.14 0.87 -0.08 0.92 -0.41 0.67
Mother's Religiosity -0.16 *** 0.85 -0.08 0.93 -0.18 0.84 0.22 1.24
Mother-Child Sex Discussions 0.37 *** 1.45 0.32 1.38 0.36 * 1.44 0.32 1.38
Mother Disapprove Sex -0.39 *** 0.68 0.00 (r) 1.00 0.09 (r)(g) 1.09 -0.46 0.63
-2 LL 1254.44 561.30 225.15 44.61
Chi-Square 470.12 *** 90.48 *** 53.75 *** 8.05
Nagelkerke R
2
0.41 0.25 0.30 0.23
N 1372 476 232 50
BIC -397.88 -28.83 0.72 31.07
Notes: * = p < .05; ** = p < .01; *** = p < .001; (r) = racial difference; (g) = gender difference;
a
= two biological parents as the
reference group
188
Table 8-10: Logistic Regression of Adolescent Sexual Activity on the Family Support Variable Within Race-Gender Categories
White Males Black Males Latinos Asian Males
Beta Exp (B) Beta Exp (B) Beta Exp (B) Beta Exp (B)
Intercept -7.16 *** -2.45 -8.13 ** -9.12
Child's Age 0.57 *** 1.77 0.32 *** 1.37 0.56 *** 1.76 0.44 1.55
Unwedded Mothers
a
1.50 * 4.48 0.64 1.90 0.32 1.38 -7.27 0.00
Divorced/Separated Mothers
a
0.67 *** 1.95 0.33 1.39 1.13 * 3.11 0.07 1.07
Stepparents
a
0.26 1.30 0.63 1.88 0.62 1.86 -1.27 0.28
Mother's Education -0.27 *** 0.76 -0.10 0.91 0.02 1.02 -0.42 0.66
Household Income -0.18 0.84 -0.09 0.92 -0.17 0.84 1.09 2.97
Missing Household Income 0.15 1.17 0.01 1.01 -0.34 0.71 -8.17 0.00
Family Support -0.10 *** 0.90 -0.12 ** 0.89 -0.07 0.93 -0.11 0.90
-2 LL 1302.59 487.57 214.45 41.77
Chi-Square 272.90 *** 46.73 *** 47.00 *** 12.83
Nagelkerke R
2
0.27 0.15 0.29 0.33
N 1242 396 194 54
BIC -215.90 1.12 -4.86 19.08
Notes: * = p < .05; ** = p < .01; *** = p < .001; (r) = racial difference; (g) = gender difference;
a
= two biological parents as the
reference group
189
Table 8-10: Continued
White Females Black Females Latinas Asian Females
Beta Exp (B) Beta Exp (B) Beta Exp (B) Beta Exp (B)
Intercept -9.91 *** -4.23 ** -8.42 *** 1.48
Child's Age 0.76 *** 2.13 0.47 *** 1.61 0.48 *** 1.62 0.29 1.34
Unwedded Mothers
a
1.54 ** 4.67 0.07 1.07 1.59 * 4.89 -7.75 0.00
Divorced/Separated Mothers
a
0.88 *** 2.41 0.21 1.23 0.98 * 2.66 -8.30 0.00
Stepparents
a
0.75 *** 2.12 0.57 1.76 1.63 *** 5.11 -0.04 0.96
Mother's Education -0.08 0.92 -0.11 0.90 -0.09 0.91 -0.04 0.96
Household Income -0.14 0.87 -0.21 0.81 0.19 1.21 -0.32 0.73
Missing Household Income -0.03 0.97 -0.04 0.96 0.04 1.04 0.93 2.54
Family Support -0.16 *** 0.85 -0.17 *** 0.84 -0.09 0.92 -0.40 0.67
-2 LL 1292.70 548.40 231.46 43.89
Chi-Square 431.87 *** 103.34 *** 47.44 *** 8.80
Nagelkerke R
2
0.38 0.26 0.26 0.25
N 1372 476 232 50
BIC -374.08 -54.02 -3.87 22.50
Notes: * = p < .05; ** = p < .01; *** = p < .001; (r) = racial difference; (g) = gender difference;
a
= two biological parents as the
reference group
190
that the relationship between family support and sexual activity is significantly
different between white and Hispanic adolescents.
Table 8-11 displays the adolescent sexual activity regressions on the peer
influence variable within the six race-gender categories. Acquaintance with deviant
peers is positively associated with sexual activity for all adolescents. There is a
gender difference for this measure among white adolescents where this effect is
positively associated with sexual activity for white males. In addition, there is a
racial difference for this measure between white and black males where this effect is
positively associated with sexual activity for whites.
According to the BIC statistic, peer influence is the most parsimonious
adolescent sexual activity explanation for the six race-gender subgroups. When the
peer influence explanation is excluded, parental attitude is most parsimonious family
process explanation for white adolescents of both genders and Latinos. Family
support is the most parsimonious family process explanation for black adolescents of
both genders. Family transition is the most parsimonious family process explanation
for Latinas. However, parental supervision is the least parsimonious adolescent
sexual activity explanation for five of the race-gender subgroups except for white
females. For white females, family transition is the least parsimonious adolescent
sexual activity explanation according to the BIC statistic.
Table 8-12 summarizes the family structure and adolescent sexual activity
relationship for the three non-traditional households within the six race-gender
subgroup categories. The first column displays the non-traditional household
configurations by the adolescent’s race-gender subgroup. The second column
191
Table 8-11: Logistic Regression of Adolescent Sexual Activity on the Peer Influence Variable Within Race-Gender Categories
White Males Black Males Latinos Asian Males
Beta Exp (B) Beta Exp (B) Beta Exp (B) Beta Exp (B)
Intercept -8.38 *** -3.96 ** -8.23 *** -8.32
Child's Age 0.48 *** 1.62 0.24 *** 1.27 0.45 *** 1.56 0.22 1.25
Unwedded Mothers
a
0.95 2.58 0.65 1.92 0.33 1.40 -6.55 0.00
Divorced/Separated Mothers
a
0.51 * 1.67 0.37 1.45 0.85 2.34 -0.99 0.37
Stepparents
a
0.18 1.20 0.80 * 2.23 0.24 1.27 -1.54 0.21
Mother's Education -0.29 *** 0.75 -0.10 0.91 -0.05 0.95 -0.37 0.69
Household Income -0.22 * 0.80 -0.03 0.97 -0.17 0.84 1.15 3.16
Missing Household Income 0.15 1.16 0.01 1.00 -0.48 0.62 -7.71 0.00
Deviant Peers 0.43 ***(g) 1.54 0.25 *** (r) 1.29 0.40 *** 1.49 0.32 1.37
-2 LL 1095.58 466.23 183.08 39.09
Chi-Square 479.92 *** 68.07 *** 78.37 *** 15.50
Nagelkerke R
2
0.45 0.23 0.45 0.39
N 1242 396 194 54
BIC -422.92 -20.22 -36.23 16.41
Notes: * = p < .05; ** = p < .01; *** = p < .001; (r) = racial difference; (g) = gender difference;
a
= two biological parents as the
reference
192
Table 8-11: Continued
White Females Black Females Latinas Asian Females
Beta Exp (B) Beta Exp (B) Beta Exp (B) Beta Exp (B)
Intercept -12.96 *** -7.11 *** -10.85 *** -3.84
Child's Age 0.71 *** 2.03 0.43 *** 1.54 0.52 *** 1.67 0.11 1.11
Unwedded Mothers
a
1.30 * 3.66 0.15 1.16 1.68 ** 5.39 -11.28 0.00
Divorced/Separated Mothers
a
0.72 *** 2.06 0.28 1.33 0.92 * 2.51 -9.94 0.00
Stepparents
a
0.66 ** 1.94 0.76 * 2.14 1.65 *** 5.22 -0.12 0.88
Mother's Education -0.04 0.96 -0.06 0.94 -0.06 0.94 -0.10 0.90
Household Income -0.13 0.88 -0.17 0.84 0.13 1.14 0.03 1.03
Missing Household Income -0.01 0.99 -0.22 0.80 0.06 1.07 0.42 1.52
Deviant Peers 0.36 *** 1.43 0.32 *** 1.38 0.25 ** 1.28 0.45 * 1.57
-2 LL 1158.43 529.08 221.82 40.58
Chi-Square 566.13 *** 122.70 *** 57.08 *** 12.11
Nagelkerke R
2
0.47 0.31 0.31 0.33
N 1372 476 232 50
BIC -508.34 -73.38 -13.51 19.19
Notes: * = p < .05; ** = p < .01; *** = p < .001; (r) = racial difference; (g) = gender difference;
a
= two biological parents as the
reference group
193
Table 8-12: Summary of the Family Structure and Adolescent Sexual Activity Relationship Within Race-Gender Categories
Significant Gender Difference Race Difference Partial Explanation Full Explanation
White Males
Unwedded Mothers Yes PA, DP *
Divorced/Separated Mothers Yes PS, PA, DP FT
Stepparent No
Black Males
Unwedded Mothers No
Divorced/Separated Mothers No
Stepparent Yes FT, PS *, PA, FS
Latinos
Unwedded Mothers No
Divorced/Separated Mothers Yes PA, DP *
Stepparent No
Notes: FT = Family Transitions; PS = Parental Supervision; PA = Parental Attitudes; FS = Family Support; DP = Deviant Peers
194
Table 8-12: Continued
Significant Gender Difference Race Difference Partial Explanation Full Explanation
White Females
Unwedded Mothers Yes PS
Divorced/Separated Mothers Yes PS *, DP
Stepparent Yes Yes FT
Black Females
Unwedded Mothers No Yes
Divorced/Separated Mothers No Yes
Stepparent Yes FT *, PS, PA, FS
Latinas
Unwedded Mothers Yes PA
Divorced/Separated Mothers Yes PS, PA *
Stepparent Yes Yes Yes (PS)
Notes: FT = Family Transitions; PS = Parental Supervision; PA = Parental Attitudes; FS = Family Support; DP = Deviant Peers
195
indicates whether the relationship between the non-traditional household and
adolescent sexual activity is statistically significant. The third column indicates
whether there is a significant non-traditional household effect by race-gender when
accounting only for the demographic variables. The fourth and fifth columns
indicate which explanation partially and statistically respectively accounts for the
non-traditional household effect. An asterisk indicates the explanation that produces
the largest drop in beta (otherwise known as the explanation that best mediates the
family structure effect) for that particular non-traditional household. The parenthesis
surrounding an explanation indicates that this particular explanation would have
been the best one to mediate that particular non-traditional household effect in cases
where none of the explanations was able to meet the level of partial mediation.
For white males, peer influence and family transition is the best reason for
the unwedded mother and divorced/separated mother household effects. Parental
supervision is the best reason for both the unwedded mother and the
divorced/separated mother household effects for white females. Finally, family
transition is the best reason for the stepparent household effect for white females.
Parental supervision is the best reason for the stepparent household effect for black
males. Family transition is the best reason the stepparent effect for blackfemales.
Peer influence is the best reason for the divorced/separated mother effect for Latinos.
However, parental attitude is the best reason for the unwedded mother and
divorced/separated mother household effects for Latinas.
196
Discussion
These analyses examined the relationship between family structure and
adolescent sexual activity for three non-traditional households within six race-gender
subgroups using various family influence and the peer influence explanations. There
are two parts to this section. The first part discusses the non-traditional household
effect within six race-gender subgroup categories. The second part discusses the
relationship between the five family influence measures and adolescent sexual
activity within these race-gender subgroups.
Non-traditional Households: White Adolescents
The stepparent effect among white adolescents was found only for females.
This finding, which is supported by Hetherington & Clingempeel (1992), seems to
suggest that white females may have a more difficult time than their male
counterpart with adjusting to the new parental figure at home. None of the five
explanations was able to at least partially explain this effect. This finding would
seem to imply that the stepparent household effect for white females could be due to
multiple reasons. Since family transition was found to be the best reason for the
stepparent household effect among the five explanations, it may be that the primary
reason for the stepparent household effect is due to the children feeling the stresses
of adjusting to the new family environment.
197
Finding that the stepparent household effect was not significant for sexual
activity for white males may imply that the new parental figure presence provides
additional stability to the former one-parent household. Stepparent households were
found to be similar to two biological parent households with respect to most of the
parental supervision variables. Another plausible reason as to why the stepparent
household effect is not significant for white males may be due to stepparents in these
analyses being males. Since the stepparent and stepsons are of similar gender, a
quicker bond may develop between these two individuals than between stepfathers
and their stepdaughters. The development of a close relationship with the stepfathers
may mitigate the stepson’s neglected feeling that he may experience during a family
transition event (Newcomer & Udry, 1987). This is because a stepfather could
represent someone who can provide support to his stepson by answering questions,
giving advice, and being there for him during difficult times. In addition, stepfathers
may also become a role model to their stepsons by demonstrating acceptable
behaviors. These actions could then result in a family environment similar to that of
two biological parent households. Thus, these findings provide further support for
the argument that there are gender differences in the way that the family environment
influences adolescent sexual activity.
A white male’s sexuality risk was greater for those who reside in single
parent households than for those who reside with two biological parents. Peer
influence was the best reason for the unwedded household effect. As mentioned in
the previous investigation, this finding may be due to several factors. First, these
unwedded mother households lack the presence of an adult male figure upon whom
198
sons can identify as a mentor. Thus, these sons may be using friends and peers as
role models in order to learn how they should act as young men.
The second factor is the neighborhood environment. It could be that
unwedded mother households of white males are significantly poorer than those of
two biological parent households. If this were the situation, these households may be
located in neighborhoods that provide a greater opportunity for children to be
acquainted with friends or peers who engage in deviant activities. Consequently,
having greater exposure to these type of friends may result in these individuals
having a very strong influence on an adolescent’s decision to engage in sexual
activity.
Family transition was the best reason for the divorced/separated mother
household effect for white males. This finding would seem to suggest that the
positive association with sexual activity for adolescents who reside in these
households is either due to problems with adjusting to a difficult family environment
during the transition period (Newcomer & Udry, 1987) or to the observation of their
parents dating behaviors (Newcomer & Udry, 1984; Thornton & Camburn, 1987;
Moore & Chase-Lansdale, 2001). For the dating behavior explanation, parents may
send these children a message that engaging in sexual activities outside of marriage
is acceptable, especially if they are part of the courtship process.
None of the explanations was able to at least partially account for the
unwedded mother and the divorced/separated mother household effects for white
females. However, parental supervision was found to be the best reason for this
effect. As mentioned in chapter 6, the amount of time the mother spends with her
199
daughter may be important for those who reside with a single mother because a lack
of it may directly lead to sexual activity during adolescence.
Non-traditional Households: Black Adolescents
The unwedded mother and divorced/separated mother households were not
positively associated with sexual activity for black males and females in the
regression analyses. However, the single household effects for black males were not
statistically different from those of white males. Thus, the non-significant
relationship between the single parent households and sexual activity found for black
adolescents in the previous investigation seems to be applicable only for black males.
The reason for the race-gender differences among black adolescents
regarding the single parent and adolescent sexual activity relationship may be due to
their position in the hierarchy relative to white adolescents. For example, the single
parent effect for white females is stronger but not significantly different from the
single parent effect for white males. In addition, these effects for black females are
less strong but not significantly different from the effects for black males. This
results in a hierarchy of the single parent effect from the strongest to the weakest for
adolescent sexual activity in the order of white females, white males, black males,
and black females. From this ordering, the single parent effect for black males is
closer to those of white males while this effect between black females and white
females is further apart. Thus, the results would show a racial difference for the
200
single parent household and adolescent sexual activity relationship between white
and black females.
Stepparent households were positively associated with sexual activity for
black adolescents of both genders. This finding is in contrast with studies by
McLanahan & Bumpass (1988), McLanahan & Sandefur (1994), and Moore (2001),
who did not find a positive association with sexual activity for black adolescents who
reside with a stepparent compared to those who reside with two biological parents.
All of the family influences were able to fully explain the relationship between
stepparent and adolescent sexual activity for black adolescents of both genders.
However, the parental supervision and family transition explanations were the best
reasons for the stepparent effect for black males and black females respectively.
Interestingly, peer influence did not explain the stepparent household effect for black
males and females. These finding suggests that the relevant issues regarding the
stepparent household effect has to do with factors associated with the household
rather than influences outside of the family.
Parental dating is probably the reason why family transition best explains the
stepparent household effect for black females. Although family members of
stepparent households may be currently going through, or have gone through an
adjustment period, the parental supervision measure does not reach the level of at
least partially explaining the stepparent effect for black females. This finding
suggests that black females probably get along with their stepfathers relatively better
than black males. In addition, it seems that black mothers with daughters who
remarry, or who had a child before marriage, may have relatively conservative
201
sexual attitudes compared to unwedded black mothers. This explanation may be the
reason why there was a minimal drop for stepparent household effect when
controlling for the parental attitude measure. Eliminating both of these reasons
would leave the parental dating explanation as a viable rationale as a reason why
family transition best explains the stepparent household effect. In another words, the
reason why black females who reside with a stepparents are at an increased risk for
sexual activity is because they may have previously observed their mother’s dating
behaviors when the family was in a transitory state and conclude that non-marital
sexual activity is acceptable (Whitbeck, Simons, & Kao, 1994).
Non-traditional Households: Hispanic Adolescents
The best reason for the non-traditional household and adolescent sexual
activity relationship varies by gender among Hispanic adolescents. These
adolescents tend to grow up in families that possess conservative attitudes regarding
sexuality due to the culture of familism where individual needs have less priority
compared to family ties and loyalty (Mirande, 1977). Consequently, Hispanic
parents are more likely to teach to their children traditional values that discourage
sexual behaviors during adolescence (East, 1998). Thus, one assumption would be
that there would be no increased risk for sexual activity among Hispanic adolescents
who reside in non-traditional households compared those who reside with two
biological parents due to the family’s emphasis on uphold traditional values and the
parents transmitting these to their children.
202
Despite this reasoning, the initial results for Hispanic adolescents revealed a
gender difference for the stepparent household effect where this effect is significant
only for Latinas. The stepparent effect was not significant for Latinos because of the
reasons that were suggested in earlier investigations such as the stepfather bonding
with and providing a role model for the stepson. In contrast, none of the
explanations was able to at least partially account for this effect for Latinas. This
finding would seem to suggest that the stepparent household effect could require
multiple explanations. However, parental supervision is the best explanation for this
effect for Latinas, which seems to suggest that the stepparent effect may be due to
either lower levels of parent-child interaction or a weakened parental authority.
Peer influence was the best reason for the divorced/separated mother effect
for Latinos. Meanwhile, parental attitude was the best reason for the unwedded
mother and the divorced/separated mother effects for Latinas. It seems that peers
who engage in deviant activities may have a strong influence on Latinos residing
with divorced/separated mothers. For Latinas, it may be that the unwedded mother
and the divorced/separated mother household effects is due to mothers being more
tolerant about adolescent sexual activity.
Summary of the Non-traditional Households Discussion
This investigation seemed to show an emergence of several themes. The first
theme is that adolescents who reside in non-traditional households may not
necessarily have an increased risk for sexual activity compared to those who reside
203
with two biological parents. For example, the risks for sexual activity for black
females who reside with a single mother are not significantly higher than those who
reside with two biological parents. One reason this finding may be due to help from
extended family members (Hines & Boyd-Franklin, 1996). Having extended family
members being involved in the family’s affairs may reduce the stresses and strains of
everyday life for family members that live in these households by providing
emotional support to the family and being able to become another source for the
monitoring of the children’s activities whenever the mother is not present in the
household. Thus, extended family members who are involved in households that are
headed by a single parent may help to maintain the household’s hierarchically
structure and reduce the adolescent’s need to seek emotional support from outside
sources.
The sexuality risk for white males and Latinos who reside with a stepparent
may not necessarily be greater compared to those who reside with two biological
parents. One reason for this finding may be due to the gender similarity between the
stepparent and his stepchild. This gender similarity may encourage a development of
a close relationship between these family members relatively quickly. The stepfather
may also be a role model for their stepsons in terms of demonstrating acceptable
behaviors and be a person who can support them during times of need.
The second theme is that the best reason for the non-traditional household
effect seemed to vary by family configuration across the six adolescent race-gender
subgroups. For example, the best reason for the single parent household and the
stepparent effects for white females and black males respectively seemed to be lower
204
levels of parent-child interaction or a weakened parental authority. Meanwhile, the
best reason for the single parent household effects for Latinas seemed to be mother’s
tolerant attitude toward adolescent sexual activity. In addition, the best reason for
the stepparent household effect for both black and white females seemed to be family
turbulence, which is possibly due to the stresses of integrating a new family member.
These examples demonstrate that these results revealed new findings of the non-
traditional household effects that could not be obtained by only analyzing the
adolescent’s gender or the adolescent’s race contexts.
Family and Peer Influence Variables Discussion
This investigation supported the hypothesis that family transitions would be
positively associated with sexual activity for all adolescents regardless of their race-
gender subgroup in most situations. These results do support the findings by
Newcomer & Udry (1987) and Wu & Thomson (2001). A number of reasons, such
as family turbulence, or parental dating during the family transition event, may
explain this relationship. However, significant racial differences regarding the
family transition effect between white females and black females were not found and
this finding does not support the study by Wu & Thomson (2001).
One exception to the hypothesis that family transition would be positively
associated with sexual activity was found for Latinas. Thus, the negative association
between family transition and sexual activity for Hispanic adolescents found in the
previous investigation is due to its effects on Latinas. In a separate cross tabulation
205
analysis (results not shown), there were a greater number of virgin Latinas who
experienced at least three family transitions compared to sexually experienced
Latinas. This finding seems to imply that Latinas who experience family transitions
are less likely to engage in sexual behaviors because they may have a closer
relationship to their mothers in times of familial change compared to those from
relatively stable households.
This investigation supported the hypothesis that mother-child activities,
parental strictness, and parental presence during meals would be negatively
associated with sexual activity for all adolescents regardless of their race-gender
subgroup. However, there was a gender difference among white adolescents for
parental presence during meals where this effect is stronger for white females than
males. This gender difference may be due to white females viewing these events as
times when they are able to interact with their parents in order to build a stronger
relationship with them.
The parental monitoring variables were proposed to not have any significant
association with sexual activity. However, there were several exceptions to this
hypothesis. Mother’s night presence was associated with preventing sexual activity
only for white males and this effect is significantly different from black males and
Latinos. Thus, the racial difference for this variable seen in the previous
investigation between whites and both blacks and Hispanics is mainly due to its
effects on white males. This finding suggests that the presence of these mothers at
home during evenings may be effective in preventing their sons from engaging in
sexual behaviors by being more vigilant about monitoring their activities.
206
In addition, mother’s night presence was positively associated with sexual
activity for Latinos but not for Latinas. There are several plausible reasons for this
finding. The first reason for this finding may be that Latinos have less responsibility
to help around the household when their mothers are present. This may be the
situation if mothers adhere to the marianismo concept where the female’s role as a
mother is emphasized and the mother’s self-sacrifice and suffering for her children is
celebrated (Ginorio et al., 1995). Mother’s presence in the household could then
provide their sons the opportunity to leave the household and engage in other
activities. This premise could be an instance where a parent’s presence around the
house may not necessarily equate to monitoring of their children’s activities.
The second reason for this finding may be that these mothers stay at home
more often when they have knowledge of or have suspicions of their son’s sexual
activity. This argument would be similar to either the mother-child general or sex
discussions where Hispanic mothers become more involved in their sons lives
because they view their behaviors as troubling. Thus, increasing parental presence
may be one way for mothers to try and prevent their sons from engaging in future
sexual activity.
The results of this investigation revealed three exceptions to the proposed
hypothesis that mother’s religion and her disapproval of adolescent sexual activity
are negatively associated while mother-child sex discussions are positively
associated with sexual activity for all adolescents regardless of their race-gender
subgroup. First, mother’s religiosity was negatively associated with sexual activity
for white but not for black males. Thus, the racial difference for this variable that
207
was seen in the previous investigation was mainly due to the non-significant
association between mother’s religiosity and sexual activity for black males.
However, the relationship between mother’s disapproval of adolescent sexual
activity and sexual activity are similar for both black and white males. Finding that
mother’s religiosity is not associated with sexual activity for black males, even
though mothers of black males are more religious compared to mothers of white
males, seems to suggest that mothers who use religious values when expressing
concern about this issue will not influence their son’s sexual behaviors. Instead,
these males may be more willing to honor their mothers values of not engaging in
sexual activity during puberty if the issue was framed in terms of how it may have a
social impact their lives such as getting someone pregnant and being a father at a
young age (Sonnenstein, 1986).
The second exception to the proposed hypothesis was that mother’s
disapproval of adolescent sexual activity is negatively associated with sexual activity
only for white females. Thus, it seems that the racial difference of mother’s
disapproval of adolescent sex that was found in the previous investigation between
whites and both blacks and Hispanics could be due to the non-significant association
between mother’s disapproval of adolescent sexual activity with sexual activity for
black females and Latinas. This finding does not support the study by Jaccard,
Dittus, & Gordon (1996) who found that that mother’s disapproval of adolescent
sexual activity was negatively associated with sexual activity for black adolescents.
In addition, the non-association between mother’s disapproval of adolescent sexual
activity and sexual activity for Latinas did not support the studies by Slonim-Nevo
208
(1992), Hovel et al. (1994), and Miller, Forehand, & Kotchick (1999). It seems that
mother’s disapproval of adolescent sexual activity was not associated with sexual
activity for black females and Latinas because their mothers are generally more
tolerant of their daughter’s engagement in this type of activity. From another point
of view, the relationship between mother’s disapproval of adolescent sexual activity
and adolescent sexual behaviors is very strong for white females which further
contributes to the race difference for this effect between white females and both
black females and Latinas. However, mother’s religious values seem to have more
of an impact on both black females’ and Latinas’ decision to engage in sexual
activity than their mother’s disapproval of these behaviors.
The third exception to the proposed hypothesis was the finding that mother’s
disapproval of adolescent sexual activity was negatively associated with sexual
activity for Latinos but not for Latinas. This gender difference refutes the idea that
was speculated in the previous investigation of mother’s disapproval of adolescent
sexual activity not being associated with sexual activity because Hispanic mothers
were more approving of their son’s sexual activity while being more disapproving of
their daughter’s sexual activity during puberty. Instead, the finding seems to suggest
that Latinos may be more concerned than Latinas about upholding their mother’s
values if they disapprove of adolescent sexual activity as a respect for the family’s
values (Mirande, 1977).
The hypothesis that family support is negatively associated with sexual
activity for all adolescents regardless of their race-gender subgroup was supported by
this investigation. These results support studies that find family support to be
209
negatively associated with sexual activity for white (Lauritsen, 1994; Whitbeck et
al., 1992) and black (Ramirez-Valles, Zimmerman, & Juarez, 2002; Moore & Chase-
Lansdale, 2001) adolescents. The family support effect was also not significantly
different between white and Hispanic adolescents of both genders. In another words,
the family support effect seemed to be significant for sexual activity for Latinos and
Latinas which also supports the study by Miller, Forehand, & Kotchick (1999).
The results of this investigation revealed support for the hypothesis that peer
influence would be positively associated with sexual activity for all adolescents
regardless of their race-gender subgroup. These results are supported by studies that
found acquaintance with deviant peers being positively associated with sexual
activity for white (Whitbeck, Simmons, & Kao, 1994), black (Furstenberg et al.,
1987), and Hispanic (Christopher, Johnson, & Roosa, 1993) adolescents. The peer
influence effect was also much stronger for white males than for both black males
and white females which seems to imply that friends may have more influence over a
white male’s decision to engage in sexual activity than these other adolescents.
Relative Importance of the Five Explanations Discussion
This study found that peer influence was the most parsimonious while family
transition was the least parsimonious adolescent sexual activity explanation for all of
the adolescent’s race-gender subgroup categories. When excluding peer influence,
parental attitude was the most parsimonious adolescent sexual activity explanation
for white adolescents of both genders and for Latinos according to the BIC statistic.
210
Meanwhile, family support was the most parsimonious sexuality explanation for both
black males and females. Finally, family transition was the most parsimonious
sexuality explanation for Latinas according to the BIC statistic.
The reason for these findings were somewhat addressed in the previous
investigation. Briefly, parental attitudes being the most parsimonious family
influence explanation for white adolescents of both genders and Latinos is probably
due to these mothers having a huge influence on their children’s decision to engage
in sexual activity through the transmission of sexual values from the parent to the
child. Family support being the most parsimonious sexuality explanation for black
adolescents of both genders is probably due to the greater likelihood for these
adolescents to reside in non-traditional households. These non-traditional
households may need help from extended family members (Hines & Boyd-Franklin,
1996), which could result in the family members of these households viewing their
relatives as being part of the family. Consequently, the family support effect may be
strengthened for adolescents who broaden the definition of family support by
considering their relatives as being part of the household compared to those who
define their household that consists only of immediate family members. Finally, this
section will later expand on the reason why family transition was the most
parsimonious adolescent sexual activity explanation for Latinas.
The least parsimonious adolescent sexual activity explanation for white
males, black males, Latinos, black females, Latinas was parental supervision. This
result is mainly due to the parental monitoring variables having no significant
association with adolescent sexual activity. Yet, family transition was the least
211
parsimonious adolescent sexual activity explanation for white females, which
probably suggests that other measures may be explaining this effect.
Chapter Summary
The results in this investigation further demonstrated that there are subtle
differences regarding the family structure and adolescent sexual activity relationship
across three non-traditional households when the data is analyzed within the
adolescent’s race-gender context than within the adolescent’s gender or adolescent’s
race contexts alone. This investigation found that there were other non-traditional
family configurations, other than the single parent households for black adolescents
that did not exhibit an increased risk for sexual activity compared to two biological
parent households. In addition, there were also subtle racial differences by gender
for the explanations that best accounted for the family structure effect across the
three non-traditional households.
The most parsimonious sexuality explanation for five of the adolescent race-
gender subgroups was similar to those found in the previous investigation. Hispanic
adolescents were an exception to this finding where parental attitude was the most
parsimonious adolescent sexual activity explanation for Latinos, but family transition
was the most parsimonious adolescent sexual activity explanation for Latinas.
However, this finding is somewhat controversial because a cross tabulation analysis
between sexual initiation status and family transition revealed a greater number of
212
virgin Latinas experiencing at least three family transition events compared to
sexually experienced Latinas.
Parental supervision was the least parsimonious adolescent sexual activity
explanation for white males, black males, Latinos, black females, and Latinas. This
finding is mainly due to many of the variables that comprise the measure not having
a significant association with adolescent sexual activity, especially for the parental
monitoring variables. Meanwhile, the least parsimonious adolescent sexual activity
explanation for white females was family transition, which suggests that this effect
may be explained by other measures.
There was plenty of support for the proposed hypotheses throughout this
investigation. However, some race and gender differences for the parental
supervision and parental attitude measures were found for several of the adolescent
race-gender subgroups. These findings further clarified which gender was
responsible for the race differences found in the previous investigation.
Consequently, the results in this investigation emphatically support the argument by
Bronfenbrenner (1979) that family processes always occurs with a specific context
and that these family processes may have different effects for different groups of
individuals in distinct family settings.
The results of this investigation revealed that a conflict measure, namely peer
influence, was a better explanation for adolescent sexual activity than any of the
functional measures for all of the race-gender subgroups. In contrast, functional
measures were mostly the best ones to explain the non-traditional household effect
within the adolescent’s race-gender categories. These findings imply that peers have
213
a strong influence on whether a child decides to engage in sexual activity while the
best reason why children who live in non-traditional households are more likely to
engage in sexual activity is predominately due to how parents socialize their
children.
214
Chapter 9: Hierarchically Ordered Regression Analyses
The second investigation of this study focuses on determining which factors
are directly associated with adolescent sexual activity after controlling for other
measures by using hierarchically ordered regression analyses where the variables are
regressed in a stepwise manner. This investigation also determines which factors are
indirectly associated with adolescent sexual activity after controlling for other
measures.
There are two main reasons for this investigation. First, many studies on
adolescent sexual activity, such as those by Newcomer & Udry (1987) and Wu &
Martinson (1993), have only tested one or two family influence explanations at a
time. This method compels researchers to speculate whether other explanations
could be responsible for their findings rather than being able to test this premise.
Thus, these studies only present a partial view of the relationship between family
influence and adolescent sexual activity because other untested explanations that
could explain these findings were not controlled for in these studies.
Second, some researchers have suggested that the distinctions between some
of the family socialization measures are not absolute and that they may be associated
with each other, such as the lack of parental control in the household explaining the
family transition effect on adolescent sexual activity (Newcomer & Udry, 1987; Wu
& Thomson, 2001). It would be difficult for researchers to determine which of these
explanations may be spurious or are indirectly associated with adolescent sexual
activity when other explanations have not been controlled in the analyses. Thus, the
215
measures that are used throughout this may be associated with each other in some
ways and the contribution of each measure on explaining adolescent sexual activity
can only be resolved by controlling for other measures.
There are nine analyses that are conducted for this investigation: the total
sample, within the adolescent’s gender, within the adolescent’s race (whites and
black), and within the adolescent’s race-gender (white and black adolescents of both
genders) contexts. Most of the discussion for this investigation will focus on the full
model (the model where all of the variables are regressed simultaneously). This
investigation will also determine whether there are significant gender and racial
differences for the relationship between the variables that comprise the five measures
and sexuality by analyzing the data within the adolescent’s gender, the adolescent’s
race, and the adolescent’s race-gender contexts and using the t-test for difference
between slopes analyses (Jaccard, Turrisi, & Wan, 1990).
Analyses
Table 9-1 displays the results of the hierarchically ordered regression analysis
on the total sample. The full regression model (model 6) shows that mother-child
activity, parental meal presence, parental strictness, mother’s disapproval of
adolescent sex, and family support are negatively associated while mother-child
discussions, mother-child sex discussions, and being acquainted with deviant peers
are positively associated with sexual activity. The analysis also reveals the
216
Table 9-1: Hierarchically Ordered Regression of Adolescent Sexual Activity on the
Demographic, Family Influence, and Peer Influence Variables for the
Total Sample (N = 4016)
Model 1 Model 2
Beta Exp (B) Beta Exp (B)
Intercept -9.19 *** -9.59 ***
Child's Age 0.57 *** 1.76 0.57 *** 1.77
Male
a
-0.06 0.94 -0.06 0.94
Black
b
0.35 ** 1.42 0.40 ** 1.50
Hispanic
b
-0.52 ** 0.59 -0.51 ** 0.60
Asian
b
-0.66 0.51 -0.65 0.52
Unwedded Mothers
c
0.68 *** 1.97 0.60 ** 1.81
Divorced/Separated Mothers
c
0.68 *** 1.98 0.54 *** 1.72
Stepparent
c
0.65 *** 1.92 0.42 *** 1.53
Mother's Education -0.13 *** 0.88 -0.13 *** 0.88
Household Income (ln) -0.13 * 0.88 -0.13 * 0.88
Missing Household Income -0.08 0.92 -0.06 0.94
Black*Gender 0.67 *** 1.95 0.66 *** 1.94
Hispanic * Gender 0.67 ** 1.96 0.68 ** 1.97
Asian * Gender 0.09 1.09 0.09 1.09
Family Transitions 0.23 *** 1.26
-2 LL 4340.07 4324.34
Chi-Square 926.59 *** 942.32 ***
Nagelkerke R
2
0.28 0.29
BIC -810.42 -817.85
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= females as the reference group;
b
= whites as the reference group;
c
= two biological parents as the reference
group
217
Table 9-1: Continued
Model 3 Model 4
Beta Exp (B) Beta Exp (B)
Intercept -6.60 *** -6.17 ***
Child's Age 0.47 *** 1.60 0.45 *** 1.57
Male
a
0.04 1.04 0.12 1.13
Black
b
0.44 ** 1.55 0.42 ** 1.52
Hispanic
b
-0.50 ** 0.61 -0.43 * 0.65
Asian
b
-0.51 0.60 -0.32 0.72
Unwedded Mothers
c
0.44 * 1.55 0.29 1.34
Divorced/Separated Mothers
c
0.31 ** 1.36 0.22 1.25
Stepparent
c
0.32 * 1.37 0.27 * 1.32
Mother's Education -0.14 *** 0.87 -0.12 ** 0.89
Household Income (ln) -0.20 *** 0.82 -0.20 *** 0.82
Missing Household Income -0.04 0.96 -0.06 0.95
Black*Gender 0.70 *** 2.02 0.68 *** 1.98
Hispanic * Gender 0.69 ** 1.99 0.66 * 1.94
Asian * Gender -0.10 0.90 -0.11 0.90
Family Transitions 0.20 ** 1.22 0.15 * 1.16
Mother-Child Activities -0.29 *** 0.75 -0.25 *** 0.78
Mother-Child Discussions 0.23 *** 1.26 0.20 *** 1.22
Parent Strictness -0.16 *** 0.85 -0.16 *** 0.85
Parent Meal Presence -0.08 *** 0.92 -0.09 *** 0.92
Full Time Employment 0.09 1.09 0.06 1.06
Mother's Morning Presence -0.05 0.95 -0.05 0.95
Mother's Afternoon Presence -0.04 0.96 -0.04 0.96
Mother's Night Presence -0.04 0.96 -0.02 0.98
Mother's Religiosity -0.07 ** 0.93
Sex Communication 0.34 *** 1.40
Disapprove Sex -0.18 *** 0.83
-2 LL 4123.46 4034.19
Chi-Square 1143.20 *** 1232.47 ***
Nagelkerke R
2
0.34 0.36
BIC -944.05 -1016.72
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= females as the reference group;
b
= whites as the reference group;
c
= two biological parents as the reference
group
218
Table 9-1: Continued
Model 5 Model 6
Beta Exp (B) Beta Exp (B)
Intercept -5.00 *** -6.39 ***
Child's Age 0.45 *** 1.58 0.41 *** 1.51
Male
a
0.15 1.16 0.13 1.14
Black
b
0.44 ** 1.56 0.82 *** 2.27
Hispanic
b
-0.46 * 0.63 -0.27 0.76
Asian
b
-0.36 0.70 -0.37 0.69
Unwedded Mothers
c
0.32 1.38 0.30 1.35
Divorced/Separated Mothers
c
0.21 1.23 0.16 1.18
Stepparent
c
0.23 1.26 0.26 1.29
Mother's Education -0.12 ** 0.88 -0.11 ** 0.89
Household Income (ln) -0.20 ** 0.82 -0.18 ** 0.83
Missing Household Income -0.04 0.96 -0.05 0.95
Black*Gender 0.73 *** 2.07 0.57 ** 1.77
Hispanic * Gender 0.73 ** 2.08 0.57 * 1.77
Asian * Gender -0.11 0.90 -0.02 0.98
Family Transitions 0.14 * 1.15 0.10 1.10
Mother-Child Activities -0.20 *** 0.82 -0.14 ** 0.87
Mother-Child Discussions 0.20 *** 1.23 0.16 *** 1.18
Parent Strictness -0.16 *** 0.85 -0.14 *** 0.87
Parent Meal Presence -0.07 *** 0.94 -0.06 ** 0.95
Full Time Employment 0.08 1.09 0.13 1.14
Mother's Morning Presence -0.04 0.96 -0.03 0.97
Mother's Afternoon Presence -0.03 0.97 -0.01 0.99
Mother's Night Presence 0.01 1.01 0.04 1.04
Mother's Religiosity -0.07 ** 0.93 -0.05 0.95
Sex Communication 0.35 *** 1.42 0.32 *** 1.37
Disapprove Sex -0.19 *** 0.83 -0.18 *** 0.82
Family's Support -0.11 *** 0.90 -0.07 *** 0.94
Deviant Peers 0.30 *** 1.35
-2 LL 3987.21 3655.25
Chi-Square 1279.45 *** 1611.41 ***
Nagelkerke R
2
0.37 0.45
BIC -1055.40 -1379.06
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= females as the reference group;
b
= whites as the reference group;
c
= two biological parents as the reference
group
219
mediation of the family transition and mother’s religiosity variables after controlling
for all measures.
Table 9-2 displays the results of the hierarchically ordered regression analysis
for males. The full regression model shows that mother’s disapproval of adolescent
sexual activity and family support explanations are negatively associated while
mother-child discussions, mother-child sex discussions, and being acquainted with
deviant peers are positively associated with sexual activity. The analysis also reveals
the mediation of the family transition and the mother-child activity variables after
controlling for every measure. Family support mediates parental meal presence,
which suggests, for males, that this activity is an indication of family support.
Table 9-3 displays the results of the hierarchically ordered regression analysis
for females. The result for females is similar to that for the male’s analysis except
that mother-child activities and parental strictness are also negatively associated with
sexual activity for females. However, a gender difference is only found for peer
influence where this effect is much stronger for males than females. Thus, the
factors associated with adolescent sexual activity are essentially the same for both
males and females. The analysis also reveals the mediation of the family transition
and mother’s religiosity variables after controlling for every measure where peer
influence mediates the mother’s religiosity effect while the parental supervision
measure mediates the family transition effect.
Table 9-4 displays the results of the hierarchically ordered regression analysis
for white adolescents. The results of the full model show that parental meal
presence, parental strictness, mother’s disapproval of adolescent sex, and family
220
Table 9-2: Hierarchically Ordered Regression of Adolescent Sexual Activity on the
Demographic, Family Influence, and Peer Influence Variables For Males
(N = 1886)
Model 1 Model 2
Beta Exp (B) Beta Exp (B)
Intercept -8.17 *** -8.68 ***
Child's Age 0.51 *** 1.66 0.52 *** 1.68
Blacks
a
1.00 *** 2.73 1.07 *** 2.90
Hispanics
a
0.11 1.12 0.13 1.13
Asians
a
-0.56 0.57 -0.54 0.59
Unwedded Mothers
b
0.67 * 1.95 0.59 * 1.81
Divorced/Separated Mothers
b
0.65 *** 1.92 0.47 ** 1.60
Stepparent
b
0.43 ** 1.53 0.13 1.14
Mother's Education -0.18 *** 0.84 -0.18 ** 0.84
Household Income (ln) -0.12 0.88 -0.12 0.89
Missing Household Income -0.04 0.96 -0.03 0.97
Family Transitions 0.29 ** 1.33
-2 LL 2106.21 2095.09
Chi-Square 415.82 *** 426.95 ***
Nagelkerke R
2
0.27 0.28
BIC -340.40 -343.99
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= whites as the reference group;
b
= two biological parents as the reference group
221
Table 9-2: Continued
Model 3 Model 4
Beta Exp (B) Beta Exp (B)
Intercept -5.98 *** -5.13 ***
Child's Age 0.42 *** 1.52 0.39 *** 1.47
Blacks
a
1.15 *** 3.15 1.11 *** 3.05
Hispanics
a
0.15 1.16 0.18 1.20
Asians
a
-0.54 0.58 -0.36 0.70
Unwedded Mothers
b
0.43 1.54 0.26 1.30
Divorced/Separated Mothers
b
0.27 1.31 0.17 1.19
Stepparent
b
0.01 1.01 -0.02 0.98
Mother's Education -0.20 *** 0.82 -0.17 ** 0.84
Household Income (ln) -0.19 * 0.83 -0.19 * 0.83
Missing Household Income 0.01 1.01 -0.02 0.98
Family Transitions 0.26 ** 1.29 0.19 * 1.21
Mother-Child Activities -0.27 *** 0.77 -0.24 *** 0.79
Mother-Child Discussions 0.23 *** 1.26 0.19 *** 1.21
Parent Strictness -0.13 ** 0.87 -0.13 ** 0.88
Parent Meal Presence -0.06 * 0.94 -0.06 * 0.94
Full Time Employment 0.04 1.04 -0.01 0.99
Mother's Morning Presence -0.03 0.97 -0.02 0.98
Mother's Afternoon Presence -0.06 0.94 -0.07 0.93
Mother's Night Presence -0.02 0.98 0.01 1.01
Mother's Religiosity -0.06 0.94
Sex Communication 0.33 *** 1.39
Disapprove Sex -0.21 *** 0.81
-2 LL 2019.88 1975.58
Chi-Square 502.15 *** 546.46 ***
Nagelkerke R
2
0.32 0.34
BIC -358.85 -380.53
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= whites as the reference group;
b
= two biological parents as the reference group
222
Table 9-2: Continued
Model 5 Model 6
Beta Exp (B) Beta Exp (B)
Intercept -3.99 *** -4.84 ***
Child's Age 0.38 *** 1.47 0.31 *** ^ 1.36
Blacks
a
1.16 *** 3.18 1.42 *** ^ 4.14
Hispanics
a
0.21 1.23 0.23 1.25
Asians
a
-0.40 0.67 -0.32 0.73
Unwedded Mothers
b
0.32 1.37 0.27 1.30
Divorced/Separated Mothers
b
0.16 1.17 0.09 1.09
Stepparent
b
-0.06 0.94 -0.05 ^ 0.96
Mother's Education -0.18 ** 0.83 -0.21 ** ^ 0.81
Household Income (ln) -0.18 * 0.83 -0.20 * 0.82
Missing Household Income -0.02 0.98 -0.03 0.97
Family Transitions 0.19 * 1.21 0.16 1.18
Mother-Child Activities -0.20 ** 0.82 -0.11 0.90
Mother-Child Discussions 0.20 *** 1.22 0.17 *** 1.18
Parent Strictness -0.14 ** 0.87 -0.12 * 0.89
Parent Meal Presence -0.04 0.96 -0.03 0.97
Full Time Employment 0.02 1.02 0.10 1.11
Mother's Morning Presence -0.01 0.99 -0.03 0.97
Mother's Afternoon Presence -0.06 0.94 -0.03 0.97
Mother's Night Presence 0.04 1.04 0.09 1.09
Mother's Religiosity -0.06 0.94 -0.04 0.96
Sex Communication 0.34 *** 1.40 0.30 *** 1.35
Disapprove Sex -0.21 *** 0.81 -0.18 ** 0.83
Family's Support -0.09 *** 0.91 -0.05 * 0.95
Deviant Peers 0.35 *** ^ 1.42
-2 LL 1959.38 1749.90
Chi-Square 562.66 *** 772.14 ***
Nagelkerke R
2
0.35 0.46
BIC -389.19 -591.13
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= whites as the reference group;
b
= two biological parents as the reference group; ^ = gender difference
223
Table 9-3: Hierarchically Ordered Regression of Adolescent Sexual Activity on the
Demographic, Family Influence, and Peer Influence Variables For
Females (N = 2130)
Model 1 Model 2
Beta Exp (B) Beta Exp (B)
Intercept -10.27 *** -10.60 ***
Child's Age 0.62 *** 1.86 0.63 *** 1.87
Blacks
a
0.38 ** 1.46 0.42 ** 1.52
Hispanics
a
-0.50 ** 0.61 -0.49 ** 0.61
Asians
a
-0.71 0.49 -0.70 0.50
Unwedded Mothers
b
0.67 ** 1.95 0.59 * 1.80
Divorced/Separated Mothers
b
0.71 *** 2.04 0.60 *** 1.83
Stepparent
b
0.88 *** 2.41 0.69 *** 2.00
Mother's Education -0.09 0.92 -0.08 0.92
Household Income (ln) -0.14 0.87 -0.14 0.87
Missing Household Income -0.08 0.93 -0.06 0.94
Family Transitions 0.19 * 1.21
-2 LL 2222.45 2216.73
Chi-Square 512.06 *** 517.78 ***
Nagelkerke R
2
0.30 0.30
BIC -435.42 -433.48
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= whites as the reference group;
b
= two biological parents as the reference group
224
Table 9-3: Continued
Model 3 Model 4
Beta Exp (B) Beta Exp (B)
Intercept -7.24 *** -7.06 ***
Child's Age 0.52 *** 1.69 0.51 *** 1.67
Blacks
a
0.46 ** 1.58 0.43 ** 1.54
Hispanics
a
-0.47 * 0.62 -0.40 * 0.67
Asians
a
-0.58 0.56 -0.40 0.67
Unwedded Mothers
b
0.41 1.51 0.27 1.32
Divorced/Separated Mothers
b
0.32 * 1.38 0.24 1.27
Stepparent
b
0.60 ** 1.81 0.54 ** 1.72
Mother's Education -0.10 0.91 -0.07 0.94
Household Income (ln) -0.22 ** 0.80 -0.23 ** 0.80
Missing Household Income -0.05 0.95 -0.05 0.95
Family Transitions 0.15 1.17 0.11 1.12
Mother-Child Activities -0.32 *** 0.73 -0.26 *** 0.77
Mother-Child Discussions 0.24 *** 1.27 0.21 *** 1.24
Parent Strictness -0.20 *** 0.82 -0.20 *** 0.82
Parent Meal Presence -0.11 *** 0.90 -0.12 *** 0.89
Full Time Employment 0.14 1.15 0.13 1.14
Mother's Morning Presence -0.08 0.92 -0.08 0.92
Mother's Afternoon Presence -0.02 0.98 -0.02 0.98
Mother's Night Presence -0.07 0.93 -0.04 0.96
Mother's Religiosity -0.07 * 0.93
Sex Communication 0.35 *** 1.42
Disapprove Sex -0.18 *** 0.84
-2 LL 2081.55 2035.19
Chi-Square 652.96 *** 699.32 ***
Nagelkerke R
2
0.37 0.39
BIC -507.35 -530.71
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= whites as the reference group;
b
= two biological parents as the reference group
225
Table 9-3: Continued
Model 5 Model 6
Beta Exp (B) Beta Exp (B)
Intercept -5.85 *** -7.64 ***
Child's Age 0.53 *** 1.70 0.52 *** 1.68
Blacks
a
0.47 ** 1.61 0.84 *** 2.31
Hispanics
a
-0.42 * 0.66 -0.20 0.82
Asians
a
-0.46 0.63 -0.46 0.63
Unwedded Mothers
b
0.28 1.33 0.24 1.27
Divorced/Separated Mothers
b
0.21 1.23 0.19 1.21
Stepparent
b
0.49 ** 1.64 0.52 ** 1.68
Mother's Education -0.07 0.94 -0.03 0.97
Household Income (ln) -0.22 ** 0.81 -0.18 * 0.83
Missing Household Income -0.01 0.99 -0.01 0.99
Family Transitions 0.10 1.11 0.05 1.05
Mother-Child Activities -0.20 ** 0.82 -0.17 ** 0.85
Mother-Child Discussions 0.22 *** 1.24 0.17 *** 1.18
Parent Strictness -0.20 *** 0.82 -0.17 ** 0.85
Parent Meal Presence -0.08 *** 0.92 -0.07 ** 0.93
Full Time Employment 0.16 1.17 0.16 1.18
Mother's Morning Presence -0.07 0.93 -0.04 0.96
Mother's Afternoon Presence -0.01 0.99 0.02 1.02
Mother's Night Presence -0.03 0.97 -0.01 0.99
Mother's Religiosity -0.08 * 0.93 -0.05 0.95
Sex Communication 0.38 *** 1.47 0.35 *** 1.42
Disapprove Sex -0.18 *** 0.84 -0.18 *** 0.83
Family's Support -0.13 *** 0.88 -0.09 *** 0.91
Deviant Peers 0.27 *** 1.32
-2 LL 2001.12 1865.91
Chi-Square 733.39 *** 868.60 ***
Nagelkerke R
2
0.40 0.46
BIC -557.12 -684.67
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= whites as the reference group;
b
= two biological parents as the reference group
226
Table 9-4: Hierarchically Ordered Regressions of Sexual Activity on the
Demographic, Family Influence, and Peer Influence Variables for White
Adolescents (N = 2614)
Model 1 Model 2
Beta Exp (B) Beta Exp (B)
Intercept -10.61 *** -11.22 ***
Child's Age 0.67 *** 1.94 0.68 *** 1.97
Male
a
-0.07 0.93 -0.07 0.93
Unwedded Mothers
b
1.60 *** 4.93 1.40 *** 4.05
Divorced/Separated Mothers
b
0.82 *** 2.27 0.58 *** 1.79
Stepparent
b
0.57 *** 1.77 0.18 1.20
Mother's Education -0.18 *** 0.84 -0.16 ** 0.85
Household Income (ln) -0.17 * 0.85 -0.17 * 0.85
Missing Household Income 0.06 1.06 0.06 1.06
Family Transitions 0.36 *** 1.43
-2LL 2665.26 2647.73
Chi-square 634.99 *** 652.52 ***
Nagelkerke R
2
0.30 0.31
BIC -572.04 -581.70
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= females as the reference group;
b
= two biological parents as the reference group
227
Table 9-4: Continued
Model 3 Model 4
Beta Exp (B) Beta Exp (B)
Intercept -7.47 *** -6.64 ***
Child's Age 0.56 *** 1.75 0.54 *** 1.71
Male
a
0.02 1.02 0.12 1.13
Unwedded Mothers
b
1.07 ** 2.93 0.97 * 2.64
Divorced/Separated Mothers
b
0.32 * 1.38 0.20 1.22
Stepparent
b
0.07 1.07 -0.01 0.99
Mother's Education -0.18 *** 0.84 -0.14 ** 0.87
Household Income (ln) -0.24 ** 0.78 -0.25 ** 0.78
Missing Household Income 0.10 1.10 0.05 1.05
Family Transitions 0.32 *** 1.38 0.27 ** 1.31
Mother-Child Activities -0.32 *** 0.72 -0.26 *** 0.77
Mother-Child Discussions 0.25 *** 1.28 0.21 *** 1.23
Parent Strictness -0.16 ** 0.85 -0.15 ** 0.86
Parent Meal Presence -0.08 *** 0.92 -0.09 *** 0.91
Full Time Employment 0.03 1.03 0.01 1.01
Mother's Morning Presence -0.04 0.96 -0.04 0.96
Mother's Afternoon Presence -0.06 0.94 -0.07 0.94
Mother's Night Presence -0.15 * 0.86 -0.12 0.89
Mother's Religiosity -0.08 ** 0.92
Sex Communication 0.38 *** 1.46
Disapprove Sex -0.29 *** 0.75
-2 LL 2513.90 2428.90
Chi-Square 786.35 *** 871.35 ***
Nagelkerke R
2
0.36 0.40
BIC -652.58 -713.98
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= females as the reference group;
b
= two biological parents as the reference group
228
Table 9-4: Continued
Model 5 Model 6
Beta Exp (B) Beta Exp (B)
Intercept -5.63 *** -7.31 ***
Child's Age 0.54 *** 1.72 0.52 *** 1.68
Male
a
0.16 1.17 0.12 1.13
Unwedded Mothers
b
0.93 * 2.52 0.66 1.93
Divorced/Separated Mothers
b
0.18 1.19 0.16 1.17
Stepparent
b
-0.06 0.94 -0.02 0.98
Mother's Education -0.14 ** 0.87 -0.12 * 0.89
Household Income (ln) -0.24 ** 0.79 -0.22 ** 0.80
Missing Household Income 0.04 1.04 0.03 1.03
Family Transitions 0.26 ** 1.30 0.18 1.20
Mother-Child Activities -0.20 *** 0.82 -0.11 0.90
Mother-Child Discussions 0.21 *** 1.24 0.15 *** 1.17
Parent Strictness -0.16 ** 0.85 -0.11 * 0.90
Parent Meal Presence -0.07 ** 0.93 -0.05 * 0.95
Full Time Employment 0.05 1.05 0.04 1.04
Mother's Morning Presence -0.03 0.98 -0.01 0.99
Mother's Afternoon Presence -0.05 0.95 -0.02 0.98
Mother's Night Presence -0.09 0.91 -0.09 0.92
Mother's Religiosity -0.08 ** 0.92 -0.06 0.94
Sex Communication 0.40 *** 1.50 0.38 *** 1.46
Disapprove Sex -0.29 *** 0.75 -0.29 *** 0.75
Family's Support -0.10 *** 0.90 -0.06 * 0.94
Deviant Peers 0.34 *** 1.40
-2 LL 2403.27 2150.99
Chi-Square 896.98 *** 1149.26 ***
Nagelkerke R
2
0.41 0.50
BIC -731.74 -976.15
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= females as the reference group;
b
= two biological parents as the reference group
229
support are negatively while mother-child discussions, mother-child sex discussions,
and being acquainted with deviant peers are positively associated with sexual
activity. The analysis also reveals the mediation of the family transition, mother-
child activity, and mother’s religiosity variables, which only occurs after controlling
for all of the measures.
Table 9-5 displays the results of the hierarchically ordered regression analysis
for black adolescents. There is a significant race difference from white adolescents
for mother’s disapproval of adolescent sex and peer influence where these effects are
significantly weaker for black adolescents regarding sexual activity. This analysis
does not reveal an eventual mediation for any of the significant variables that were
found in the previous regression models.
Table 9-6 displays the results of the hierarchically ordered regression analysis
for white males. The full model shows that mother’s disapproval of adolescent
sexual activity is negatively while mother-child discussions, mother-child sex
discussions, and acquaintance with deviant peers are positively associated with
sexual activity. The analysis also reveals an eventual mediation of the family
transitions, mother-child activities, mother’s night presence, and family support
variables. Both the parental attitude and deviant peer measures contribute to the
mediation of the family transition measure, which suggests that the family transition
effect for white males is due to the combination of mother’s less conservative sexual
attitudes and coping with the stress of a turbulent household by associating
themselves with peers who engage in deviant activities. Family support mediates
mother’s night presence, which suggests that mother’s presence at night is
230
Table 9-5: Hierarchically Ordered Regression of Sexual Activity on the
Demographic, Family Influence, and Peer Influence Variables for Black
Adolescents (N = 872)
Model 1 Model 2
Beta Exp (B) Beta Exp (B)
Intercept -6.26 *** -6.60 ***
Child's Age 0.41 *** ^ 1.51 0.42 *** ^ 1.52
Male
a
0.58 *** ^ 1.79 0.58 *** ^ 1.79
Unwedded Mothers
b
0.31 ^ 1.37 0.29 ^ 1.34
Divorced/Separated Mothers
b
0.31 ^ 1.37 0.23 1.26
Stepparent
b
0.69 ** 1.99 0.53 * 1.70
Mother's Education -0.10 0.91 -0.10 0.90
Household Income (ln) -0.14 0.87 -0.13 0.88
Missing Household Income -0.18 0.83 -0.15 0.86
Family Transitions 0.22 * 1.24
-2 LL 1066.67 1061.79
Chi-Square 141.95 *** 146.84 ***
Nagelkerke R
2
0.20 0.21
BIC -87.78 -85.90
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= females as the reference group;
b
= two biological parents as the reference group; ^ = race difference
231
Table 9-5: Continued
Model 3 Model 4
Beta Exp (B) Beta Exp (B)
Intercept -4.36 *** -5.10 ***
Child's Age 0.33 *** 1.39 0.31 *** 1.36
Male
a
0.73 *** 2.07 0.78 *** 2.18
Unwedded Mothers
b
0.16 1.18 0.19 1.21
Divorced/Separated Mothers
b
0.01 1.01 0.05 1.05
Stepparent
b
0.42 1.52 0.42 1.52
Mother's Education -0.13 0.88 -0.13 0.88
Household Income (ln) -0.21 0.81 -0.19 0.82
Missing Household Income -0.19 0.83 -0.19 0.82
Family Transitions 0.13 1.14 0.12 1.13
Mother-Child Activities -0.23 ** 0.79 -0.25 ** 0.78
Mother-Child Discussions 0.26 *** 1.29 0.25 *** 1.28
Parent Strictness -0.20 ** 0.82 -0.20 *** 0.82
Parent Meal Presence -0.06 0.94 -0.06 0.94
Full Time Employment 0.16 1.18 0.10 1.11
Mother's Morning Presence -0.09 0.92 -0.09 0.91
Mother's Afternoon Presence -0.02 0.98 -0.02 0.98
Mother's Night Presence 0.05 1.05 0.03 1.03
Mother's Religiosity 0.06 1.06
Sex Communication 0.26 *** 1.29
Disapprove Sex -0.02 0.98
-2 LL 1009.89 1001.15
Chi-Square 198.37 *** 207.47 ***
Nagelkerke R
2
0.27 0.28
BIC -83.27 -72.05
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= females as the reference group;
b
= two biological parents as the reference group;
232
Table 9-5: Continued
Model 5 Model 6
Beta Exp (B) Beta Exp (B)
Intercept -3.36 * -4.16 **
Child's Age 0.31 *** 1.36 0.26 *** ^ 1.29
Male
a
0.86 *** 2.36 0.72 *** ^ 2.06
Unwedded Mothers
b
0.20 1.22 0.28 1.32
Divorced/Separated Mothers
b
0.03 1.03 0.09 1.09
Stepparent
b
0.33 1.39 0.44 1.56
Mother's Education -0.14 0.87 -0.14 0.87
Household Income (ln) -0.21 0.81 -0.20 0.82
Missing Household Income -0.14 0.87 -0.13 0.88
Family Transitions 0.12 1.13 0.12 1.13
Mother-Child Activities -0.21 ** 0.81 -0.18 * 0.84
Mother-Child Discussions 0.25 *** 1.28 0.22 *** 1.24
Parent Strictness -0.20 ** 0.82 -0.20 ** 0.82
Parent Meal Presence -0.03 0.97 -0.03 0.97
Full Time Employment 0.13 1.14 0.25 1.29
Mother's Morning Presence -0.09 0.92 -0.08 0.92
Mother's Afternoon Presence -0.02 1.00 0.01 1.01
Mother's Night Presence 0.05 1.05 0.10 1.11
Mother's Religiosity 0.06 1.06 0.09 ^ 1.10
Sex Communication 0.27 ** 1.31 0.26 ** 1.30
Disapprove Sex -0.02 0.98 0.01 ^ 1.00
Family's Support -0.13 *** 0.88 -0.10 ** 0.91
Deviant Peers 0.25 *** ^ 1.28
-2 LL 983.66 938.98
Chi-Square 224.96 *** 269.65 ***
Nagelkerke R
2
0.30 0.36
BIC -82.77 -120.69
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= females as the reference group;
b
= two biological parents as the reference group; ^ = race difference
233
Table 9-6: Hierarchically Ordered Regression of Adolescent Sexual Activity on the
Demographic, Family Influence, and Peer Influence Variables for White
Males (N = 1242)
Model 1 Model 2
Beta Exp (B) Beta Exp (B)
Intercept -8.98 *** -9.67 ***
Child's Age 0.58 *** 1.79 0.60 *** 1.82
Unwedded Mothers
a
1.50 * 4.48 1.28 * 3.60
Divorced/Separated Mothers
a
0.69 *** 1.98 0.38 1.46
Stepparent
a
0.31 1.37 -0.17 0.84
Mother's Education -0.27 *** 0.76 -0.26 *** 0.77
Household Income (ln) -0.18 0.84 -0.19 0.83
Missing Household Income 0.18 1.19 0.18 1.19
Family Transitions 0.43 ** 1.54
-2 LL 1317.59 1306.67
Chi-Square 257.91 *** 268.83 ***
Nagelkerke R
2
0.26 0.27
BIC -208.04 -211.83
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= two biological parents as the
reference group
234
Table 9-6: Continued
Model 3 Model 4
Beta Exp (B) Beta Exp (B)
Intercept -6.05 *** -5.44 ***
Child's Age 0.49 *** 1.63 0.46 *** 1.59
Unwedded Mothers
a
1.03 2.80 0.75 2.12
Divorced/Separated Mothers
a
0.15 1.16 0.02 1.02
Stepparent
a
-0.32 0.73 -0.39 0.68
Mother's Education -0.28 *** 0.76 -0.25 ** 0.78
Household Income (ln) -0.26 * 0.77 -0.26 * 0.77
Missing Household Income 0.20 1.22 0.15 1.16
Family Transitions 0.43 ** 1.54 0.37 ** 1.44
Mother-Child Activities -0.33 *** 0.72 -0.30 *** 0.74
Mother-Child Discussions 0.26 *** 1.29 0.21 *** 1.24
Parent Strictness -0.15 * 0.86 -0.13 * 0.87
Parent Meal Presence -0.04 0.96 -0.05 0.95
Full Time Employment -0.10 0.90 -0.12 0.88
Mother's Morning Presence 0.01 1.01 0.02 1.02
Mother's Afternoon Presence -0.09 0.91 -0.10 0.90
Mother's Night Presence -0.21 * 0.81 -0.18 0.83
Mother's Religiosity -0.06 0.94
Sex Communication 0.40 *** 1.49
Disapprove Sex -0.22 ** 0.80
-2 LL 1246.63 1211.63
Chi-Square 328.87 *** 363.87 ***
Nagelkerke R
2
0.32 0.35
BIC -214.88 -228.50
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= two biological parents as the
reference group
235
Table 9-6: Continued
Model 5 Model 6
Beta Exp (B) Beta Exp (B)
Intercept -4.60 *** -6.23 ***
Child's Age 0.46 *** 1.59 0.43 *** 1.53
Unwedded Mothers
a
0.73 2.07 0.34 1.41
Divorced/Separated Mothers
a
0.00 1.01 0.04 1.04
Stepparent
a
-0.43 0.65 -0.34 0.71
Mother's Education -0.25 ** 0.78 -0.27 ** 0.76
Household Income (ln) -0.26 * 0.77 -0.27 * 0.76
Missing Household Income 0.12 1.13 0.09 1.10
Family Transitions 0.37 ** 1.45 0.27 1.31
Mother-Child Activities -0.25 ** 0.77 -0.12 0.89
Mother-Child Discussions 0.21 *** 1.24 0.16 ** 1.18
Parent Strictness -0.14 * 0.87 -0.08 0.92
Parent Meal Presence -0.03 0.97 0.00 1.00
Full Time Employment -0.10 0.91 -0.05 0.95
Mother's Morning Presence 0.03 1.03 0.01 1.01
Mother's Afternoon Presence -0.09 0.91 -0.05 0.95
Mother's Night Presence -0.14 0.87 -0.11 0.90
Mother's Religiosity -0.06 0.94 -0.06 0.95
Sex Communication 0.41 *** 1.51 0.38 *** 1.47
Disapprove Sex -0.22 ** 0.80 -0.19 * 0.83
Family's Support -0.08 ** 0.92 -0.04 0.96
Deviant Peers 0.39 *** 1.48
-2 LL 1203.90 1045.99
Chi-Square 371.60 *** 529.50 ***
Nagelkerke R
2
0.36 0.48
BIC -229.11 -379.89
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= two biological parents as the
reference group
236
interpreted as a sign of support for these males. The reason why the deviant peer
variable provides the biggest contribution to the mediation of the mother-child
activity variable may be because of the overwhelming influence of peers on an
adolescent’s life when their mothers do not build a relationship with their sons.
Table 9-7 displays the results of the hierarchically ordered regression analysis
for black males. The influence of peers on adolescent sexual activity is significantly
stronger for white males than for black males. This analysis does not reveal an
eventual mediation for any of the variables that are significant in the previous
regression models. Instead, the analysis reveals a mediation of child’s age in the full
regression model. Thus, child’s age is associated with greater adolescent sexual
activity risk because older adolescents may be more likely to experience supervision
by the parents and support by family members since parents may feel that older
children should have more responsibility over their lives. In addition, mothers of
older adolescents may also be more tolerant of adolescent sexual activity and that
would allow these adolescents to have greater opportunities to be acquainted with
peers who engage in deviant activities as they grow older.
Table 9-8 displays the results of the hierarchically ordered regression analysis
for white females. Significant gender differences are found from white males for the
parental meal presence, mother’s disapproval of adolescent sexual activity, and the
peer influence variables. This analysis reveals an eventual mediation of the family
transition, mother-child activities, parental strictness, and mother’s religiosity
variables. Unlike white males, the mediation of the family transition effect does not
require the use of all the explanations for white females. In addition, both the
237
Table 9-7: Hierarchically Ordered Regression of Adolescent Sexual Activity on the
Demographic, Family Influence, and Peer Influence Variables for Black
Males (N = 396)
Model 1 Model 2
Beta Exp (B) Beta Exp (B)
Intercept -4.90 *** -5.22 ***
Child's Age 0.34 *** 1.40 0.34 *** 1.41
Unwedded Mothers
a
0.63 1.87 0.64 1.89
Divorced/Separated Mothers
a
0.39 1.47 0.32 1.38
Stepparent
a
0.71 * 2.03 0.58 1.79
Mother's Education -0.08 0.92 -0.08 0.92
Household Income (ln) -0.07 0.94 -0.05 0.95
Missing Household Income -0.04 0.96 -0.03 0.98
Family Transitions 0.17 1.18
-2 LL 494.97 493.36
Chi-Square 39.33 *** 40.93 ***
Nagelkerke R
2
0.13 0.13
BIC 2.54 6.92
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= two biological parents as the
reference group
238
Table 9-7: Continued
Model 3 Model 4
Beta Exp (B) Beta Exp (B)
Intercept -3.48 *** -3.72
Child's Age 0.26 1.30 0.22 ** 1.25
Unwedded Mothers
a
0.38 1.47 0.41 1.51
Divorced/Separated Mothers
a
0.06 1.07 0.09 1.09
Stepparent
a
0.43 1.53 0.43 1.54
Mother's Education -0.12 0.89 -0.12 0.89
Household Income (ln) -0.23 0.79 -0.21 0.81
Missing Household Income -0.12 0.89 -0.17 0.85
Family Transitions 0.08 1.09 0.05 1.05
Mother-Child Activities -0.22 0.80 -0.25 * 0.78
Mother-Child Discussions 0.30 ** 1.34 0.29 ** 1.34
Parent Strictness -0.18 * 0.83 -0.20 * 0.82
Parent Meal Presence -0.02 0.98 -0.02 0.98
Full Time Employment 0.51 1.66 0.39 1.48
Mother's Morning Presence -0.07 0.93 -0.10 0.90
Mother's Afternoon Presence -0.01 1.00 -0.01 0.97
Mother's Night Presence 0.12 1.13 0.10 1.10
Mother's Religiosity 0.12 1.13
Sex Communication 0.26 1.29
Disapprove Sex -0.11 0.90
-2 LL 467.50 482.12
Chi-Square 66.80 *** 73.46 ***
Nagelkerke R
2
0.21 0.23
BIC 28.90 40.19
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= two biological parents as the
reference group
239
Table 9-7: Continued
Model 5 Model 6
Beta Exp (B) Beta Exp (B)
Intercept -1.78 -2.38
Child's Age 0.21 ** 1.23 0.13 1.13
Unwedded Mothers
a
0.45 1.57 0.50 1.65
Divorced/Separated Mothers
a
0.04 1.04 0.06 1.06
Stepparent
a
0.35 1.41 0.44 1.55
Mother's Education -0.14 0.87 -0.16 0.85
Household Income (ln) -0.21 0.81 -0.20 0.82
Missing Household Income -0.12 0.89 -0.03 0.97
Family Transitions 0.04 1.04 0.07 1.07
Mother-Child Activities -0.21 0.81 -0.19 0.83
Mother-Child Discussions 0.30 ** 1.34 0.26 ** 1.30
Parent Strictness -0.20 * 0.82 -0.22 * 0.80
Parent Meal Presence 0.00 1.00 0.00 1.00
Full Time Employment 0.41 1.51 0.53 1.71
Mother's Morning Presence -0.09 0.92 -0.08 0.92
Mother's Afternoon Presence 0.01 1.01 0.02 1.02
Mother's Night Presence 0.14 1.15 0.21 1.23
Mother's Religiosity 0.11 1.12 0.15 1.16
Sex Communication 0.25 1.29 0.26 1.30
Disapprove Sex -0.12 0.89 -0.06 0.94
Family's Support -0.12 * 0.89 -0.09 0.91
Deviant Peers 0.25 *** (r) 1.29
-2 LL 454.01 430.17
Chi-Square 80.28 *** 104.12 ***
Nagelkerke R
2
0.25 0.31
BIC 39.35 21.49
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= two biological parents as the
reference group; (r) = race difference
240
Table 9-8: Hierarchically Ordered Regression of Adolescent Sexual Activity on the
Demographic, Family Influence, and Peer Influence Variables for White
Females (N = 1372)
Model 1 Model 2
Beta Exp (B) Beta Exp (B)
Intercept -12.32 *** -12.90 ***
Child's Age 0.75 *** 2.11 0.76 *** 2.14
Unwedded Mothers
a
1.68 *** 5.39 1.51 ** 4.54
Divorced/Separated Mothers
a
0.96 *** 2.62 0.76 *** 2.14
Stepparent
a
0.82 *** 2.27 0.48 * 1.61
Mother's Education -0.09 0.91 -0.07 0.93
Household Income (ln) -0.16 0.85 -0.15 0.86
Missing Household Income -0.05 0.95 -0.04 0.96
Family Transitions 0.32 ** 1.37
-2 LL 1332.75 1324.94
Chi-Square 391.81 *** 399.63 ***
Nagelkerke R
2
0.35 0.35
BIC -341.24 -341.84
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= two biological parents as the
reference group
241
Table 9-8: Continued
Model 3 Model 4
Beta Exp (B) Beta Exp (B)
Intercept -8.92 *** -7.73 ***
Child's Age 0.64 *** 1.90 0.62 *** 1.86
Unwedded Mothers
a
1.09 * 2.97 1.08 * 2.95
Divorced/Separated Mothers
a
0.44 * 1.55 0.34 1.40
Stepparent
a
0.37 1.45 0.32 1.37
Mother's Education -0.10 0.91 -0.06 0.94
Household Income (ln) -0.25 * 0.78 -0.26 * 0.77
Missing Household Income 0.02 1.02 -0.03 0.97
Family Transitions 0.28 * 1.32 0.22 1.24
Mother-Child Activities -0.33 *** 0.72 -0.22 ** 0.80
Mother-Child Discussions 0.25 *** 1.29 0.23 *** 1.26
Parent Strictness -0.18 * 0.84 -0.19 * 0.83
Parent Meal Presence -0.12 *** 0.88 -0.14 *** 0.87
Full Time Employment 0.16 1.17 0.16 1.17
Mother's Morning Presence -0.09 0.92 -0.10 0.91
Mother's Afternoon Presence -0.04 0.96 -0.04 0.96
Mother's Night Presence -0.12 0.89 -0.08 0.92
Mother's Religiosity -0.09 * 0.91
Sex Communication 0.38 *** 1.46
Disapprove Sex -0.38 *** 0.69
-2 LL 1240.83 1185.56
Chi-Square 483.73 *** 539.00 ***
Nagelkerke R
2
0.42 0.45
BIC -368.15 -401.74
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= two biological parents as the
reference group
242
Table 9-8: Continued
Model 5 Model 6
Beta Exp (B) Beta Exp (B)
Intercept -6.47 *** -8.35 ***
Child's Age 0.64 *** 1.89 0.63 *** 1.88
Unwedded Mothers
a
1.03 2.81 0.85 2.34
Divorced/Separated Mothers
a
0.29 1.34 0.26 1.29
Stepparent
a
0.29 1.33 0.27 1.31
Mother's Education -0.06 0.94 -0.02 0.98
Household Income (ln) -0.24 * 0.78 -0.20 0.82
Missing Household Income -0.01 0.99 -0.02 0.98
Family Transitions 0.20 1.22 0.13 1.14
Mother-Child Activities -0.14 0.87 -0.08 0.92
Mother-Child Discussions 0.24 *** 1.28 0.17 ** 1.19
Parent Strictness -0.20 ** 0.82 -0.15 0.86
Parent Meal Presence -0.11 ** 0.90 -0.09 ** (g) 0.91
Full Time Employment 0.22 1.25 0.16 1.17
Mother's Morning Presence -0.08 0.92 -0.04 0.96
Mother's Afternoon Presence -0.02 0.98 -0.01 0.99
Mother's Night Presence -0.07 0.93 -0.07 0.93
Mother's Religiosity -0.10 * 0.90 -0.07 0.93
Sex Communication 0.41 *** 1.51 0.38 *** 1.46
Disapprove Sex -0.38 *** 0.68 -0.40 *** (g) 0.67
Family's Support -0.13 *** 0.88 -0.09 ** 0.92
Deviant Peers 0.30 *** (g) 1.35
-2 LL 1165.39 1065.50
Chi-Square 559.17 *** 659.06 ***
Nagelkerke R
2
0.47 0.53
BIC -414.69 -507.36
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= two biological parents as the
reference group; (g) = gender difference
243
parental attitude and the family support measures are the most responsible for
mediating the mother-child activity effect. Thus, it seems, for white females, that
mothers with conservative attitudes may have a stronger bond with their daughters
and that these activities help to create feelings of support by family members.
Peer influence is the only measure that mediates the parental strictness effect.
While parental strictness discourages sexual activity, this finding seems to suggest
that the presence of deviant peers in these daughter’s lives may be an indication of
low parental strictness in the household. In addition, deviant peers may also have a
greater influence than their parents over their daughter’s behaviors.
This analysis also reveals that the mediation of the mother’s religiosity
variable occurs only after accounting for the peer influence measure. The finding
suggests that the influence of deviant peers in a white female’s life may overcome
the influence of conservative sexual values as prescribed by the mother’s religion.
Table 9-9 displays the results of the hierarchically ordered regression analysis
for black females. There is a significant race difference from white females
regarding mother’s disapproval of adolescent sexual activity. However, there is no
significant gender difference from black males for any of these measures. The
analysis reveals an eventual mediation of the mother-child activities and parental
meal presence variables after controlling for the family support measure. This
finding suggests that, for these females, mother-child activities and parental meal
presence are indications of support from parents and family members.
244
Table 9-9: Hierarchically Ordered Regression of Adolescent Sexual Activity on the
Demographic, Family Influence, and Peer Influence Variables for Black
Females (N = 476)
Model 1 Model 2
Beta Exp (B) Beta Exp (B)
Intercept -7.15 *** -7.55 ***
Child's Age 0.48 *** 1.61 0.49 *** 1.63
Unwedded Mothers
a
0.11 1.11 0.06 1.06
Divorced/Separated Mothers
a
0.26 1.29 0.15 1.17
Stepparent
a
0.70 * 2.02 0.50 1.65
Mother's Education -0.10 0.90 -0.10 0.90
Household Income (ln) -0.17 0.84 -0.18 0.84
Missing Household Income -0.17 0.84 -0.11 0.89
Family Transitions 0.29 1.34
-2 LL 567.70 563.81
Chi-Square 84.07 *** 87.97 ***
Nagelkerke R
2
0.22 0.23
BIC -40.91 -38.65
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= two biological parents as the
reference group
245
Table 9-9: Continued
Model 3 Model 4
Beta Exp (B) Beta Exp (B)
Intercept -4.89 ** -5.90 **
Child's Age 0.41 *** 1.51 0.40 *** 1.50
Unwedded Mothers
a
0.00 1.00 -0.01 0.99
Divorced/Separated Mothers
a
-0.06 0.94 -0.06 0.94
Stepparent
a
0.44 1.56 0.43 1.53
Mother's Education -0.12 0.89 -0.11 0.89
Household Income (ln) -0.16 0.85 -0.16 0.85
Missing Household Income -0.14 0.85 -0.10 0.91
Family Transitions 0.21 1.24 0.21 1.24
Mother-Child Activities -0.26 ** 0.77 -0.26 * 0.77
Mother-Child Discussions 0.22 ** 1.25 0.21 * 1.23
Parent Strictness -0.22 ** 0.80 -0.22 ** 0.80
Parent Meal Presence -0.09 * 0.91 -0.09 * 0.91
Full Time Employment -0.17 0.85 -0.18 0.84
Mother's Morning Presence -0.10 0.90 -0.09 0.91
Mother's Afternoon Presence -0.02 0.98 -0.03 0.97
Mother's Night Presence -0.03 0.97 -0.04 0.96
Mother's Religiosity 0.01 1.01
Sex Communication 0.29 * 1.33
Disapprove Sex 0.03 1.03
-2 LL 531.83 527.46
Chi-Square 119.95 *** 124.32 ***
Nagelkerke R
2
0.30 0.31
BIC -15.14 -7.18
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= two biological parents as the
reference group
246
Table 9-9: Continued
Model 5 Model 6
Beta Exp (B) Beta Exp (B)
Intercept -4.24 * -5.38 **
Child's Age 0.42 *** 1.52 0.39 *** 1.47
Unwedded Mothers
a
-0.03 0.97 0.06 1.06
Divorced/Separated Mothers
a
-0.06 0.94 0.05 1.05
Stepparent
a
0.33 1.39 0.45 1.58
Mother's Education -0.12 0.89 -0.10 0.91
Household Income (ln) -0.18 0.83 -0.18 0.83
Missing Household Income 0.02 1.02 -0.03 0.97
Family Transitions 0.24 1.27 0.20 1.22
Mother-Child Activities -0.22 * 0.80 -0.19 0.83
Mother-Child Discussions 0.19 * 1.21 0.17 1.18
Parent Strictness -0.21 * 0.81 -0.19 * 0.83
Parent Meal Presence -0.05 0.95 -0.05 0.95
Full Time Employment -0.16 0.85 -0.05 0.95
Mother's Morning Presence -0.10 0.91 -0.09 0.91
Mother's Afternoon Presence -0.02 0.98 -0.01 0.99
Mother's Night Presence -0.03 0.97 0.02 1.02
Mother's Religiosity 0.02 1.02 0.05 1.06
Sex Communication 0.34 * 1.40 0.30 * 1.35
Disapprove Sex 0.02 1.02 0.03 (r) 1.03
Family's Support -0.14 ** 0.87 -0.11 * 0.89
Deviant Peers 0.26 *** 1.30
-2 LL 515.98 494.10
Chi-Square 135.80 *** 157.68 ***
Nagelkerke R
2
0.33 0.38
BIC -12.49 -28.21
Notes: * = p < .05; ** = p < .01; *** = p < .001;
a
= two biological parents as the
reference group; (r) = race difference
247
Discussion
This investigation analyzed the relationship between the five explanations
and adolescent sexual activity using hierarchically ordered regression analyses where
the measures are regressed in a stepwise manner. The main reason for this analysis
was to determine which measures are directly and indirectly associated with
adolescent sexual activity after all of them are jointly considered. Several
researchers have suggested that some of the family influence measures may be
associated with one another (Newcomer & Udry, 1987; Wu & Thomson, 2001).
However, none of the studies in the family literature has truly tested this premise.
This investigation was conducted on the total sample, by adolescent’s gender,
by adolescent’s race, and by adolescent’s race/gender subgroups. However, this
investigation excluded Hispanics and Asians from both the race and race-gender
subgroup analyses due to their small samples. In addition, these investigations
examined whether there are significant gender and racial differences regarding the
relationship between both the family influence and deviant peer measures and
adolescent sexual activity by gender, race, and race-gender subgroups.
White adolescents of both genders were the only subgroups that revealed
indirect relationships between the family influence measures and adolescent sexual
activity. Family transitions, mother’s religiosity, mother-child activities, and
parental strictness were found to be indirectly associated with adolescent sexual
activity for white adolescents of both genders. Although the full model revealed a
mediation of the family transition effect for both white males and females, the
248
explanation for this effect seem to differ by gender. For white females, only one
other explanation, in this case parental supervision, was needed to mediate this
effect. However, for white males, all of the family influence measures including
peer influence were required to mediate this effect. A conclusion that can be reached
from this finding is that the family transitions effect may be multidimensional for
white males because the mediation of this effect required the use of all the
explanations. This conclusion seems to be supported by Newcomer & Udry (1987)
who found that the family transition event affected white children who are males
much more than white children who are females.
However, in contrast to their study where it was speculated that the family
transition event affected males more than females because of lack of parental
support, the results of this investigation revealed that other family influences could
have greater responsibility for this effect. When examining the results of white
males more closely, the greatest drop in the beta of the family transition variable
occurred when controlling for the parental attitude and the peer influence measures.
These findings would suggest that the combination of mothers transmitting less
conservative sexual attitudes to their sons and the presence of peers who engage in
deviant activities in the adolescent’s life rather than lack of parental support that
would bear the most responsible for the family transition effect on adolescent sexual
activity for white males.
For white females, mother’s religiosity was indirectly associated with
adolescent sexual activity. This process occurred after taking peer influence into
account, which suggests that the influences of friends and peers may overwhelm the
249
conservative sexual values as prescribed by the mother’s religion. However,
mother’s religiosity was not significant for adolescent sexual activity for white males
when accounting for the mother’s attitude measure. Thus, for white adolescents,
mother’s religiosity seems to have the weakest effect on sexual activity among the
parental attitude variables.
The reason for the mother-child activity effect for white adolescents seems to
differ by gender. For white females, the mother-child activity effect on adolescent
sexual activity was partially mediated by the parental attitude measure. Adding the
family support variable to the model fully mediated the mother-child activity effect.
However, for white males, neither the parental attitude nor the family support
measures were able to partially mediate the mother-child activity effect. Instead,
peer influence fully explained this effect. This finding suggests that, for white
females, mothers with conservative attitudes may be engaging in more activities with
their daughters and that these activities are one indication of support by family
members. In contrast, for white males, mother-child activity seems to be more of a
social control issue. Peer influence being able to fully account for the mother-child
activity effect for white males seems to suggest that friends and peers may have
more influence over these adolescents because their mothers are not closely
monitoring their activities.
In another situation, peer influence was found to mediate the parental
strictness effect for white adolescents of both genders. This finding suggests that
households with low parental strictness levels may allow adolescents to have greater
opportunity to be acquainted with friends who engage in deviant activities.
250
Among white adolescents, gender differences were also found for some of
the variables that are directly associated with adolescent sexual activity. Parental
meal presence was negatively associated with sexual activity for white females but
not for white males. The finding suggests that these occasions may be opportunities
for the parents to strengthen the relationship with their daughters. In addition, the
effect of mother’s disapproval of adolescent sex on adolescent sexual activity was
stronger on the daughters compared to her sons. This gender difference is primarily
due to mothers being more disapproving of her daughter sexual activity than her sons
because she may become more susceptible to the consequences of sexual behaviors
such as getting pregnant. Apparently, these mothers may fear that their daughters are
more susceptible than their sons to these types of influences so they become much
more disapproving of their daughter’s sexual activity. Furthermore, the relationship
between deviant peer influence and adolescent sexual activity was found to be
significantly stronger for white males than for white females. This finding suggests
that family influences have a greater blunting effect on peer influences for white
females than for white males.
Racial differences were found between white and black adolescents for some
of the family and peer influence variables regarding sexual behaviors. Mother’s
disapproval of adolescent sex had a stronger negative relationship with sexual
activity for white females than for black females. As mentioned in the previous
investigation, this result could be due to the fact that mothers of black females tend
to have more liberal sexual values than mothers of white females.
251
There were racial differences found between white males and black males
regarding the peer influence effect. As mentioned in the previous investigation, this
finding suggests that friends and peers seem to have more influence over a white
male’s decision to engage in sexual activity in comparison to black males even
though the number of friends and peers that engage in these behaviors are
statistically similar to each other.
Chapter Summary
The premise that the family and peer influence measures may be associated
with one another as proposed by Newcomer & Udry (1987) and Wu & Thomson
(2001) was somewhat supported in this study. White adolescents were the only
subgroup where the family influence variables are found to be indirectly associated
with adolescent sexual activity. The variables that are indirectly associated with
adolescent sexual activity include family transition, mother-child activity, and
mother’s religiosity. However, there were gender differences regarding the reasons
why these variables were indirectly associated with adolescent sexual activity.
The factors that are positively directly associated with sexual activity for
white and black adolescents include mother-child discussions, mother-child sex
discussions, and peer influence. Meanwhile, the factors that are negatively directly
associated with sexual activity for these adolescents are mother’s disapproval of
sexual activity and family support. In addition, various race and gender differences
252
were found for the relationship between these variables and adolescent sexual
activity.
In essence, this investigation has shown that only a few family influence
variables are being explained by other family influence measures. However, this
finding may be culturally specific because these indirect relationships were only
found for white adolescents. In addition, the reasons for some of these effects may
be gender specific. Thus, this investigation has shown that the relationships between
the five explanations and adolescent sexual activity are highly contexualized by both
race and gender.
253
Conclusion
The percentage of sexually experienced adolescents living in the United
States increased dramatically from the early 1970s to the mid 1990s (Moore, 1992;
Parrillo, Felts, & Mikow-Porto, 1997). A second major trend in the same period was
the increase in children who do not reside with two biological parents (U.S. Census,
2003). Both of these trends have been a concern to adults such as parents, educators,
healthcare professionals, and counselors because numerous studies have shown that
children who do not reside with two biological parents are more likely to have
engaged in sexual activity before reaching young adulthood. Adolescent sexual
activity has been associated with consequences such as acquiring a sexually
transmitted disease or AIDS for adolescents of both genders which can jeopardize
one’s health (Hayes, 1987), early pregnancy for females (Hayes, 1987; Brooks-Gunn
& Furstenberg, 1989), and early fatherhood for males (Sonnenstein, 1986). These
behaviors also have economic consequences that can diminish the prospect of
productive adulthood such as receiving fewer employment opportunities and lower
wages throughout their employment history for adolescents who become pregnant or
experience early fatherhood (Card & Wise, 1978; Hofferth & Moore, 1979).
This study conducted two major investigations. The first investigation
attempted to determine the best reason for the family structure effect among the three
non-traditional households using four family influence explanations and a peer
explanation. This goal was achieved by regressing each explanation individually
rather than simultaneously or in a stepwise manner. There were several reasons for
254
analyzing the data using this method such as being able to determine if there may be
subtle differences among the various non-traditional households and to avoid the
casual order issues that one may need to address when investigating the explanations
in a stepwise manner.
Both of these investigations were conducted on the 1994-1995 National
Study of Adolescent Health Wave I dataset. This study examined white, black,
Hispanic, and Asian adolescents and their biological mothers. These investigations
were conducted on the total sample and within the adolescent’s gender, the
adolescent’s race, and the adolescent’s race-gender subgroup categories.
This chapter is divided into several sections. The first section discusses the
answers to the questions that were posed in the introduction. The second section
discusses some of the significant findings from this study for the non-traditional
household -- adolescent sexual activity relationship. This section also discusses
some of the significant findings for the relationship between both the family and peer
influence variables and adolescent sexual activity. The third section discusses the
limitations of this study and issues for further research. The fourth section discusses
related research, social policy, and counseling implications of this study.
Answers to the Questions Posed in the Introduction
Question 1: Does the risk for adolescent sexual activity differ across the three non-
traditional household configurations?
255
There are several ways in which this question can be answered. The first way
would be to determine whether adolescent sexual activity rates differ across the non-
traditional households without controlling for the demographic variables. The
second way would be to determine whether adolescent sexual activity risk differs
across the non-traditional households when considering these demographic variables.
The answer to both of these questions would be a “yes” but with some qualifications.
Adolescents who reside in the non-traditional households were more likely
than those who reside with two biological parents to engage in sexual activity for the
total sample in both situations. However, this result was not necessarily the outcome
when the analyses were contexualized within the adolescent’s race-gender subgroup
categories. There were several situations where risks of sexual activity for
adolescents who reside in non-traditional households were no different from those
who reside with two biological parents when controlling for the demographic
variables. The first situation was the stepparent households for white males and
Latinos. The second situation was the unwedded mother and divorced/separated
mother households for black females.
The stepparent household effect for white males and Latinos was much less
pronounced compared to this effect on white females and Latinas respectively. This
finding seems to imply that the new parental figure presence in stepparent
households may provide additional stability to the former one-parent household.
Since the new parental figure in this situation would be a male, the gender similarity
between the stepparent and the stepchild may encourage a close relationship
relatively quickly and help to mitigate the possible neglected feeling that the male
256
children may experience during a family transition event (Newcomer & Udry, 1987).
Stepfathers could be someone who can provide support to their stepsons by
answering questions, giving advice, and being there for him during difficult times.
In addition, stepfathers can also be role models to them by demonstrating acceptable
behaviors. Both of these type of actions may result in white males and Latinos
perceiving that the stepparent household family environment as similar to those of
two biological parent households.
Both the unwedded mother and divorced/separated mother effects were much
less significant for black females than for white females after considering the
demographic variables. However, this finding was not found for black males
because of the single parent effect for white adolescents. As mentioned in chapter 8,
the reason for the race-gender differences among black adolescents may be due to
their position in the hierarchy regarding this relationship relative to white
adolescents. For example, the single parent effect for white females is stronger but
not significantly different from the single parent effect for white males while these
effects for black females are weaker but not significantly different from the effects
for black males. Meanwhile, the single parent effect for black males is similar to
those of white males while there is a much greater difference for this effect between
white females and black females. The result is a statistical racial difference for the
single parent effect between white and black females while there is no statistical
racial difference for this effect between white and black males.
257
Question 2: Which of the five influences is the best reason for the family structure
and adolescent sexual activity relationship for the three non-traditional households?
For the total sample, parental attitude was the best reason for the unwedded
mother effect. Meanwhile, parental supervision was the best reason for the
relationship between divorced/separated mother household and adolescent sexual
activity. Finally, family transition was the best reason for the stepparent --
adolescent sexual activity relationship. Analyzing the data within gender categories
also revealed similar findings. In another words, tolerant attitudes in regards to
adolescent sexual activity may be the best reason for the unwedded mother
household effect, the lack of a parental presence at home or lower levels of parent-
child interactions may be the best reason for the divorced/separated mother
household effect, and the stresses of adjusting to a new family form may be the best
reason for the stepparent household effect.
The best reason for the non-traditional household and adolescent sexual
activity relationship did differ within race categories. For stepparent household
effect, family transition was the best reason for white adolescents while parental
supervision was the best reason for black adolescents. For white adolescents, the
stepparent effect is probably due to these children feeling tremendous stress trying to
adjust to the new family configuration. This adjustment process may contribute to
the feeling of a disorganized household, which would then affect the interaction
quality between the parent and his or her children (Wu & Martinson, 1993).
Children who are going through these events may engage in impulsive, rebellious,
and aggressive actions or spend less time at home (Wallerstein & Kelly, 1980;
258
Wallerstein & Blakeslee, 1989). In contrast, black adolescents who live with a
stepparent may not have gone through multiple transitions because the mother’s
marriage to a stepparent may be more like a first marriage than a remarriage due to
the non-marital birth rates tending to be higher for low-income individuals and
African American families compared to the general population (Taylor et al., 1997).
While black adolescents who reside with a stepparent may feel relatively more
advantaged than those who reside in one-parent households (Moore, 2001), it still
could be difficult for them to adjust to the new family environment because the
stepparent may be seen as a “relative stranger” (Beer, 1988). This difficulty can lead
to considerable conflict within the household and the stepparent feeling that their
role as a parent is ambiguous and being less strict during attempts to discipline their
stepchildren (Cherlin, 1978; Amato, 1987; Furstenberg, 1987).
Parental supervision was the best reason for the divorced/separated mother
household effect while peer influence was the best reason for the unwedded mother
household effect for white adolescents. For the divorced/separated mother
households, this finding suggests that divorced/separated mothers may not spend
enough time to look after or bond with their children. It is possible that lack of
mother’s presence at home could make her children feel unsupported and become
more influenced by their peers. The reason why peer influence was the best
explanation for the unwedded mother household effect for white adolescents is
probably because of the neighborhood environment where unwedded mothers are
more likely than divorced/separated mothers to reside in poorer neighborhoods for a
longer period. Since, these adolescents may have grown up in this type of
259
neighborhood for their whole life, they may have greater exposure to values that
encourage deviant behaviors since poorer neighborhoods are associated with
antisocial values (Wilson, 1987).
Parental attitude was the best reason for the unwedded mother and the
divorced/separated mother household effects for Hispanic adolescents. The finding
seems to confirm the thesis by Thornton (1991) that unwedded mothers may
influence their children by being more liberal in their sexual views in comparison to
mothers of two biological parent households. In addition, this finding suggests that
familism (Davis & Harris, 1982; Padilla & O’Grady, 1997) may be weaker in one-
parent households than in two parent households for Hispanic families.
There are further differences by the adolescent’s race-gender subgroup
regarding the explanations that best mediate the relationship between the three non-
traditional households and adolescent sexual activity. For white males, peer
influence and family transition were the best explanations for the unwedded mother
and divorced/separated mother household effects respectively. Several reasons were
given as to why peer influence might best account for the unwedded mother
household effect such as a lack of an adult male presence around the household who
can socialize with or be a role model to these children and the neighborhood
environment. In addition, family transition was the best explanation for the
divorced/separated household -- adolescent sexual activity relationship for white
males. This finding suggests that these children may either have difficulty with the
family environment during the family transition period (Newcomer & Udry, 1987) or
260
that they are more likely to follow their parents dating behaviors (Thornton &
Camburn, 1987).
For white females, none of the explanations was able to at least partially
explain the unwedded mother and the divorced/separated mother household effects.
However, the parental supervision explanation seemed to be the best reason for the
single parent household effects. The amount of time that the mother spends with her
daughter may be important for those who reside in one-parent households because a
lack of it may directly lead to sexual activity during adolescence.
Parental supervision was also the best reason for the stepparent household
effect for black males. This finding may be due to mothers having greater
discussions with their sons about their behaviors and parents not eating meals as
often compared to those who reside with two biological parents. Although those
who reside with a stepparent may be relatively more advantaged than those who
reside in one-parent households (Moore, 2001), they may still have difficultly
adjusting to the new family environment.
For black females, family transition was the best reason for the stepparent
household effect. This effect seems to be due to parental dating instead of difficulty
in adjusting to a new parental figure or mothers having relatively tolerant attitudes in
regards to adolescent sexual activity. It could be that black females who reside with
a stepparent may have previously observed their mother’s dating behaviors and
concluded that non-marital sexual activity is acceptable (Whitbeck, Simons, & Kao,
1994).
261
Acquaintance with deviant peers was the best reason for the
divorced/separated household effect for Latinos. This reason for this finding may be
similar to those for white males. However, for Latinas, mother’s attitude is the best
reason for both the unwedded mother and divorced/separated mother household
effects. Although the Hispanic culture tends to advocate conservative attitudes
regarding sexuality due to the culture of familism where the individual needs has less
priority compared to family ties and loyalty (Mirande, 1977), this finding suggests
that familism may be weaker in one-parent Hispanic households than in two
biological parent Hispanic households.
Overall, these findings demonstrated that the best reasons for the family
structure and adolescent sexual activity relationship for several non-traditional
households varies tremendously within the adolescent’s race-gender contexts.
Consequently, studies that investigate this relationship using race and gender as
control variables are only able to produce a limited view of this association since the
adolescent’s race-gender contexts may have a unique effect on sexual behaviors that
cannot not be explained using race or gender context alone.
Question 3: Which variables among the five explanations are significantly
associated with adolescent sexual activity when the explanations are regressed
individually?
The variables that were positively associated with sexual activity in the total
sample are family transitions, mother-child general discussions, mother-child sex
discussions, and acquaintance with deviant peers. Meanwhile, mother-child
262
activities, parental strictness, parental meal presence, mother’s religion, mother
disapproval of adolescent sexual activity, and family support were negatively
associated with a sexual activity for the total sample. These results remained the
same for both males and females when analyzed by gender. However, the family
support effect was much stronger for females than for males. This finding suggests
that family support is more important for a female’s development than for a male’s
development because the experience of support by family members may help them
to develop the skill to nurture others as an adult (Moore & Rosenthal, 1993).
Some racial differences were found for the relationship between some of the
family influence variables and adolescent sexual activity. However, these racial
differences were predominately due to gender differences. For example, family
transition was positively associated with sexual activity for both white and black
adolescents but not for Hispanic adolescents. The race-gender analyses revealed that
this racial difference is mainly due to family transition being negatively associated
with sexual activity for Latinas. Further analyses revealed that Latinas whose
mothers had engaged in more marriages or marriage-like relationships over the past
18 years were more likely to be virgins. It is unknown at this time the reason for this
result and that further studies are needed to confirm to this finding.
Mother’s night presence was found to be negatively associated with sexual
activity for white but not for black and Hispanic adolescents. However, this racial
difference is found only for males. It was speculated that mother’s presence at home
during evenings is effective in preventing their sons from engaging in sexual
behaviors by closely monitoring their activities.
263
In contrast, mother’s night presence was positively associated with sexual
activity for Latinos when compared to white adolescents. Several reasons were
given for this finding. First, Latinos may have less responsibility to help around the
household when the mother is present. Latino mothers may adhere to the
marianismo concept where the female’s role as a mother is emphasized and the
mother’s self-sacrifice and suffering for her children is celebrated (Ginorio et al.,
1995). Mothers who adhere to this ideal may allow their sons the opportunity to
leave the household and engage in other activities including sexual activity. Second,
mothers stay at home more often when they have knowledge of or are suspicious of
their son’s sexual activity. This argument would be similar to that regarding mother-
child general or sex discussions where Hispanic mothers become more involved in
their sons lives because they view their behaviors as troubling. Thus, increasing
parental presence may be one way for mothers to try and prevent their sons from
engaging in future sexual activity.
Parental presence during meals was found to be negatively associated with
sexual activity for white females but not for white males. This gender difference was
speculated to be due to white females viewing these events as times when they are
able to interact with their parents in order to build a stronger relationship with them.
Race and gender differences were also found for some of the relationships
between the parental attitude variables and adolescent sexual activity. The first race
difference is that mother’s religiosity had no association with sexual activity for
black males. The main reason given for this finding is that black mothers may be
264
using practical reasons rather than religious values for their sons not to engage in
sexual activity such as the hardships of being a teenage father.
The second race difference is that mother’s disapproval of adolescent sexual
activity had no association with sexual activity for black females and Latinas
compared to white females. Several reasons were given for this finding. From one
point of view, mothers of black females and Latinas seem to have relatively tolerant
attitudes toward adolescent sexual activity compared to mothers of white females.
This result may have lead to this effect being somewhat muted for both black
females and Latinas. From another point of view, the relationship between this
measure and adolescent sexual activity is very strong for white females, which
further contributes to race differences for this effect between white females and both
black females and Latinas.
A gender difference was found for the mother’s disapproval of adolescent
sexual activity effect between Latinos and Latinas. The main reason that was given
for this finding is that Latinos are more concerned than Latinas about upholding their
mother’s values if they disapprove of adolescent sexual activity as a respect for the
family’s values (Mirande, 1977).
Finally, a gender difference was found among white adolescents and a racial
difference was found between white and black males for the peer influence effect.
These findings suggest that peers have a greater influence on the sexual behaviors of
white males in comparison to both black males and white females even though the
number of peers who engage in deviant activities that they have acquaintance with is
similar between these three race-gender subgroups.
265
Question 4: Which of the five explanations is the most and least parsimonious
adolescent sexual activity explanation?
The deviant peer explanation was consistently the most parsimonious
adolescent sexual activity explanation in all of the analyses. However, by only
considering family influence explanations, the most parsimonious family influence
explanation regarding sexual activity for the total sample was parental supervision.
Analyzing the data by gender revealed that mother’s attitude was the most
parsimonious adolescent sexual activity explanation for males while parental
supervision was the most parsimonious adolescent sexual activity explanation for
females among the family influence explanations. This gender difference may be
due to the greater importance of mother-child interactions for the daughter’s identity
formation than for the sons as explained by Chodorow (1974). Although the mother-
child interactions may be more important for females than for males, mother’s
attitudes regarding sexuality still exert a strong influence on her son’s decision to
engage in sexual activity.
However, there was a race difference for the most parsimonious family
influence explanation. Parental attitude was the most parsimonious family influence
explanation for sexuality for both white and Hispanic adolescents. This finding
seems to suggest that the sexual values that are transmitted from the mothers to her
children have a huge influence on their decision to engage in sexual activity during
puberty for these adolescents. Family support was the most parsimonious family
influence explanation for sexual activity for black adolescents. This finding is
probably due to the greater likelihood for these adolescents to reside in non-
266
traditional households. These non-traditional households may need help from
extended family members in order for it to function normally. Since extended family
members are usually present during times of need for black households (Hines &
Boyd-Franklin, 1996), it may be that black adolescents view these individuals as
being part of the family. Consequently, the definition of family support would be
broader for adolescents who also consider their extended family members as being
part of the household compared to those who only consider their immediate family
members as being part of the household.
With the exception of Hispanic adolescents, the most parsimonious family
influence explanation for sexual activity within four race-gender subgroups was
similar to the ones found when analyzing the data by race. Among Hispanic
adolescents, parental attitude was the most parsimonious family influence
explanation for sexual activity for Latinos, but family transition was the most
parsimonious family explanation for sexual activity for Latinas. However, this
finding is somewhat controversial because a cross tabulation analysis between sexual
initiation status and family transition revealed a greater number of virgin Latinas
experiencing at least three family transitions compared to sexually experienced
Latinas.
Parental attitude being the most parsimonious family influence explanation in
some of the findings also may be somewhat controversial. It is possible that parental
monitoring could have been the most parsimonious family influence explanation in
some of the analyses if parental involvement and the parental monitoring variables
were analyzed separately. Some analyses revealed a significant association between
267
the family involvement variables and adolescent sexual activity, especially for white
adolescents. However, most of the investigations revealed that the parental
monitoring variables were not significantly associated with adolescent sexual
activity. Combining these variables together into a single measure may have resulted
in this explanation being weaker than parental attitudes for some of the analyses.
This is because model parsimony depends on both the significance of and the
number of variables in the regression analysis.
Family transition was the least parsimonious sexuality explanation for the
total sample and within the adolescent’s gender analyses. Several reasons were
given for this finding. First, the family transition measure may only be representing
the frequency of marital or family transition events children may potentially
experience in their life rather than representing the quality or degree of disruption
and emotional chaos of these events. Second, the family transition measure may
have an indirect relationship with adolescent sexual activity and other explanation(s)
could explain this effect.
The least parsimonious sexuality explanation differed within the adolescent’s
race and the adolescent’s race-gender subgroups. Family transition was the least
parsimonious sexuality explanation for white adolescents probably because this
effect may be explained by other measures. Parental attitude was the least
parsimonious sexuality explanation for black adolescents probably because of the
relatively liberal sexual values of the black culture (Staples, 1978). Finally, parental
supervision was the least parsimonious sexuality explanation for Hispanic
268
adolescents probably because many of these variables do not have an association
with sexual activity.
However, parental supervision was the least parsimonious sexual activity
explanation for white males, black males, Latinos, black females, and Latinas. This
is primarily due to many of the parental supervision variables not being significantly
associated with adolescent sexual activity. Family transition was the least
parsimonious sexual activity explanations for white females. This finding suggests
that this measure may be explained by other measures for this particular subgroup.
Question 5: Pertaining to the second analysis of this study, which variables among
the five explanations are directly and indirectly associated with adolescent sexual
activity when they are considered simultaneously?
The factors that were positively directly associated with sexual activity after
they are all considered jointly for white and black adolescents are mother-child
discussions, mother-child sex communication, and acquaintance with deviant peers.
Meanwhile, mother’s disapproval of adolescent sexual activity and family support
were negatively directly associated with sexual activity. Parental strictness was also
negatively associated with sexual activity for black adolescents of both genders, but
this effect is not racially different from white adolescents. Thus, it cannot be
determined whether parental strictness is truly associated with sexual activity after
considering other family influences for black and white adolescents.
White adolescents were the only subgroup that had indirect relationships
between the variables that comprise the five explanations and adolescent sexual
269
activity. Family transitions, mother’s religiosity, mother-child activities, and
parental strictness were the only variables that have an indirect association with
adolescent sexual activity. Although the full model shows the mediation of the
family transition variable, this process differs by gender. For white females, the
mediation of family transition only required only one other explanation, which for
this situation is parental supervision. In contrast, for white males, the mediation of
family transition required all of the remaining explanations including peer influence.
It may be that the family transition effect is multidimensional for white males
because all of the explanations are required to mediate this effect. This interpretation
would indicate that the family transition event could be more traumatic for white
males than for white females. If this were the situation, the finding would support
the findings by Newcomer & Udry (1987).
However, in contrast to their study where it was speculated that the lack of
parental support is the reason why the family transition event affects males more
than females, this study found that other explanations could also explain this effect.
When looking at the data, the greatest drop in the family transition variable beta
occurred after controlling for the parental attitude and peer influence measures. This
finding suggests that the combination of mothers more tolerant attitudes regarding
adolescent sexual activity and the presence of peers who engage in deviant activities
are the main reasons for the family transition effect for white males.
Mother’s religiosity had an indirect relationship with adolescent sexual
activity for white females. The indirect relationship occurred after accounting for
peer influence, which suggests that the influences of these friends may be stronger
270
than the influence of conservative sexual values as prescribed by the mother’s
religion. Mother’s religiosity is not initially significant for white males when
controlling for this variable.
The reason for the indirect relationship between mother-child activity and
sexual activity for white adolescents somewhat differs by gender. For white females,
the mother-child activity effect is fully accounted for by both the mother’s attitudes
and family support measures. However, both of these measures only partially
explain the mother-child activity effect for white males. Controlling for the peer
influence measure fully explains the mother-child activity effect. This finding
suggests that for white females, mothers with conservative attitudes maybe engaging
in greater amount of activities with their daughters and that these activities indicate
support by family members. For white males, mother-child activity seems to
indicate the absence of peers who engage in deviant activities in the adolescent’s life.
The data suggests that mother-child activities may have a protective effect from peer
influences for these males. The peer influence measure being able to fully accounted
for this effect seems to suggest that peers who engage in deviant activities appear in
these male’s lives when their mothers do not spend the time engaging in activities
with them.
Some of the Significant Findings From This Study
Since there are a large number of contexualized analyses in this study, there
is the risk that some of the major findings do not stand out from the results. This
271
section will point out what I regard as interesting major findings from this study.
These findings are highlighted because they were either contrary to the proposed
hypotheses, ran counter to the literature, or further expanded the literature by
providing more specific detail as to when the non-traditional household effect or the
five explanations either were or were not predictive for adolescent sexual activity.
Significant Family Structure Findings
First, this study found that the unwedded mother households were not
necessarily as pathological as described in the family literature. Although unwedded
mothers were significantly more tolerant of adolescent sexual activity than mothers
of two biological parent households, the unwedded mother’s supervision levels were
mostly similar to those of two biological parent households. In most cases,
unwedded mothers spent as much time around the home as the mothers of two
biological parent households. In addition, unwedded mothers were also stricter with
their children than two biological parent household mothers. Adolescents who reside
with an unwedded mother reported similar levels of parent-child activities, family
support levels, and acquaintances with peers who smoke, drink alcohol, and take
drugs as those who reside with two biological parents.
These findings are surprising because researchers have talked about how
children who live with an unwedded mother may be disadvantaged because there is
not another parental figure in the household who can either help to discipline the
children or to monitor the children’s activities (Smetana, 1989; Thornton, 1991).
272
The unwedded mother households may not be as disadvantaged as the two biological
parent households in the parental supervision department because these mothers are
the only parental figure in the family that children recognize since there is
presumably no permanent father figure in their lives. In addition, unwedded mothers
do not have to struggle with the parental authority issue such as those regarding the
absent biological father for the divorced/separated mother household or the
stepfather for the stepparent household. By contrast, divorced/separated mothers and
mothers of stepparent households may have weaker parental authority over their
children because there is a father figure present in their children’s lives.
The presence of these father figures in the divorced and stepparent household
situation may cause mothers to reenact the traditional motherhood role of deferring
child discipline duties to the fathers (Thomson, McLanahan, & Curtin, 1992).
However, the father’s parental authority over his children who reside in these
households is also weakened. In the divorced/separated household situation, fathers
are unable to reinforce their parental authority because they do not reside at home.
In the stepparent situation, the stepfathers may not be seen to have legitimate
parental authority in the eyes of the stepchildren because they are seen as a “relative
stranger” (Beer, 1988). Since there is only one parental figure in the unwedded
mother household situation, these mothers do not have the luxury to play the
traditional motherhood role of being lax when it comes to disciplining their children.
Consequently, these mothers have to be stricter with their children than mothers of
two biological parent households in order for the household to function smoothly
273
since there is presumably no other parental figure at home that can help with the
parental duties.
Besides not having to struggle with the issue of parental authority, unwedded
mothers also presumably do not have to struggle with issues that come with either
losing or gaining a significant other as described by Newcomer & Udry (1987) for
divorced households and Furstenberg (1987) for stepparent households. This
premise implies that unwedded mother households may be relatively more stable
than divorced/separated and stepparent households since children who live with
unwedded mothers presumably have not gone through a family transition event. If
this interpretation is correct, children who live with unwedded mothers may be less
likely to seek emotional support from their peer group in order to find solace from
their home environment. This type of household stability may also contribute to the
greater influence of unwedded mother’s attitudes and values in her children’s lives.
Consequently, this reason could explain why the mother’s attitude explanation,
rather than the parental supervision explanation, was found to be the best explanation
for the unwedded mother household effect.
Despite the unwedded mothers best efforts at controlling their home
environment by being stricter and relatively more present at home than other non-
traditional households, these households may be vulnerable to the surrounding
neighborhood. The households of unwedded mothers are significantly poorer than
any of the other households, which implies that unwedded mothers may tend to live
in neighborhoods that are more conducive to criminal activities (Wilson, 1987).
Even though adolescents who reside with unwedded mothers are no more likely to
274
have acquaintance with peers who engage in deviant activities, peer influence was
the best explanation for the unwedded mother household effect for white males and
Latinos. The reason for this finding may be due to the strength of peer influence
being greater in low socioeconomic neighborhoods where deviant behaviors is
common. Another interpretation is that these adolescents look toward their peers as
role models since there is presumably no father figure around that can be a role
model to these children.
However, there were some difficulties in determining whether the unwedded
mother households are truly different from divorced/separated during the contextual
analyses because of their small sample. Most of the differences that were found
between the unwedded mother household and the divorced/separated mother
household were found for black males. Analyzing a greater number of unwedded
mother households would make a significant difference in determining whether they
are truly different from divorced/separated households for both white and Hispanic
adolescents. Nevertheless, the results from this dissertation seem to support Moore’s
(2001) premise that unwedded mother households are different from
divorced/separated households and that they should be examined separately.
A second major finding regarding family structure is that non-traditional
households are not necessarily associated with greater adolescent sexual activity risk.
These findings do indeed challenge the commonly held assumption on the pathology
of non-traditional households as a cause of adolescent sexual experience. This study
has found three instances of this situation. First, black females who reside with a
single parent are no more likely than black females who reside with two biological
275
parents to engage in sexual activity after controlling for the demographic variables.
The finding is especially interesting because much as been written on the pathology
of female headed African American families with the assumption that low income
and low parental monitoring would lead to greater likelihood of sexual behavior even
after having a child (Kaplan, 1997). It seems that the best reason for this finding is
the institutionalization of the non-traditional households within the black culture
where adolescents may feel that living with a single parent is just as normal as living
with two parents.
The second and third instances of the non-significant relationships between
non-traditional household and adolescent sexual activity were found for white males
and Latinos who live with a stepparent. This was a surprising because stepparent
households are generally thought to be associated with increased risk for sexual
activity due to the issues that family members of these households face (Cherlin,
1978). However, these findings may be due to the new parental figure being a male
where the gender similarity between the stepparent and the stepchild for white males
and Latinos may encourage a relatively quicker bonding process. This bonding
process may allow stepfathers to quickly help their stepsons through the difficulties
of adolescence and to be a role model to them in order to develop their own gender
identity (Chodorow, 1979). Consequently, this close relationship with the new
parental figure may help to mitigate the possible neglected feeling that males may
experience during a family transition event (Newcomer & Udry, 1987).
276
Significant Family and Peer Influence Variables Findings
This study also found several significant family and peer influence findings
regarding adolescent sexual activity. The first significant finding is that most of the
analyses did not find a direct association between parental monitoring variables
(mother’s employment and presence at home during the day and night) and
adolescent sexual activity which was quite surprising. However, it was not
surprising to find that the parental interaction variables (mother-child activities and
discussions, parental presence during meals, and parental strictness) were
significantly associated with adolescent sexual activity in many of the analyses.
These findings seem to suggest that the parent-child interaction quality may be more
effective in influencing a child’s decision to engage in sexual behavior than how
often parents monitor their children’s activities in most situations. Thus, it seems
that parents may significantly lower their children’s sexual activity risk by having
meaningful interactions with their children such as spending time with them doing
activities together, eating meals with them, and imposing household rules.
The second significant family and peer influence finding is that mother-child
general discussions and mother-child sex discussions are positively rather than
negatively associated with adolescent sexual activity. This was quite surprising to
me since my assumptions for these variables were that mothers would engage in the
general and sex discussions with their children in order to enhance the relationship
and prevent problem behaviors respectively. Instead, the study consistently found
277
that increases in these discussions between the mother and the child were positively
associated with sexual activity.
Several reasons were given for the positive relationship between mother-child
sex discussion and adolescent sexual activity. Briefly, the first reason is the reverse
causality explanation where mothers have more discussions about sex with their
children when they become aware of their sexual activity in hopes of preventing
them from engaging in future sexual activity during puberty. The second reason is
that mothers who engage in these discussions openly may be more liberal about
sexual expression and pass on this philosophy to their children. These discussions
may involve practical matters such as using protection such as condoms or other
forms of birth control when engaging in sexual activity. Although such mothers may
not be advocating early sexual activity for their children, these discussions may
normalize these behaviors by suggesting that it is acceptable to engage in these
activities as long as they are well protected. Further research is needed to determine
which of these explanations is the better reason for the results of this study.
The third significant finding deals with racial and gender differences for the
relationship between both the family and peer influence explanations and adolescent
sexual activity. These differences were discussed previously and will not be
repeated here. Some of these findings, such as family transition being negatively
associated with sexual activity for Latinas, mother’s evening presence being
associated with an increased risk for sexual activity for Latinos were surprising and
seem to contradict the literature. Other findings, such as mother’s disapproval of
adolescent sexual activity not being associated with adolescent sexual activity for
278
black females and Latinas, may help to clarify the findings of other studies by
helping to determine whether the applicability of the results mainly applies to either
males or females. Despite these race and gender differences for the family and peer
influence measures, these findings are preliminary and needs to be validated by other
studies.
Study Limitations and Issues for Further Research
Despite all of the different results that are found in this dissertation, there are
some cautions in the interpretation of the findings. One limitation of this study is
that these results predominately apply to the relationship between the biological
mother and her children. This study focused on resident biological mothers because
they were the primary choice for answering the parental questionnaire (Udry &
Bearman, 1998). However, other researchers have shown that parents may have
greater influences over their children for those that are of similar gender. For
example, Thornton & Camburn (1987) found that mothers with higher levels of
education had daughters who were less likely to engage in sexual activity during
adolescence while fathers with higher levels of education had sons who were less
likely to be sexually experienced. Since this study does not include biological
fathers or other male parental figures in the analyses, the results of this study only
presents a partial depiction of how family influences are related to adolescent sexual
activity. Future studies could investigate whether these results would also apply to
other parental figures.
279
This study has presented several issues for further research the first of which
results from the problems of using cross-sectional data. The disadvantage of using
this type of data is that it only captures the respondent’s circumstances at a particular
moment in time. This disadvantage could lead to difficulty in determining the
direction of causality for some of the associations between the family influence
variables and adolescent sexual activity, especially for those where the relationship
may be debatable. One example of this issue is the mother-child sex discussion
variable where it is not known from the findings of this study whether parents who
have frequent discussions about sex with their children encourages sexual activity or
if these discussions occur because parents have some suspicion or knowledge of their
children’s sexual activity.
One way to clarify the direction of causality for these ambiguous
relationships would be to perform longitudinal analyses. Clarifying the direction of
the relationship between some of these ambiguous relationships may either support
or refute the interpretations presented in this study. Determining whether these
factors are either leading or lagging predictors for sexual activity may help educators
and therapists to understand the underlying causes for this connection.
Second, some of the findings of this study may not truly reflect the
relationship between family influence and adolescent sexual activity because this
study had to use questions that were either too ambiguous or were used as a proxy
for other socialization measures. A possible ambiguous question that was used in
this study was the family transition measure. This question asked the parent how
many serious relationships, defined as a marriage or marriage-like relationship, they
280
had had in the past 18 years. Not only does the number of family transitions not
capture the degree of rupture, conflict, and chaos accompanied by a transition event,
it is also unclear from this question as to how the respondent defines a “marriage-like
relationship”. This definition could range from a very serious and close relationship
where the adult figures are not physically who reside together to cohabitation. If the
respondent considered cohabitation as a marriage-like relationship, there may be a
debate as to whether this event fits the definition of a family transition because the
mother and/or the children may not consider these boyfriends to be a parental figure.
This question’s ambiguity may lead respondents to overstate the number of true
family transition events and further obscures the real effect that these events have on
adolescent sexual activity. Using a question that asks the respondent about the
frequency of true family transition events would be a more accurate measure of the
family transition effect.
A measure used in this study that could be interpreted as proxy for another
socialization measure is the peer influence variable. Peer influence measure was
represented by the variable that asked the child how many friends he or she knew
who smoked cigarettes, drank alcohol, and took drugs. Unfortunately, this dataset
did not contain a specific question that asks the respondent whether their friends are
sexually experienced. A question that asks the respondent the number of friends or
acquaintances he or she knows who had engaged in sexual activity may be a more
accurate measure of how a friend’s sexual experiences may influence a child’s
decision to engage in sexual behaviors during puberty.
281
Third, the results of this study assumed that adolescents are responding
truthfully to questions regarding their sexual activity. Respondents in this dataset
may underreport sexual activity since this topic is a sensitive issue especially for
adolescents. However, the extent of this bias would be difficult to estimate and
could vary across ethnic groups (Cantina et al., 1990). In addition, young men might
be inclined to over report having intercourse, which could introduce age related
errors estimates (Newcomer & Udry, 1988). Consequently, one would need to be
cautious when interpreting the findings of this study since it would be impossible to
determine whether children are giving truthful answers about their sexual behaviors
from the data.
Fourth, this study was not able to conduct cultural subgroup analyses for
Hispanic and Asian respondents because of their small numbers. Studies have
shown that that sexual activity rates do differ for various Hispanic and Asian
adolescents (East, 1998; Aneshensel, Fielder, & Becerra, 1989). Those who wish to
replicate these studies on certain cultural subgroups would need to use large samples
in order to achieve valid statistical findings.
Fifth, this study did not analyze how acculturation may influence adolescent
sexual activity for Hispanic and Asian adolescents. Acculturation is one factor that
may affect an adolescent’s decision to engage in sexual activity because those who
are acculturated may be more likely to internalize the values and mores of their host
country even if they run contrary to those from their country of origin. Researchers
have found that the sexual behaviors of acculturated Hispanic (Ford & Norris, 1993)
and Asian (East, 1998; Hou & Basen-Engquist, 1997) adolescents differ significantly
282
from their non-acculturated peers. Consequently, studies that investigate the sexual
behaviors of adolescents with either Asian or Hispanic backgrounds may need to
take into account acculturation factors in order to present a more accurate view of the
relationship between family influences and adolescent sexual activity for these
populations.
Finally, this study did not explore whether both family structure and family
influences are associated with other sexuality measures such as the adolescent’s age
of first sex, the number of sexual partners, and whether birth control was used during
sexual activity. One could investigate whether there is a family structure effect for
adolescent’s age of first sex and for number of sexual partners in order to determine
whether these results would be different from this study’s findings. However, there
are several implications when using these measures since these adolescents are
already sexually active. First, one would have to use a greater number of
respondents in the analysis in order to produce statistically valid results since the
investigation will be on approximately half of all adolescents. Second, the analyses
on sexually active adolescents may result in fewer family influence variables being
associated with sexual activity because these adolescents may have more
characteristics in common with each other as opposed to analyses that investigate
two distinct groups.
283
Research Implications
This study contains several implications for researchers who are interested in
the relationship between family environment and adolescent sexual activity. The
first implication is the need to contextualize the analyses within race and gender
when conducting research on adolescent outcomes for future studies.
Bronfenbrenner (1979) argued that each individual brings to the family certain
personal attributes and characteristics and since family processes always occur
within a specific context, these processes may have different effects for different
groups of individuals in distinct family settings. The results of this study
wholeheartedly support this argument. As shown by the results in chapter 8, the
explanations that best account for the family structure effect regarding sexuality for
adolescents who live in similar households do vary within race-gender contexts. In
addition, one could not have predicted these differences if the data were analyzed
within the adolescent’s gender or adolescent’s race categories alone. These findings
suggest that the convergence of culture and gender have somewhat of a unique effect
on a child’s life. Consequently, future studies would need to contexualize the
adolescent sexual activity analyses within race-gender subgroup categories rather
than within race or gender categories alone in order to discover the nuances that
make each subgroup different from each other.
The second implication is the need to examine non-traditional households
separately when investigating the relationship between family structure and
adolescent sexual activity. This is because the family structure effect may differ
284
across households that share similar family configurations. In addition, this study
found that the best explanation for the unwedded mother and divorced/separated
mother effects sometimes differed from each other. These findings support Moore’s
(2001) contention that there are environmental differences between unwedded
mother and divorced/separated mother households, which justifies analyzing these
typologies separately. Other family configurations that would need further research
to determine whether these households are different environmentally include same
sex parent and cohabitating households.
Social Policy and Counseling Implications
The results of this study also have several implications for educators and
counselors in addressing adolescent sexual activity. First, this study consistently
found that the conflict perspective, especially acquaintance with deviant peers, is the
most parsimonious explanation for adolescent sexual activity. This is not very
surprising since adolescents spend more time with their friends than by themselves
(Csikszentmihalyi & Larson, 1984). On the surface, this finding would suggest a
treatment or social policy of having parents understand their children’s peer value
system so they can better identify discrepancies with their own values and emphasize
those values that will encourage lowered sexual activity. However, this study has
shown an increasing influence by friends and peers when adolescents feel
disconnected from their family. Thus, educators and counselors may need to address
285
the family system in order to reduce the peer group influence when attempting to
discourage children from engaging in sexual behaviors during puberty.
Second, the best explanation for the non-traditional household effect
regarding sexual activity varies by both the household configuration and the
adolescent’s race-gender contexts. The following paragraphs provide some
guidelines, based on the results of this study, for adults who are concerned about this
issue as to how to effectively address the family structure issues among the three
non-traditional households. However, there needs to be some caution when one sets
social policy or counseling suggestions for families by only using the findings in this
study because these guidelines are not absolute. The best explanation for the family
structure and adolescent sexual activity relationship among the non-traditional
households does not necessarily mean that other explanations are irrelevant when
trying to deal with this issue. These guidelines are just recommendations that one
may want to explore further or place special emphasis on when addressing this issue
in order to efficiently prevent children from engaging in sexual activity during
puberty.
Addressing mother’s attitudes may best serve white males who reside with
unwedded mothers and Latinos who reside with divorced/separated mothers. For
white males, it would probably be very difficult for unwedded mothers to move from
the neighborhood environments that they live in especially since these households do
not have many financial resources. However, it would be much easier to change
one’s attitudes toward social issues than to change one’s neighborhood. Since
unwedded parents households seem to be relatively stable compared to
286
divorced/separated parent households, it could be possible that the white males who
reside with an unwedded mother may be more respectful of their parent’s authority.
Having these unwedded mothers to set higher standards of conduct for their son’s
actions may result in them being less likely to engage in adolescent sexual activity
compared to those mothers who espouse relatively liberal sexual values.
However, exploring the concept of machismo and familism may best serve
Latinos who reside with divorced/separated mothers. Helping the mother and her
son to understand that machismo does not necessarily mean that one needs to engage
in sexual activity to reaffirm a male’s identity and that familism does not necessarily
only apply to two biological parent households may help these families to reduce
likelihood of sexual activity during puberty for Latinos (Mikawa, et al., 1992).
Addressing the issues regarding the loss of the parent and the stresses of
going through a family transition may best serve white males who reside with
divorced/separated mothers. The loss of the parent for this study is that of a father
figure. This issue may be very difficult for these males to struggle with because he is
going through a time of change while losing a person in his life who could have been
a role model for them or who could be there to support them while confronting the
challenges of puberty. One way to help ease the family transition effect is for the
parents to have the divorce process be less contentious and relatively free of the
issues regarding the previous marriage so that the father-son relationship could
continue (Ahrons & Rodgers, 1987). Another way to help ease the family transition
effect would be to recognize that males are the ones who receive less care when the
family goes through a divorce (Newcomer & Udry, 1987). Divorced mothers who
287
interact with their sons more frequently during this time may help to ease the effects
of the family transition and lower the likelihood that their son would engage in
sexual activity as a response to the family situation.
Forming a closer relationship between the mother and daughter may best
assist white females who reside with a single mother. Mothers of these households
may be less likely to be around the house because they would have to work longer
hours in order to maintain the financial status of the household. However, the
findings of this study has consistently shown that the quality of the relationship (such
as engaging in mother-daughter activities and the mother eating meals with her
children) rather than the amount of time parents spend at home is more of a
determining factor for white female’s sexual activity. Consequently, single mothers
may lower their daughter’s risk of sexual activity by developing a close relationship
with them.
Black males who reside with a stepparent may be best served by the
addressing the issues of integrating a father figure into the household. These
households may have issues such as confusion in knowing the appropriate roles of
stepparents and non-custodial parents in disciplining children and possible emotional
attachments to the biological father or other father figures. Helping the stepfather
and his stepson to connect with each other may lower the risk of sexual activity for
black males who reside in these households.
Addressing the meaning of sex outside of the marriage may best serve black
females who reside with a stepparent. It may be that females who reside in these
households are more likely to engage in sexual activity during adolescence because
288
they saw their mothers engaging in these behaviors outside of wedlock. Having
these mothers address the issue that sexual activity should be limited to adults rather
than children may help to lower the risk of adolescent sexual activity for black
females.
Finally, addressing the mother’s attitudes toward sexual activity may best
serve Latinas who reside with a single mother. It may be that these mothers do not
necessarily subscribe to either the concept of marianismo due to their upbringing of
having a child out of wedlock or familism because they do not see their family as the
idealistic household. Thus, the most effective approach to this issue would probably
be for these household members to understand the importance of marianismo in a
Latina’s life and that familism does not necessarily only apply to two biological
parent households.
Final Comments
This study presented results from two investigations. The first investigation
examined the primary reason why adolescents who reside in non-traditional
households were more likely than those who reside with two biological parents to
engage in sexual activity using family and deviant peer influence explanations. The
second investigation examined the relationship between both the family and peer
influences and adolescent sexual activity. These investigations are complex due to
the analyzing the data within a variety of adolescent race and gender contexts.
289
The results of these investigations are varied and interesting. However, these
investigations also have numerous shortcomings and limitations. Thus, this study is
only just the beginning in the analysis of the relationship between family influences
and adolescent sexual activity.
290
References
Abrahamse, A. F., Morrison, P. A., & Waite, L. J. (1988). Teenagers willing to
consider single parenthood: Who is at greatest risk? Family Planning
Perspectives, 30, 1, 12-18.
Adolph, C., Ramons, D. E., Linton, K. L. P., & Grimes, D. A. (1995). Pregnancy
among Hispanic teenagers: Is good parental communication a deterrent?
Contraception, 51, 303-306.
Ahrons, C. R., & Rodgers, R. H. (1987). Divorced families: Meeting the challenge of
divorce and remarriage. New York: W.W. Norton & Company, Inc.
Allen, L., & Mitchell, C. (1996). Poverty and adolescent health. In M. Kagawa-
Singer, P. Katz, D. Taylor, & J. Vanderryn (Eds.), Health issues for minority
adolescents (pp. 1-35). Lincoln: University of Nebraska Press.
Amato, P. R. (1987). Family processes in one-parent, stepparent, and intact families:
The child’s point of view. Journal of Marriage and the Family, 49, 327-337.
Aneshensel, C. S., Fielder, E. P., Becerra, R. M. (1989). Fertility and fertility-related
behavior among Mexican-American and non-Hispanic white female
adolescents. Journal of Health and Social Behavior, 30, 56-76.
Bach-y-Rita, G. (1982). The Mexican American: Religious and cultural influences.
In R.M. Becerra, M. Karno, & J. I. Escobar (Eds.), Mental health and
Hispanic Americans: Clinical perspectives (pp. 29-40). New York: Grune &
Stratton.
Baumeister, L. M., Flores, E., & Marin, B. V. (1995). Sex information given to
Latina adolescents by parents. Health Education Research, 10, 2, 233-239.
Beer (1988). Relative strangers: Studies of stepfamily processes. Totowa, NJ:
Rowman & Littlefield.
Benda, B. B., & Corwyn, R. F. (1998). Race and gender differences in theories of
sexual behavior among rural adolescents who reside in AFDC families. Youth
& Society, 30, 1, 59-88.
Biblarz, T. J., & Raftery, A. E. (1999). Family structure, educational attainment, and
socioeconomic success: Rethinking the “pathology of matriarchy”. American
Journal of Sociology, 105, 2, 321-365.
291
Billy, J. O. G., & Udry, J. R. (1985). The influence of male and female best friends
on adolescent sexual behavior. Adolescence, 20, 21-32.
Bock, E. W., Beeghley, L., & Mixon, A. J. (1983). Religion, socioeconomic status,
and sexual morality: An application of reference group theory. The
Sociological Quarterly, 24, 545-559.
Bredemeier, H. C., & Stephenson, R. M. (1962). The analysis of social systems. New
York: Holt, Rinehart, & Winston, Inc.
Bronfenbrenner, U. (1979). The ecology of human development: Experiments by
nature and design. Cambridge, MA: Harvard University Press.
Brooks-Gunn, J., & Ruble, D. N. (1982). The development of menstrual-related
beliefs and behaviors about early adolescence. Child Development, 53, 1567-
1577.
Brooks-Gunn, J., & Furstenberg, F. F. (1989). Adolescent sexual behavior. American
Psychologist, 44, 249-257.
Brown, B. B., Clasen, D. R., & Eicher, S. A. (1986). Perceptions of peer pressure,
peer conformity dispositions, and self-reported behavior among adolescents.
Developmental Psychology, 22, 521-530.
Cantina, J. A., Gibson, D. R., Chitwood, D. D., & Coates, T. J. (1990).
Methodological problems in AIDS behavioral research: Influences on
measurement error and participation bias in studies of sexual behavior.
Psychological Bulletin, 108, 339-362.
Capaldi, D. M. & Patterson, G. R. (1991). Relation of parental transitions to boys’
adjustment problems: I. A linear hypothesis. II. Mothers at risk for transitions
and unskilled parenting. Developmental Psychology, 27, 489-504.
Capaldi, D. M., Crosby, L., & Stoolmiller, M. (1996). Predicting the timing of first
sexual intercourse for at-risk adolescent males. Child Development, 67, 344-
359.
Card, J., & Wise, L. (1978). Teenage mothers and teenage fathers: The impact of
early childbearing on the parents personal and professional resides. Family
Planning Perspectives, 10, 199-205.
Casper, L. (1990). Does family interaction prevent adolescent pregnancy? Family
Planning Perspectives, 22, 109-114.
292
Chase-Lansdale, P. L., Cherlin, A. J., & Kiernan, K. E. (1995). The long-term effects
of parental divorce on the mental health of young adults: A developmental
perspective. Child Development, 66, 1614-1634.
Cherlin, A. J. (1978). Remarriage as an incomplete institution. American Journal of
Sociology, 83, 634-650.
Chodorow, N. (1974). Family Structure and Feminine Personality. In M. Z. Rosaldo
and L. Lamphere (Eds.), Woman, Culture and Society. Stanford: Stanford
University Press.
Christopher, F. S., Johnson, D. C., & Roosa, M. W. (1993). Family, individual, and
social correlates of early Hispanic adolescent sexual expression. The Journal
of Sex Research, 30, 1, 54-61.
Clark, S., Zabin, L., & Hardy, J. (1984). Sex, contraception and parenthood:
Experience and attitudes among urban black young men. Family Planning
Perspectives, 16, 77-82.
Cohen, B. H. (2001). Explaining psychological statistics. New York: Wiley.
Coleman, J. C. (1981). Friendship and the peer group in adolescence. In J. Adelson
(Ed.). Handbook of Adolescent Psychology. New York: Wiley.
Collins, R. (1975). Conflict sociology: Regarding an explanatory science. New
York: Academic Press.
Connor, J. W. (1976). Family bonds, material closeness, and the suppression of
sexuality in three generations of Japanese Americans. Ethos, 4, 189-221.
Cooper, C. R. (1988). The role of conflict in adolescent-parent relationships. In M.
R. Gunnar (Eds.), Minnesota symposia on child psychology: Development
during the transition to adolescence (Vol. 21, pp. 1181-1187). Hillsdale, NJ:
Erlbaum.
Csikszentmihalyi, M., & Larson, R. (1984). Being adolescent: Conflict and growth
in the teenage years. New York: Basic Books.
Darabi, K. F., & Ortiz, V. (1987). Childbearing among young Latino women in the
United States. American Journal of Public Health, 77, 25-28.
Davis, S. M., & Harris, M. B. (1982). Sexual knowledge, sexual interests, and
sources of sexual information of rural and urban adolescents who reside
inthree cultures. Adolescence, 17, 471-492.
293
Davis, E. C., & Friel, L. V. (2001). Adolescent sexual activity: Disentangling the
effects of family structure and family context. Journal of Marriage and
Family, 63, 3, 669-681.
Day, R. D. (1992). The transition to first intercourse among racially and culturally
diverse youth. Journal of Marriage and the Family, 54, 749-762.
DeLamater, J. (1981). The social control of sexuality. Annual Review of Sociology,
7, 263-290.
Devereaux, E. C. (1970). The role peer group experience in moral development. In J.
P. Hill (ed.), Minnesota symposia on child development. Vol. 4. Minneapolis:
University of Minnesota Press, pp. 94-140.
Dittus, P. J., & Jaccard, J. (2000). Adolescents’ perception of maternal disapproval
of sex: Relationship to sexual outcomes. Journal of Adolescent Health, 26,
268-278.
Dornbusch, S. M., Carlsmith, J. M., Bushwall, S. J., Ritter, P. L., Leiderman, H.,
Hastorf, A. H., & Gross, R. T. (1985). Single parents, extended households,
and the control of adolescents. Child Development, 56, 326-341.
Driscoll, A. K., Biggs, M. A., Brindis, C. D., & Yankah, E. (2001). Adolescent
Latino reproductive health: A review of the literature. Hispanic Journal of
Behavioral Sciences, 23, 3, 255-326.
Dryfoos, J. G. (1990). Adolescent at Risk: Prevalence and Prevention, New York:
Oxford University Press.
East, P. L. (1998). Racial and ethnic differences in girls’ sexual, marital, and birth
expectations. Journal of Marriage and the Family, 60, 150-162.
Emery, R. E., Hetherington, E. M., & Dilalla, L. F. (1985). Divorce, children, and
social policy. In H. Stevenson & A. Seigal (Eds.), Child development
research and social policy (pp. 189-266). Chicago: University of Chicago
Press.
Flewelling, R. L., & Bauman, K. E. (1990). Family structure as a predictor of initial
substance use and sexual intercourse in early adolescence. Journal of
Marriage and the Family, 52, 171-181.
Ford, K., & Norris, A. E. (1993). Urban Hispanic adolescents and young adults:
Relationship of acculturation to sexual behavior. The Journal of Sex
Research, 30, 4, 316-323.
294
Fox, G. L., & Inazu, J. (1980). Mother-daughter communication about sex. Family
Relations, 29, 347-
Furstenberg, F. F. (1987). The new extended family: The experience of parents and
children after remarriage. In K. Pasley & M. Ihinger-Tallman (Eds.),
Remarriage and stepparenting: Current research and theory (pp. 42-61).
New York: Guilford Press.
Furstenberg, F. F., Morgan, S. P., Moore, K. A., & Peterson, J. L. (1987). Race
difference in the timing of adolescent intercourse. American Sociological
Review, 52, 511-518.
Gibson, J. W. & Kempf, J. (1990). Attitudinal predictors of sexual activity in
Hispanic adolescent females. Journal of Adolescent Research, 5, 4, 414-430.
Ginorio, A., Gutierrez, L., Cauce, A. M., & Acosta, M. (1995). The psychology of
Latinas. In C. Travis (Ed.), Feminist perspectives on the psychology of
women (pp. 331-342). Washington, DC: American Psychological
Association.
Gove, W. R., & Crutchfield, R. D. (1982). The family and juvenile delinquency.
Sociological Quarterly, 23, 301-319.
Hansson, R. O., O’ Connor, M. E., Jones, W. H., & Blocker, T. J. (1981). Maternal
employment and adolescent sexual behavior. Journal of Youth and
Adolescence, 10, 1, 55-60.
Harlan, W., Harlan, E., Grillo, G. (1980). Secondary sex characteristics of girls 12 to
17 years of age: The U.S. health examination survey. Journal of Pediatrics,
95, 293-297.
Havighurst, R. J. (1967). Developmental tasks and education. New York: David
McKay.
Hayes, C. (Ed.). (1987). Risking the future: Adolescent sexual activity, pregnancy,
and childbearing (Vol. 1.). Washington, D.C.: National Academy Press.
Hayward, M. D., Grady, W. R., & Billy, J. O. (1992). The influence of
socioeconomic status on adolescent pregnancy. Social Science Quarterly, 73,
751-772.
Hetherington, E. M. (1979). Children of divorce. In R. Henderson (Ed.). Parent-
Child Interaction: Theory, Research and Prospects. New York: Academic
Press.
295
Hetherington, E. M., & Camara, K, A. (1984). Families in transition: The process of
dissolution and reconstitution. In R. Parke (Ed.), Review of child development
research (Vol 3, pp. 398-439). Chicago: University of Chicago Press.
Hetherington, E. M. (1989). Coping with family transitions: Winners, losers, and
survivors. Child Development. 60, 1, 1-14.
Hetherington, E. M., & Clingempeel, W. G. (1992). Coping with marital transitions:
A family systems perspective. Monographs of the Society for Research in
Child Development. 57, 2-3, 227, 1-242.
Hines, P. M. & Boyd-Franklin, N. (1996). African American Families. In M.
McGoldrick, J. Giordano, & J. K. Pearce (Eds.), Ethnicity & Family Therapy
(pp. 66-84). New York: The Guilford Press.
Hirschi, T. (1969). Causes of delinquency. Berkeley: University of California Press.
Hirayama, H., & Hirayama, K. K. (1986). The sexuality of Japanese Americans.
Journal of Social Work and Human Sexuality, 4, 3, 81-98.
Hofferth, S. L., & Moore, K. A. (1979). Early childbearing and later economic well
being. American Sociological Review, 44, 784-815.
Hofferth, S. L., Kahn, J. R., & Baldwin, W. (1987). Premarital sexual activity among
U.S. teenage women over the past three decades. Family Planning
Perspectives, 19, 46-53.
Hogan, D. P., & Kitagawa, E. M. (1985). The impact of social status, family
structure, and neighborhood on the fertility of black adolescents. American
Journal of Sociology, 90, 4, 825-855.
Holmes, T. H., & Rahe, R. H. (1967). The social readjustment rating scale. Journal
of Psychosomatic Research, 11, 213-218.
Holtzman, D., & Rubinson, R. (1995). Parent and peer communication effects on
AIDS-related behavior among U.S. high school students. Family Planning
Perspectives, 27, 235-240 & 268.
Hou, Su-I, & Basen-Engquist, K. (1997). Human immunodeficiency virus risk
behavior among White and Asian/Pacific Islander high school student in the
United States: Does culture make a difference? Journal of Adolescent Health,
20, 68-74.
296
Hovell, M., Sipan, C., Blumber, E., Atkins, C., Hofstetter, C. R., & Kreitner, S.
(1994). Family influences on Latino and Anglo adolescents’ sexual behavior.
Journal of Marriage and the Family, 56, 973-986.
Hunter & Youniss (1982). Changes in functions of three relations during
adolescence. Developmental Psychology, 18, 6, 806-811.
Inazu, J. K., & Fox, G. L. (1980). Maternal influence on the sexual behavior of
teenage daughters. Journal of Family Issues, 1, 81-102.
Jaccard, J., Turrisi, R., & Wan, C. K. (1990). Interaction effects in multiple
regression. Sage University paper series on quantitative applications in the
social sciences, 07-072. Newbury Park, CA: Sage.
Jaccard, J., & Dittus, P. J. (1991). Parent-teen communication: Regarding the
prevention of unintended pregnancies. New York: Springer-Verlag.
Jaccard, J., Dittus, P. J., & Gordon, V. V. (1996). Maternal correlates of adolescent
sexual and contraceptive behavior. Family Planning Perspectives, 28, 4, 159-
165, 185.
Jacobson, K. C., & Crockett, L. J. (2000). Parental monitoring and adolescent
adjustment: An ecological perspective. Journal of Research on Adolescence,
10, 1, 65-97.
Kaplan, E. B. (1997). Not our kind of girl: Unraveling the myths of black teenage
motherhood. Berkeley: University of California Press.
Landsbaum, J., & Willis, R. (1971). Conformity in early and late adolescence.
Developmental Psychology, 4, 334-337.
Laumann, E. O., Gagnon, J. H., Michael, R. T., & Michaels, S. (1994). The social
organization of sexuality: Sexual practices in the United States. Chicago:
University of Chicago Press.
Lauritsen, J. L. (1994). Explaining race and gender differences in adolescent sexual
behavior. Social Forces, 72, 3, 859-883.
Liebowitz, S. W., Castellano, D. C., & Cuellar, I. (1999). Factors that predict sexual
behaviors among young Mexican American adolescents: An exploratory
study. Hispanic Journal of Behavioral Sciences, 21, 4, 470-479.
Maccoby, E. E., & Martin, J. A. (1983). Socialization in the context of the family:
Parent-child interaction. In E. M. Hetherington (Ed.)., P. H. Mussen (Series
297
Ed.), Handbook of child psychology: Vol. 4. Socialization, personality, and
social development (1-101). New York: Wiley.
McCord, J. (1990). Problem behaviors. In S. S. Feldman & G. R. Elliott (Eds.), At
the threshold: The developing adolescent. Cambridge, MA: Harvard
University Press.
McLanahan, S., & Bumpass, L. (1988). Intergenerational consequences of family
disruption. American Journal of Sociology, 94, 1, 130-152.
McLanahan, S., & Sandefur, G. (1994). Growing up with a single parent.
Cambridge, Mass.: Harvard University Press.
McNeely, C., Shew, M. L., Beuhring, T., Sieving, R., Miller, B. C., Blum, R. W.
(2002). Mother’s influence on the timing of first sex among 14- and 15-year-
olds. Journal of Adolescent Health, 31, 256-265.
Mead, G. H. (1964). On Social Psychology. Chicago, IL: University of Chicago
Press.
Merton, R. (1957). Social theory and social structure. Glencoe, IL: Free Press.
Metzler, C. W., Noell, J., Biglan, A., Ary, D., & Smolkowski, K. (1994). The social
context for risky sexual behavior among adolescents. Journal of Behavioral
Medicine, 17, 4, 419-438.
Mikawa, J. K., Morones, P. A., Gomez, A., Case, H. L, Olsen, D., & Gonzales-Huss,
M. J. (1992). Cultural practices of Hispanics: Implications for prevention of
AIDS. Hispanic Journal of Behavioral Sciences, 14, 4, 421-433.
Miller, B.C., McCoy, J. K., Olson, T. D., & Wallace, C. M. (1986). Parental
discipline and control attempts in relation to adolescent sexual attitudes and
behavior. Journal of Marriage and the Family, 48, 503-512.
Miller, B. C., Higginson, R., McCoy, J. K., & Olson, T. D. (1987). Family
configuration and adolescent sexual attitudes and behavior. Population and
Environment, 9, 2, 111-123.
Miller, B. C., & Moore, K. A. (1990). Adolescent sexual behavior, pregnancy, and
parenting: Research through the 1980s. Journal of Marriage and the Family,
52, 1025-1044.
Miller, K. S., Forehand, R., & Kotchick, B. A. (1999). Adolescent sexual behavior in
two minority samples: The role of family variables. Journal of Marriage and
the Family, 61, 1, 85-98.
298
Mirande, A. (1977). The Chicano family: A reanalysis of conflicting views. Journal
of Marriage and the Family, 39, 747-756.
Moore, K. (1992). Facts at a glance. Washington, DC: Child Trends.
Moore, K. A., Peterson, J. L., & Furstenberg, F. F. (1986). Parental attitudes and the
occurrence of early sexual activity. Journal of Marriage and the Family, 48,
777-782.
Moore, S., & Rosenthal, D. (1993). Sexuality in adolescence. New York: Routledge.
Moore, K. A., Simms, M. C., & Betsey, C. L. (1986). Choices and circumstance.
New Brunswick, NJ: Transaction Books.
Moore, M. R. (2001). Family environment and adolescent sexual debut in alternative
household structures. In R. T. Michael (Ed.), Social awakening: Adolescent
behavior as adulthood approaches (pp. 109-136). New York: Russell Sage.
Moore, M. R., & Chase-Lansdale, P. L. (2001). Sexual intercourse and pregnancy
among African American girls in high-poverty neighborhoods: The role of
family and perceived community environment. Journal of Marriage and
Family, 63, 1146-1157.
Murry, V. M. (1992). Sexual career paths of Black adolescent females: A study of
socioeconomic status and other life experiences. Journal of Adolescent
Research, 7, 4-27.
Murry, V. M. (1994). Black adolescent females: A comparison of early versus late
coital initiators. Family Relations, 43, 342-348.
Nagelkerke, N. J. D. (1991). A note on a general definition of the coefficient of
determination. Biometrika, 78, 3, 691-692.
Nathanson, C. A. (1991) Dangerous Passage: The Social Control of Sexuality in
Women’s Adolescence, Philadelphia: Temple University Press.
Newcomer, S. F., & Udry, J. R. (1983). Adolescent sexual behavior and popularity.
Adolescence, 18, 515-522.
Newcomer, S. F., & Udry, J. R. (1984). Mother’s influence on the sexual behavior of
their teenage children. Journal of Marriage and the Family, 46, 477-485.
Newcomer, S. F., & Udry, J. R. (1985). Parent-child communication and adolescent
sexual behavior. Family Planning Perspectives, 17, 4, 169-174.
299
Newcomer, S. F., & Udry, J. R. (1987). Parental marital status effects on adolescent
sexual behavior. Journal of Marriage and the Family, 49, 235-240.
Newcomer, S. F., & Udry, J. R. (1988). Adolescent honesty in a survey of sexual
behavior. Journal of Adolescent Research, 3, 419.
Newton, R. R., & Rudestam, K. E. (1999). Your statistical consultant: Answers to
your data analysis questions. Thousand Oaks, CA: Sage.
Noller, P. & Callan, C. (1991). The adolescent in the family. New York: Routledge.
Offer, D. (1969). The psychological world of the teenager. New York: Basic Books.
Offer, D., Ostrov, E. & Howard, K. (1981). The adolescent: A psychological self-
portrait. New York: Basic Books.
Ooms, T. (1981). Teenage pregnancy in a family context. Philadelphia: Temple
University.
O’ Sullivan, L. F., Jaramillo, B. M. S., Moreau, D., & Meyer-Bahlburg, H. F. L.
(1999). Mother-daughter communication about sexuality in a clinical sample
of Hispanic adolescent girls. Hispanic Journal of Behavioral Sciences, 21, 4,
447-469.
Padilla, E. R., & O’Grady, K. E. (1997). Sexuality among Mexican Americans: A
case of sexual stereotyping. Journal of Personality and Social Psychology,
52, 5-10.
Papillo, A. R., Franzetta, K., Manlove, J., Moore, K. A., Ryan, S., Irkamullah, E., &
Terry-Humen, E. (2003). 2003 Facts at a glance. Washington, D.C.: Child
Trends.
Parrillo, A. V., Felts, W. M., & Mikow-Porto, V. (1997). Early initiation of sexual
intercourse and its co-occurrence with other health-risk behaviors in high
school students: The 1993 North Carolina youth risk behavior survey.
Journal of Health Education, 28, 2, 85-93.
Parson, T. & Bales, R. F. (1955). Family, socialization and interaction process.
Glencoe, IL: The Free Press.
Patterson, G. R., & Dishion, T. J. (1985). Contributions of families and peers to
delinquency. Criminology, 23, 1, 63-79.
300
Peterson, J. L., & Zill, N. (1986). Marital disruption, parent-child relationships, and
behavior problems in children. Journal of Marriage and the Family, 48, 295-
307.
Pick, S., & Palos, P. A. (1995). Impact of the family on the sex resides of
adolescents. Adolescence, 30, 119, 667-675.
Powell, B., & Steelman, L. C. (1993). The educational benefits of being spaced out:
Sibling density and educational progress. American Sociological Review, 58,
367-381.
Raftery, A. E. (1995). Bayesian Model Selection in Social Research (with
Discussion). In P. W. Marsden (Ed.), Sociological Methodology 1995 (pp.
111-195). Cambridge, MA: Blackwell Publishers.
Ramirez-Valles, J., Zimmerman, M. A., Juarez, L. (2002). Gender differences of
neighborhood and social control processes: A study of the timing of first
intercourse among low-achieving, urban, African American youth. Youth &
Society, 33, 3, 418-441.
Ransford, H. E. (1994). Race and class in American society: Black, Latino, and
Anglo. 2
nd
ed. Rochester: Schenkman Publishing Company.
Reiss, I. L. (1967). The social context of premarital sexual permissiveness. New
York: Holt, Rinehart, & Winston.
Rodgers, J. L. (1983). Family configuration and adolescent sexual behavior.
Population and Environment, 6, 2, 73-83.
Savin-Williams, R. C., & Berndt, T. J. (1990). Relationship and peer relations. In S.
S. Feldman & G. R. Elliott (Eds.), At the threshold: The developing
adolescent. Cambridge, MA: Harvard University Press.
Schuster, M. A., Bell, R. M., Nakajima, G. A., & Kanouse, D. E. (1998). The sexual
practices of Asian and Pacific Islander high school students. Journal of
Adolescent Health, 23, 221-231.
Scott-Jones, D., & Turner, S. L. (1988). Sex, education, contraceptives and
reproductive knowledge, and contraceptive use among Black adolescent
females. Journal of Adolescent Research, 3, 171-187.
Sells, W. C., & Blum, R. W. (1996). Morbidity and mortality among U.S.
adolescents: An overview of data and trends. American Journal of Public
Health, 86, 513-519.
301
Simons, R. L., and Associates (1996). Understanding differences between divorced
and intact families: Stress, interaction, and child outcome. Thousand Oaks,
CA: Sage.
Simons, R. L., Lin, K., Gordon, L. C., Conger, R. D., & Lorenz, F. O. (1999).
Explaining the higher incidence of adjustment problems among children of
divorce compared with those in two-parent families. Journal of Marriage and
the Family, 61, 1020-1033.
Slonim-Nevo, V. (1992). First premarital intercourse among Mexican-American and
Anglo-American adolescent women: Interpreting ethnic differences. Journal
of Adolescent Research, 7, 3, 332-351.
Small, S. A., Silverberg, S. B., & Kerns, D. (1993). Adolescents’ perception of the
costs and benefits of engaging in health-compromising behaviors. Journal of
Youth and Adolescence, 22, 73-87.
Small, S. A., & Luster, T. (1994). Adolescent sexual activity: An ecological, risk-
factor approach. Journal of Marriage and the Family, 56, 181-192.
Smetana, J. G. (1989). Adolescents’ and parents’ reasoning about actual family
conflict. Child Development, 60, 1052-1067.
Smith, C. A. (1997). Factors associated with early sexual activity among urban
adolescents. Social Work, 42, 4, 334-346.
Sonenstein, F. (1986). Risking paternity: Sex and contraception among adolescent
males. In A. Elster and M. Lamb (eds.), Adolescent fatherhood (pp. 31-54).
Hillsdale, NJ: Erlbaum.
Stanton, B., Li, X., Black, M., Ricardo, I., Gailbraith, J., Kaljee, L., & Feigelman, S.
(1994). Sexual practices and intentions among preadolescent and early
adolescent low-income urban African Americans. Pediatrics, 93, 966-973.
Staples, R. (1978). The black family: Essays and studies. 2
nd
ed. Belmont, CA:
Wadsworth.
Staples, R., & Mirande, A. (1980). Racial and cultural variations among American
families: A decennial review of the literature on minority families. Journal of
Marriage and the Family, 42, 4, 887-903.
Steinberg, L. (1990). Interdependency in the family: Autonomy, conflict, and
harmony in the parent-adolescent relationship. In S. S. Feldman & G. R.
Elliot (Eds.), At the threshold: The developing adolescent (pp. 255-276).
Cambridge, MA: Harvard University Press.
302
Strunin, L. (1991). Adolescents’ perceptions of risk for HIV infection: Implications
for future research. Social Science and Medicine, 32, 221-228.
Taris, T. W., & Semin, G. R. (1997). Parent-child interaction during adolescence,
and the adolescent’s sexual experience: Control, closeness, and conflict.
Journal of Youth and Adolescence, 26, 4, 373-398.
Taylor, D., & Katz, P. (1996). Health related services available to black adolescents.
In M. Kagawa-Singer, P. Katz, D. Taylor, & J. Vanderryn (Eds.), Health
issues for minority adolescents (pp. 36-79). Lincoln: University of Nebraska
Press.
Taylor, R. J., Tucker, M. B., Chatters, L. M., & Jayakody R. (1997). Recent
demographic trends in African American family structure. In R. J. Taylor, J.
S. Jackson, & L. M. Chatters (Eds.), Family life in Black America (pp. 14-
62). Thousand Oaks, CA: Sage.
Thomson, E., McLanahan, S. S., & Curtin, R. B. (1992). Family structure, gender,
and parental socialization. Journal of Marriage and the Family, 54, 368-378.
Thornburg, H. D. (1981). Adolescents’ sources of information about sex. Journal of
School Health, 51, 4, 274-277.
Thornton, A. (1989). Changing attitudes regarding family issues in the United States.
Journal of Marriage and the Family, 51, 873-893.
Thornton, A. (1991). Influence of the marital history of parents on the marital and
cohabitational experiences of children. American Journal of Sociology, 96, 4,
868-894.
Thornton, A., & Camburn, D. (1987). The influence of the family on premarital
sexual attitudes and behavior, Demography, 24, 323-340.
Tucker, S. K. (1989). Adolescent patterns of communication about sexually related
topics. Adolescence, 94, 269-278.
Udry, J. R. (1988). Biological predispositions and social control in adolescent sexual
behavior. American Sociological Review, 53, 709-722.
Udry, J. R., Talbert, L., & Morris, N. M. (1986). Biological foundations for
adolescent female sexuality. Demography, 23, 2, 217-228.
Udry, J. R., & Billy, J. O. (1987). Initiation of coitus in early adolescence. American
Sociological Review, 52, 841-855.
303
Udry, J. R., & Bearman, P. S. (1998). New Methods for New Research on
Adolescent Sexual Behavior. In R. Jessor (Ed.), New Perspectives on
Adolescent Risk Behavior. (pp. 241-269) Cambridge: Cambridge University
Press.
U.S. Census Bureau, Annual Social and Economic Supplement: 2003 Current
Population Survey, Current Population Reports, Series P20-553 “America’s
Families and Who reside Arrangements: 2003” and earlier reports.
Upchurch, D. M., Aneshensel, C. S., Sucoff, C. A., & Levy-Storms, L. (1999).
Neighborhood and family contexts of adolescent sexual activity. Journal of
Marriage and the Family, 61, 4, 920-933.
Wallerstein, J. S., & Kelly, J. B. (1980). Surviving the breakup: How children and
parent cope with divorce. New York: Basic Books.
Wallerstein, J. S., & Kelly, J. B. (1982). The father-child relationship: Changes after
divorce. In S. Cath, A Gurwitt, & J. Ross (Eds.), Father and child:
Developmental and clinical perspectives (pp. 451-466). Boston: Little
Brown.
Wallerstien, J. S., & Blakeslee, S. (1989). Second chances: Men, women, and
children a decade after divorce. New York: Ticknor & Fields.
Warren, C. W., Santelli, J.S., Everett, S. A., Kann, L., Collins, J. L., Cassell, C.,
Morris, L., & Kolbe, L. J. (1998). Sexual behavior among U.S. high school
students, 1990-1995. Family Planning Perspectives, 30, 170-172, 200.
Weinstein, M. & Thronton, A. (1989). Mother-child relations and adolescent sexual
attitudes and behavior. Demography, 26, 563-577.
Weiss, R. S. (1979). Growing up a little faster: The experience of growing up in a
single-parent household. Journal of Social Issues, 35, 97-111.
Whitbeck, L. B., Hoyt, D. R., Miller, M., & Kao, M. (1992). Parental support,
depressed affect, and sexual experience among adolescents. Youth and
Society, 24, 2, 166-177.
Whitbeck, L. B., Conger, R. D., & Kao, M. Y. (1993). The influence of parental
support, depressed affect, and peers on the sexual behaviors of adolescent
girls. Journal of Family Issues, 14, 2, 261-278.
Whitbeck, L. B., Simons, R., & Kao, M. (1994). The effects of divorced single
mothers’ dating and sexual attitudes on the sexual attitudes and behaviors of
their adolescent children. Jouirnal of Marriage and the Family, 56, 615-621.
304
Whitbeck, L. B., Simons, R. L., & Goldberg, E. (1996). Adolescent sexual
intercourse. In R. L. Simons (Ed.), Understanding differences between
divorced and intact families: Stress, interaction, and child outcome (pp. 144-
156). Thousand Oaks, CA: Sage.
Wilson, W. J. (1987). The truly disadvantaged: The inner city, the underclass, and
public policy. Chicago: University of Chicago Press.
Wu, L. L., & Martinson, B. C. (1993). Family structure and the risk of a premarital
birth. American Sociological Review, 58, 210-232.
Wu, L. L., & Thomson, E. (2001). Race differences in family experience and early
sexual initiation: Dynamic models of family structure and family change.
Journal of Marriage and Family, 63, 682-696.
Young, E. W., Jensen, L. C., Olsen, J. A., & Cundick, B. P. (1991). The effects of
family structure on the sexual behavior of adolescents. Adolescence, 26, 104,
977-986.
Youniss, J., & Smollar, J. (1985). Adolescent relations with mother, fathers, and
friends. Chicago: University of Chicago Press.
Zabin, L. S., & Hayward, S. C. (1993). Adolescent sexual behavior and childbearing.
Newbury Park, CA: Sage.
Zelnick & Kantner (1980). Sexual activity, contraceptive use and pregnancy among
metropolitan area teenagers: 1971-1979. Family Planning Perspectives, 9,
55-71.
Zelnick, M., Kantner, J. F., & Ford, K. (1981). Sex and pregnancy in adolescence
(Sage Library of Social Research, Vol. 133). Beverly Hills: Sage.

Abstract (if available)

Abstract Using the 1994-1995 National Study of Adolescent Health dataset, I conducted two investigations on the impact of family structure and family influences on sexual activity for white, black, Hispanic, and Asian adolescents. The first investigation determined the best explanation (family transitions, parental supervision, parental attitudes, family support, and peer influence) for the relationships between non-traditional households (unwedded mother, divorced/separated mother, and stepparent households) and sexual activity. The second investigation determined which of these explanations are either directly or indirectly associated with sexual activity. These analyses were conducted on the total sample and within gender, race, and race-gender subgroups.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The link between maternal depression and adolescent daughters' risk behavior: the mediating and moderating role of family

PDF

Parenting practices among non-offending mothers of sexually abused girls and its impact on the abused girls' behavioral adjustment: perspectives from a multigenerational, longitudinal study

PDF

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

PDF

Social network influences on depressive symptoms among Chinese adolescents

PDF

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

PDF

The impact of family factors on the functioning of African-American consumers living with schizophrenia

PDF

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

PDF

Family aggression, prosocial friends, and the risk of dating and friend victimization in late adolescence

PDF

Examining the longitudinal relationships between community violence exposure and aggressive behavior among a sample of maltreated and non-maltreated adolescents

PDF

How does status influence behavior? The impact of achieved and endowed status on consumption patterns

PDF

Intergenerational transmission of values and behaviors over the family life course

PDF

In visible families: gay fatherhood and the politics of family change

PDF

Pearl harbored: race, gender and public memories of Pearl Harbor and 9/11/2001

PDF

Influence of atmospheric processes on the physico-chemical properties of primary particles and their impact on public exposure

PDF

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

PDF

Race, class, gender and the American Dream: a qualitative study of upper working-class African American mothers' situated American Dream meanings, contradictions, and resolution

PDF

Social capital and community philanthropy: the impact of social trust and social networks on individual charitable behavior and community foundation development

PDF

The effects of mindfulness on adolescent cigarette smoking: Measurement, mechanisms, and theory

PDF

Culturally relevant sex education for Black adolescents: a study of Black sexuality educators

PDF

Essays on development economics and adolescent behavior

Asset Metadata

Creator Lo, Thomas I-Ping (author)

Core Title The impact of family structure and family process on adolescent sexual behavior: race and gender variations

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Sociology

Degree Conferral Date 2007-12

Publication Date 10/01/2007

Defense Date 08/21/2007

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

Tag Adolescent,family influence,family process,family structure,gender,OAI-PMH Harvest,Race,sexuality

Language English

Advisor Ransford, H. Edward (committee chair), Biblarz, Timothy J. (committee member), Silverstein, Merril (committee member)

Creator Email loiping@hotmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m838

Unique identifier UC1477477

Identifier etd-Lo-20071001 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-587501 (legacy record id),usctheses-m838 (legacy record id)

Legacy Identifier etd-Lo-20071001.pdf

Dmrecord 587501

Document Type Dissertation

Rights Lo, Thomas I-Ping

Type texts

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

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu