ADOLESCENT MALES SEXUALITY
B. A., National Medical University of Mongolia, 1999
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Arts
i Al i
2004 by Tsetsen Tugsmandal
All rights reserved.
This thesis for the Master of Arts
has been approved
Tugsmandal, Tsetsen (M.A., Sociology)
Adolescent Males Sexuality
Thesis directed by Associate Professor Candan Duran-Aydintug
Most recent statistics show that adolescents are more likely to engage in sexual
intercourse before finishing high school than they were a decade ago. In fact, by the
time they graduate from high school, the majority of American teens report that they
are sexually experienced. Nearly 1 in 12 teens will become pregnant each year and
high levels of unprotected sexual activity put adolescents at risk of contracting AIDS
or other sexually transmitted diseases. Given these facts, adolescents sexual
behaviors and attitudes have received considerable attention from researchers in the
past two decades. Researchers mainly have focused on determinants and
consequences of adolescent sexual behavior, especially high-risk sexual behavior.
Using data from the 1995 National Survey of Adolescent Males, in this study models
predicting adolescent sexual behavior were tested. Results of multiple regression and
logistic regression models showed that some demographic variables such as race, age,
and having worked for pay, education, traditional gender ideology and a family
structure variable called strictness of rules at home were better predictors of the age at
which adolescents males have their first intercourse and whether they ever made
This abstract accurately represents the content of the candidate's thesis. I recommend
I warmly dedicate this thesis to my parents for their support, patience, and
understanding while I was writing this. My father and mother, I would not be here
without you. Your encouragement, support, love, and example have kept me studying
and doing this thesis.
This thesis would not have been possible without the support, patience, and
encouragement of many people. First of all, I would like to sincerely thank to my
great advisor Dr. Candan Duran-Aydintug for all her support, advice, patience, help,
and understanding with me during these past two years. She has motivated me to
study and has been enormous factor in my finishing school. I would also like to
thank Dr. Yili Xu for all his help and guidance. Many thanks to Dr. Susan Allison-
Endriss for her help in the final stages of this thesis. Finally, I want to thank to my
family, my parents, brother, husband and son, for their support and patience. None of
this would have been possible without all of you. Thank you very much.
2. REVIEW OF THE LITERATURE....................................7
Characteristics of Adolescent Males......................7
Race and Ethnicity.................................8
Sexual Attitudes and Behavior.....................10
Neighborhood, Work, Peers, Religion, and School.........11
Family Structure Variables..............................15
Family Process Variables................................19
Fathers are Important...................................29
Data and Sampling.......................................38
Variables and Measurement...............................39
Interpretations of the Coefficients.................49
Interpretations of the Coefficients.................53
Interpretations of the Coefficients.................60
4.1 Descriptive Statistics for Model 1 and Model 2.............................44
4.2 Regression Coefficients for Demographic Variables with Regard to Age of First
4.3 Regression Coefficients for Attitudinal Variables with Regard to Age of First
4.4 Regression Coefficients for Family Variables with Regard to Age of First
4.5 Regression Coefficients for Demographic Variables with Regard to Age of First
4.6 Regression Coefficients for Attitudinal Variables with Regard to Age of First
4.7 Regression Coefficients for Family Variables with Regard to Age of First
4.8 Descriptive Statistics for Model 3.........................................56
4.9 Regression Coefficients for Demographic Variables with Regard to Whether the
Respondent Had Ever Made Someone Pregnant.......................................57
4.10 Regression Coefficients for Attitudinal Variables with Regard to Whether the
Respondent Had Ever Made Someone Pregnant.......................................58
4.11 Regression Coefficients for Family Variables with Regard to Whether the
Respondent Had Ever Made Someone Pregnant.........................................59
The United States Centers for Disease Control and Prevention publishes the
Youth Risk Behavior Surveillance System that provides current information about
adolescent sexual behavior. The Youth Risk Behavior Surveillance System is
conducted every two years with students in grades nine through 12 at high schools
across the country.
According to the 2001 Youth Risk Behavior Surveillance System, 45.6
percent of high school students (48.5 percent of males and 42.9 percent of females)
reported having had sexual intercourse, while 6.6 percent of students reported
initiating sexual intercourse before age 13. Particularly, 60.5 percent of twelfth
graders, 51.9 percent of eleventh graders, 40.8 percent of tenth graders, and 34.4
percent of ninth graders reported having had sexual intercourse. When it comes to
race and ethnicity, 60.8 percent of Black students, 48.4 percent of Hispanic students,
and 43.2 percent of White students reported having had sexual intercourse, while 16.3
percent of Black students, 7.6 percent of Hispanic students, and 4.7 percent of White
students reported having had sexual intercourse before age 13. According to the
result, 33.4 percent of students reported they were currently sexually active.
The 2001 Youth Risk Behavior Surveillance System found following
information regarding sexual partners: 14.2 percent of students (17.2 percent of males
and 11.4 percent of females), especially, 21.6 percent of twelfth graders, 15.2 percent
of eleventh graders, 12.6 percent of tenth graders, and 9.6 percent of ninth graders
reported having had sexual intercourse with four of more partners. Moreover, 26.6
percent of Black students, 14.9 percent of Hispanic students, and 12 percent of White
students reported having had sexual intercourse with four or more partners.
The National Longitudinal Study on Adolescent Health is a national study that
uses total of 12,118 adolescents in grades 7 through 12 and studies adolescent health
behaviors and other factors influencing on their health. Results from the National
Longitudinal Study on Adolescent Health (1997) showed that 17 percent of seventh
and eighth graders, 49.3 percent of ninth through twelfth graders, and 39.9 percent of
female and 37.3 percent of male adolescents in grades seven through 12 reported
having had sexual intercourse.
The National Survey of Adolescents and Young Adults: Sexual Health
Knowledge, Attitudes, and Experiences uses a nationally representative sample of
more than 1,800 youths and examines individuals knowledge and attitudes toward
sexuality. The survey (2003) results indicated that 37 percent of adolescent ages 15
to 17 (42 percent of males and 33 percent of females), 80 percent of young adults
ages 18 to 24 (83 percent of males and 78 percent of females) reported having had
The Alan Guttmacher Institute focuses on the Sexual and Reproductive Health
Needs of American Men and a 2002 study from this Institute found that 42 percent of
men ages 15 to 17 had engaged in sexual intercourse. Of these, 86 percent had
engaged in intercourse in the past year, 47 percent had engaged in intercourse in the
past month, and 36 percent had engaged in intercourse 10 or more times in the past
year. In addition, it had been reported that 75 percent of men ages 18 to 19 had
engaged in sexual intercourse. Of these, 91 percent had engaged in intercourse in the
past year, 61 percent had engaged in intercourse in the past month, and 56 percent
had engaged in intercourse 10 or more times in the past year.
Researchers from the Henry Kaiser Family Foundation and Seventeen
Magazine conducted a national survey of 510 adolescents ages 12 to 17 and together
they examined factors that impact adolescents decisions about sex and relationships.
The results (2000) indicated that among adolescents ages 15 to 17 who had not had
sex, 83 percent said it was because they were worried about pregnancy, 74 percent
said it was a conscious decision they had made to wait, 73 percent said they were
worried about sexually transmitted diseases, 64 percent said it was because they
worry about what their parents might think, 63 percent said it was because they
have not met the right person, 63 percent said they felt they are far too young, and
52 percent said it was because of their religious beliefs. Furthermore, the
researchers reported that among participants ages 15 to 17 who had had sex, 51
percent said when they had sexual intercourse for the first time it was because they
met the right person, 45 percent said it was because the other person wanted to,
32 percent said it was because they were just curious, 28 percent said it was
because the hoped it would make the relationship closer, and 16 percent said it was
because many of their friends already had.
The National Campaign to Prevent Teen Pregnancy compiled a summary of
findings from two nationally representative surveys of 510 adolescents ages 12 to 17.
Their summary (2003) reported that 62 percent of adolescent ages 12 to 14 and 65
percent pf adolescents ages 15 to 17 said they have had helpful conversations with
their parents about sex. Moreover, according to this summary, 61 percent of
adolescents surveyed said the media have provided them with information or advice
about sex in the past month, 57 percent said their friends have, and 55 percent said
their parents have.
By looking at a sample of results from the 1991, 1995, and 2001 Youth Risk
Behavior Surveillance, one can see that 45.6 percent of students reported having ever
had sexual intercourse in 2001, 53.1 percent in 1995 and 54.1 percent in 1991.
Moreover, 6.6 percent of students reported having initiated sexual intercourse before
age 13 in 2001 compared to 8.9 percent in 1995 and 10.2 percent in 1991. They also
reported that 14.2 percent of students reported having had sexual intercourse with
four or more partners in 2001 compared to 17.8 percent in 1995 and 18.7 percent in
1991. Close to 34 percent (33.4) of students reported being currently sexually
active (defined as having had sexual intercourse with one or more partners in the
three months preceding the survey) in 2001 compared to 37.9 percent in 1995 and
37.5 percent in 1991.
The National Campaign to Prevent Teen Pregnancy recently released Fourteen
and Younger: The Sexual Behavior of Young Adolescents. The report found that
approximately one in five adolescents has engaged in sexual intercourse before his or
her fifteenth birthday. In addition, the report found that boys who are 14 and younger
are slightly more likely to have had intercourse than girls of the same age.
Nowadays, adolescents unsafe sexual intercourse practices and their related
negative outcomes are seriously discussed issues in our society. Therefore,
researchers study adolescent sexual activity from all aspects and share the social
implications of their findings.
Adolescents tend to engage in short sexual relationships that are serially
monogamous and that increase their exposure to multiple sexual partners
and, subsequently, increase their risk of contracting HIV infection and
other negative consequences of sexual risk behavior (Overby and
Kegeles, 1994 cited in Miller, Forehand, and Kotchick, 1999:86).
The sexual behavior among adolescent males is becoming an important area
of research. According to results from 1988 National Survey of Adolescent males,
three-fifths of adolescent males aged 15-19 have had sexual intercourse. The data
show that among sexually experienced adolescent males the mean number of partners
in the last 12 months is 1.9 and the mean frequency of intercourse in the last four
weeks is 2.7 times (Sonenstein, Pleck and Ku, 1991). Adolescent males are more
sexually active than females (Coales and Stokes, 1986; lessor and lessor; 1975;
Robinson and Jedlicka, 1982); however, research about adolescent males sexual
behavior is scarce (Ku, Sonenstein and Pleck, 1993). Adolescent males, minority
adolescents, and adolescents of lower socioeconomic status have higher risk of sexual
intercourse (Kann, Warren, Harris, Collins, Williams, Ross, and Kolbe, 1996; Leigh,
Morrison, Trocki, and Temple, 1994; Seidman and Reider, 1994). For example, 81%
and 62% of Black and Hispanic males in high school and 67% and 53% of Black and
Hispanic females in high, school report having engaged in sexual intercourse (Miller,
Forehand, and Kotchick, 1999).
Given the importance of this problem in our society, in this study, I aim to
answer three research questions.
1) What factors predict the age at which adolescent males start dating
the first time;
2) What factors predict the age at which adolescent males start having
sexual intercourse for the first time;
3) What factors predict whether adolescent males ever make someone
REVIEW OF THE LITERATURE
Even though many studies have examined teenage girls sexual activity, here
the focus is on male adolescents. In this literature review the focus is on boys
sexuality including results from studies dealing with teen fathers.
Characteristics of Adolescent Males
Personal factors including socio-demographic and background characteristics,
sexual attitudes and behavior are important factors affecting adolescent males sexual
Social-psychological research has adopted two broad theoretical approaches to
the male gender role: 1) trait perspectives, focusing on the sources and consequences
of the extent to which men actually have characteristics culturally defined as
masculine, and 2) normative perspectives, deriving from the social constructionist
view of gender roles, which concentrate on the nature and consequences of the
standards by which cultures define masculinity (Pleck, Sonenstein, and Ku, 1993).
Using data from the 1988 National Survey of Adolescent Males, Pleck, Sonenstein,
and Ku (1993) found that traditional masculinist ideology had correlated with
negative implications in heterosexual relationships for the level of intimacy of
adolescent males. Furthermore, they reported that adolescent males have lower
emotional investment in any one sexual relationship; therefore they had more sexual
partners in the last year and a less close relationship at last intercourse (Pleck,
Sonenstein, and Ku, 1993).
Peplau, Rubin, and Hill (1977) found that male college students are less likely
to cite lack of emotional commitment as a reason for abstaining from sex and they
reported that loss of virginity by a young man was less closely related to his love and
commitment toward his partner than loss of virginity by a woman.
Race and Ethnicity
Analyzing of a nationally representative survey of 1,880 15 to 19 year old
men, Ku, Sonenstein and Pleck (1993) reported that generally Black youth have
intercourse at a younger age and Hispanic males have greater chance of early
intercourse than White non-Hispanic youth. They found that Black adolescents
engaged in first intercourse at significantly younger ages at first intercourse than
White or Hispanic youths. Twenty percent of Black males reported that their first
intercourse was before age 13, whereas three percent of Whites and four percent of
Hispanics said having intercourse by that age. Thirty five percent of the Black males,
seven percent of Whites and six percent of Hispanics had had intercourse at least once
before age 14.
Using longitudinal data from a nationally representative sample of adolescents
aged 11 to 17, Lauritsen (1994) found that family structure and neighborhood
characteristics are relevant for race differences in the sexual activity of females, but
not of males. Moreover, Lauritsen reported that social control variables such as
attachment to family, educational aspirations were found to be important predictors of
sexuality among white adolescents; however, strain variables such as perceived
inability to achieve educational goals were found to be significant for the sexual
activity among black females. Social control and strain factors were not accountable
for explaining the sexual behavior of black male youth (Lauritsen, 1994).
Miller, Christen and Olson (1987) found that among adolescents who were
frequent church attenders and Mormons, self esteem was significantly and negatively
related to permissive sexual attitudes and behavior while self-esteem was positively
related to sexual intercourse experience among adolescents who believed that
premarital sex was usually or always right. On the other hand, self-esteem was
negatively related to sexual intercourse experience among those who believed it was
wrong. Religious participation and church attendance were related to sexual behavior
of college students (Davids, 1982; Notzer, Leuvan, Mashiach, and Soffer, 1984;
Reiss, 1967; Reiss and Miller, 1979; Sack, Keller, and Hinkle, 1984; young, 1982)
and younger teenagers (Harris and Associaty, 1986)
Sexual Attitudes and Behavior
In a convenience sample, Miller and Olson (1988) found a strong relationship
between teenagers attitudes about sexual intercourse before marriage and their
reported intercourse experience. Moreover, they stated, over 100 respondents who
reported sexual intercourse experience said that it was usually or always wrong for
unmarried teenagers to have sex. Conversely, about one third of the virgins stated
that it was neutral or all right for teenagers to have sex before marriage, but they had
not. These attitude behavior discrepancies suggest that adolescent sexuality is indeed
complex (Miller and Olson, 1988:198).
Luster and Small (1994) surveyed a nationally representative sample of 2,567
teenagers and examined factors that distinguish between sexually active adolescents
who are at risk for pregnancy and sexually transmitted diseases, and sexually active
teens who are at lower risk for those outcomes. They found that high-risk males were
more likely to have lower GPAs, higher levels of suicidal ideation, and more
frequently alcohol use. They also reported that 20% of the high-risk males were
sexually abused, 23% of the high-risk males had experienced physical abuse. Only
2% males of the low-risk group and sexual abstainers group had been sexual abused,
and 14% of males in the low risk group and 7% of males in the sexual abstainers
group had experienced physical abuse.
Neighborhood. Work. Peers. Religion, and School
Other factors such as neighborhood, work, peers, religion, and school
contribute to adolescent males sexual activity.
Small and Luster (1994) reported that particular aspects of the adolescent
ecology that may be associated with teen fatherhood include broader contextual
factors such as economic, neighborhood, and family background, as well as more
proximate factors such as family relationships, school achievement, and behavioral
characteristics of the adolescent. Small and Luster studied the relationship between
adolescent sexual activity and an ecological, risk factors including history of physical
abuse, neighborhood monitoring, and the adolescents attachment to school. Using
data from 2,168 adolescents, who were enrolled in the 7th, 9th, and 11th grades in a
midsize Southwestern city, they tested a cumulative risk factor model, which helped
to predict and understand heterosexual adolescent sexual behavior. The result
supported the concept of cumulative risk factor model and they found a strong
relationship between the number of risk factors and adolescent sexual activity for
both male and female youths (Small and Luster, 1994). They reported that similar to
previous studies, poor school performance, involvement in a committed relationship,
low parental education, and frequency of using alcohol were related to sexual
experience. Moreover, other risk factors such as history of sexual abuse, the
perception of limited economic opportunities in the future, history of physical abuse,
lack of school attachment, low neighborhood monitoring, parental monitoring, and
permissive parental values about teenage sexual behavior were also linked to
adolescent sexual intercourse. Grade point average was important factor to predicting
adolescent sexual activity for both males and females (Small and Luster, 1994).
Ku, Sonenstein, and Pleck (1993) examined how certain neighborhood,
family, and personal characteristics affect the premarital behaviors of teenage males
in the U.S. Using data from 1988 National Survey of Adolescent Males, a nationally
representative survey of 1,880 never married, noninstitutionalized males 15 to 19
years old and the 1980 census, they compared the effects of related personal and
neighborhood traits in multilevel analysis, including employment, income, education,
welfare receipt, family composition, race, and ethnicity. They reported that personal
factors were generally more powerful predictors than neighborhood factors. Their
results also indicated positive relationships between financial resources, measured by
family income and hours worked in the last year, and sexual activity, measured by
number of partners and frequency of intercourse in the last year. Young men who
worked more hours were more likely to have made someone pregnant, while higher
neighborhood unemployment rates were independently associated with greater risk of
Using the 1993 wave of the Philadelphia Teen Survey, Teitler and Weiss
(2000) examined the effects of neighborhood and school environments on youths
sexual intercourse. They found that neighborhood effects on sexual behavior exist
only to the extent that they determine the type of school an adolescent attends; school
effects are primarily a function of racial composition and school type, and school
types are associated, to a large extent, with school-level behavioral norms(pl 12).
Upchurch, Aneshensel, Sucoff, and Levy-Storms (1999) examined the
influence of neighborhood and family contexts on the transition to first sex. Using
data from a longitudinal survey of stress and mental health conducted on a
representative sample of 870 adolescents, ages 12 to 17 in Los Angeles, collected
between October 1992 and April 1994, they investigated neighborhood and family
context within adolescents gender and race, including neighborhood typologies,
experiential neighborhood, family interaction, family structure, and individual
characteristics. They found that not only neighborhood socioeconomic status and
race influenced on adolescent sexual activity, but also structural attributes were
significant variables. They also reported that living with single parent or
reconstituted families and parental overcontrol were important for high risk of
adolescent sexual behavior. In terms of race and ethnicity, Hispanic girls had lower
sexual risk comparison to White female youths, while Black and Hispanic male
youths had higher sexual risk compared to their white comerparts.
South and Baumer (2000) tested Wilsons hypothesis and determined whether
the effects of neighborhood socioeconomic disadvantages on risk of adolescent
premarital childbearing. According to Wilson (1987), racial differences in rates of
premarital childbearing (as well other social dislocations) can be attributed to racial
differences in the types of neighborhoods inhabited by black and white women
(pl381). Using the data from National Survey of Children and the 1980 U.S. census,
they examined the ability of racial differences in neighborhood attributes to explain
racial differences in premarital childbearing and the mechanisms through which
neighborhood conditions influence adolescent premarital childbearing. The National
Survey of Children is a nationally representative survey of children and black
children were oversampled. They found that over one-third of the positive effect of
neighborhood socioeconomic disadvantages on the timing of young womens first
premarital birth can be attributed to the attitudes and behaviors of peers and to young
womens more tolerant attitudes toward unmarried parenthood in distressed
communities (South and Baumer, 2000:1379). Higher rates of residential mobility in
poor neighborhoods explained a smaller proportion of the effect of neighborhood
socioeconomic status on adolescent premarital childbearing (South and Baumer,
2000). Their results also showed that adolescents educational aspirations, school
attachment, and parental supervision were not significant variables to mediate the
impact of community disadvantage on adolescent childbearing; however, racial
differences in neighborhood quality were extremely significant (South and Baumer,
Miller, Norton, Curtis, Hill, Schvaneveldt, and Young (1997) investigated
relationship between adolescents perceptions of their relationship with their parents,
peers, and other antecedents and their age at first sexual intercourse. Using data from
the National Survey of Children, a three wave (1976,1981, and 1987) longitudinal
study of a nationally representative household survey of 1,145 children who were
living in the contiguous United States, they separately analyzed the timing of sexual
intercourse foe males and females. They found that significant variables predicting
age of first sex among males were age of first date, dating often, number of friends
perceived to have had sex at age 16, being Black, having parents undergo marital
changes during the childs school years, and fighting at school. Except from fighting
at school and dating frequency, all of above mentioned variables and additional
variables such as menarche, parents education, mothers coercive behavior and love
withdrawal, and attitudes about attending religious services were the most significant
predictors of age at first sexual intercourse among females.
Family Structure Variables
The family structure factors including parental marital status, family income,
family education are significant predictors for adolescent sexual behavior.
The family context plays an important role in the sexual socialization of
adolescents (Small and Luster, 1994). Mille, Forehand, and Kotchick (1999)
demonstrated that family structure variables have received less attention than process
variables. Yet,, there are several studies indicating that family structure variables
cannot be ignored. For example, adolescents early sexual activity and the frequency
of sexual intercourse related to single parenting or the absence of a father (Hogan and
Kitagawa, 1985), and low educational attainment of parents (Flick, 1986; Hayes,
1987; Small and Luster, 1994; Udry and Billy, 1987).
Another family structure factor, parental marital status, is also important
predictor for explaining adolescent development. As Salem, Zimmerman, and Notaro
stated (1998) The rising number of children who now live in single-mother
households has led to considerable interest in the effects of family structure on
adolescent development (p331).
As a matter of fact, results showed that living with a single mother has been
associated with childrens delinquency (Dombusch et al., 1985; Ensminger, Kellum,
and Rubin, 1983; Sampson, 1987; Steinberg, 1987), alcohol and substance use
(Brook, Whiteman, and Gordon, 1985; Castro, Maddahian, Newcomb, and Bentler,
1987; Covey, and Tam, 1990; Murray, Roche, Goldman, and Whitbeck, 1988; Stem,
Northman, and Van Slyck, 1984), lower self-esteem (Harper and Ryder, 1986; Parish,
1991; Parish and Taylor, 1979), psychiatric problems (Barbarin and Soler, 1993),
earlier initiation of sexual intercourse (Newcomer and Udry, 1987) and leaving
school before graduation (Zimilies and Lee, 1991).
Several researches have demonstrated that not living with both biological
parents is a predictor of early sexual intercourse among adolescents (Hogan and
Kitagawa, 1985; Miller, Higginson, McCoy, and Olson, 1985; Mott, 1984; Thornton
and Cambum, 1983; Zelnik, Kantner, and Ford, 1981; Pleck, Sonenstein, and Ku,
Newcomer and Udry (1987) examined the relationship of change in marital
status to subsequent coital behavior of white adolescents who were virgins at the first
interview. In 1980, they interviewed 1,405 youths at the first time and 82% of them
were interviewed at the second time in 1982. Using data from a panel study of white
virgin adolescents, they studied household income, others education, age of child,
age of mother, importance of religion to the mother and to the child, premarital sexual
permissiveness of mother and child, mother and children reports of mothers control
over child activities, childrens feelings of closeness to mother, and child; subjective
expected utility from coitus. They found that living with a single mother has been
associated with childrens earlier initiation of sexual intercourse.
Jemmott and Jemmott (1992) studied African American adolescent males and
they reported that family structure was related to condom use and chances of
fathering a child; however, it was not related to the frequency of sexual intercourse or
the number of sexual partners.
Flewelling and Bauman (1990), in a study of family structure effects among
early adolescents, demonstrated that family structure was related to substance abuse
and initiation on frequency sexual intercourse for both Whites and non Whites, yet
the behavior of non Whites were less affected by family structure than it was the case
Family structure has been found to be unrelated to precocious sexual behavior
for African American youths (Ensminger, 1990).
Relatedly, Salem, Zimmerman, and Notaro (1998) examined the relationship
between family structure and a variety of psychosocial outcomes among African
American adolescents. They identified family structure variable as youth with whom
they lived and their ages and they developed five family types such as living with
single mother, both biological parents, biological mother and stepfather, mother with
extended family members, and extended family members only. They found that
family structure was not related to psychosocial outcomes for the African American
Miller, Forehand, and Kotchick (1999) used 907 Black and Hispanic families
and investigated family structural variables such as family income, parental
education, and maternal marital status, and process variables such as maternal
monitoring, mother-adolescent general communication, mother-adolescent sexual
communication, and maternal attitudes about adolescent sexual behavior as predictors
of adolescent sexual behavior. They found that family structure variables failed to
predict adolescent sexual behavior, while family process variables predicted
adolescent sexual behavior and risk.
Ramirez-Valles, Zimmerman, and Newcomb (1998) used sociological
theories of financial deprivation and collective socialization and they tested a model
of the relationships among neighborhood poverty, family structure and social class
position, parental involvement, prosocial activities, race, and gender as predictors of
sexual risk behavior among youth. Using a cross-sectional sample of 370 sexually
active high school students in ninth grade from a Midwestern city, they conducted
face-to-face interviews with their respondent. The sample consisted of 56% males
and 44% females, and 86% of them were African-American youths, while 14% were
white youths. They found that family structure predicted sexual risk behavior among
youths through neighborhood poverty, parental involvement, and prosocial activities.
Moreover, family class position was a predicting factor through neighborhood
poverty and prosocial activities.
Family Process Variables
Family process variables include parenting, parental attitudes, discipline,
control, involvement, and etc.
Parental monitoring, supervision, and control have been found to delay or
prevent problem behaviors among adolescents (Chilcoat and Anthony, 1996;
Forehand, Miller, Dutra, and Chance, 1997; Vazsonyi and Flannery, 1997).
Hogan and Kitagawa (1985) studied over 1,000 black adolescents between the
ages of 13 and 19 and they reported that a strong relationship between parental
control of daughters dating experiences and the daughters sexual activity and
Using survey data from 836 students inl983 and 1,587 students in 1984 aged
15 to 18, Miller, McCoy, Olson, and Wallace (1986) examined how parental
discipline and control were related to adolescent sexual attitudes and behavior. They
Adolescents perceptions of parental strictness and rules show a
curvilinear relationship to their sexual attitudes and behavior; sexual
permissiveness and intercourse experience was highest among
adolescents who viewed their parents as not being strict at all or not
having any rules, lowest among those who reported that their parents
were moderately strict, and intermediate among teens who perceived
their parents to be very strict and to have many rules. The parents own
reports of their dating rules were less strongly related to their adolescent
sons and daughters reports of sexual attitudes and behavior (Miller,
McCoy, Olson, and Wallace, 1986:503).
Amato and Rivera (1999) used data from the 1987-1988 National Survey of
Families and Households (n=994) and they tested the hypothesis that positive father
involvement is associated with fewer behavior problems in children. The National
Survey of Families and Household consist of a cross-sectional national probability
sample of 13,017 respondents. The researches used both fathers and mothers reports
including fathers reports of involvement with children and mothers reports of
childrens behavior problems. They found that positive paternal and maternal
involvement were independently and significantly associated with childrens behavior
problems and results were similar for biological fathers, stepfathers, White fathers,
Black fathers, and Latino fathers.
Roberts, Kline, and Gagnon (1978) interviewed parents with three-to-eleven-
year-old children about their attitudes and actual behavior in the area of sex
education. They found that 75% of the mothers and 50% of the fathers had discussed
pregnancy with their children, yet only 15% of the mothers and 8% of the fathers had
provided information about sexual intercourse.
In another study from the late 1970s, Elias (1978) reported that about two-
thirds of high school girls had discussed sexuality with their mothers, and only 2% of
the girls had discussed sexual issues with their fathers; however, only a quarter of the
boys had discussed sexuality with their fathers, and two-thirds had never received
sexual information from either parent.
Koblinsky and Atkinson (1982) surveyed parents (82 mothers and 70 fathers)
of preschoolers about their future plans fore educating their children regarding 15
aspects of human sexuality. They found that both mothers and fathers expected to
discuss a range of sexual topics with their children, but fathers were less likely to
involve in their daughters sex education than mothers. Luster and Small (1994)
found that sexually active male and female adolescents, who had multiple partners
and failed to use contraception consistently, perceived their parents as less supportive
than their lower risk peers who consistently used birth control within a monogamous
relationship. Rodgers (1999) examined parenting behaviors related to sexual risk-
taking behaviors among sexually active adolescent males and females and reported
that although parental communication about sexual matters and parental support
were not directly associated with sexual risk taking, an interaction effect between
parental support and communication was revealed for sexually active adolescent
males. (pl06). Mueller and Powers (1990) reported that adolescents who describe
their parents as attentive and supportive communicators report less sexual activity
during junior high school and high school.
Several studies suggest that parental attitudes regarding adolescent sexual
behavior have related to adolescent sexuality, for example, adolescents who perceived
that their parents disapproved of teen sexual activity have been less likely to be
sexually active (Jaccard et al., 1996; Metzler et al., 1994). On the other hand,
permissive parental attitudes have been related to earlier adolescent sexual activity
(Small and Luster, 1994; Thornton and Cambum, 1987).
Davis and Friel (2001) examined how a variety of family structures, including
cohabiting and lesbian families, and family context, specifically maternal
involvement as well as maternal sexual attitudes, are related to adolescent sexual
behavior. Using the 1995 National Longitudinal Study of Adolescent Health, a
nationally representative sample of students from schools, they tested hypotheses by
modeling the family environment. The participants were 20,745 students in grades 7
through 12 completed in-home interviews (12,105 adolescents from the core sample
and 8,640 from oversampled subgroups). They found that the family context, more
specifically the mother-child relationship, their level of interaction, and the mothers
attitudes toward and discussion of sex was related to adolescent sexuality, while with
the exception of girls in single-parent families, family structure was not associated
with adolescent sexual behavior. They also reported that when looking at the patterns
of sexual partnership for those teens who are sexually active, family status, the
mother-child relationship, and maternal sexual attitudes have no impact on the sexual
partnership of boys and explain little in terms of girls sexual partnering (Davis and
Meschke, Bartholomae, and Zentall (2000) focused on current trends in
adolescent sexual behavior and the associations of family factors and processes with
adolescent sexuality. They also reviewed current adolescent sexuality programs with
a family component. Meschke, Bartholomae, and Zentall reported that the frequency
and number of adolescents engaging in sexual activity and teen pregnancy rates have
decreased, whereas the percentage of adolescents initiating intercourse at an early age
has increased. American adolescents have higher rates of unprotected sex and
sexually transmitted infections than adults. The teen pregnancy rates are highest of
any developed nation. In terms of the relation between parenting and adolescent
sexual outcomes, they found that adolescent sexuality was associated with parent-
adolescent communication, parental values, parental monitoring and control of
adolescent activities, and closeness of parent-adolescent relationship. However, the
quality of parent-adolescent relationships significantly influences the amount of
parental influence on their adolescents sexual behaviors. In term of reviewing
current adolescent sexuality programs, they reported that 19 adolescent sexuality
programs with a parent component supported the incorporation of theory and the
ecological model in program design and evaluation.
Wemer-Wilson (1998) studied the effect of sexual attitudes on sexual
behavior by using a nonrandom availability sample of 1,587 public high school
students and 1.372 parents. Three regression models were developed separately for
females and males. The regression models run separately for individual adolescent
characteristics (including age, gender, locus of control, self-esteem, and religious
participation) and for family characteristics (including number of siblings, number of
parents in home, communication with mother and father, family strengths, parent
contribution to sexuality education, parental discussion of sexual values, and the
sexual attitudes of mother and father) as predictor variables. Finally, a multiple
regression model was developed that included both individual and family factors.
The researcher found that the integrated model was more significant than separate
models on males and females and in that integrated model, religious participation was
the strongest predictor. Results also indicated that in the integrated model, males
were influenced by more individual factors and females were influenced by more
family factors (Wemer-Wilson, 1998).
Miller (2002) investigated family influence on adolescent sexual and
contraceptive behavior. He reported that parent-child closeness or connectedness,
and parental supervision or regulation of children, in combination with parents
values against teen intercourse or unprotected intercourse decreased the adolescent
pregnancy risk. However, living with a single parent, in a lower socio-economic
status family, having older sexually active siblings or pregnant/parenting teenage
sisters, being a victim of sexual abuse, and residing in disorganized/dangerous
neighborhoods increased adolescent pregnancy risk. Miller (2002) emphasized that
because Of methodological complexities, findings on parent-child communication and
the risk of adolescent pregnancy were less conclusive.
Jaccard, Dittus, and Gordon (1996) surveyed 751 unmarried black male and
female adolescents and their mothers to examine how maternal disapproval of
premarital sex, maternal discussions of birth control, and the quality of the parent-
child relationship influenced adolescent sexual behavior. They found that each of
these maternal variables was relate to whether the adolescents had intercourse, how
frequently they had intercourse, and how consistently birth control was used.
Whitbeck, Simons, and Kao (1994) also studied mothers and adolescents,
focusing on the dating behaviors of divorced mothers, and found significant
relationship between maternal variables and adolescent sexual behavior. Their
sample included 210, white, predominantly rural, eight and ninth grade female and
male adolescents and their divorced mothers, and focused on the effect of mothers
dating behaviors and sexual permissiveness on adolescents sexual activity and sexual
permissiveness. The effects differed by gender. Sons behaviors were directly
affected by their mothers behaviors and their attitudes were unaffected by mothers
attitudes. Daughters sexual behaviors were indirectly influenced by mothers
Perkins, Luster, Villaruel, and Small (1998) examined the effects of a set of
risk factors for sexual activity, including individual, family, and extrafamilial
variables, on a sample of 15,362 female and male predominantly white adolescents,
aged 12-17. The risk factors that significantly predicted sexual activity included age
suicide ideation, grade point average, alcohol use, time spent home alone, physical
abuse, sexual abuse, religiosity, associating with a negative peer group, and low
school attachment. Time spent alone was the only family variable that was
significant; family support and parental monitoring were also tested and found to be
insignificant predictors of sexual activity.
Udry and Billy (1987) measured the effect of mothers control behavior,
mothers religiosity, and mothers permissiveness towards premarital sex and found
no relationship to the subsequent sexual behavior of their white male adolescent
children. Instead, white males initiation of sexual activity was guided by the
motivational effects of hormones and social attractiveness.
Hovell, Sipan, Blumberg, Atkins, Hofstetter, and Kreitner (1994) constructed
a developmentally based sexual activity scale to examine Latino and Anglo
adolescents sexual behavior in a more detailed light than is afforded by the usual
dichotomous yes/no variable. The researchers studied many background and
family independent variables, including age, gender, ethnicity, mothers education,
religiosity, acculturation, and single-parent status; parent (mother) models of love,
affection, expression of positive relationships, and acceptance; adolescents belief
that mother had premarital sex; communication with family and mother about sex;
and parent rules about sex and dating. They found that parent discussion of sexual
issues had no effect, and that these discussions happened rarely and produce
discomfort when they did happen. Living in a single parent household also had no
effect with this sample, The implicit acceptance of sexual behavior communicated by
adolescents perception that their mother had premarital sex was correlated with
adolescent sexual activity. The variables associated with less sexual activity included
conservative maternal attitudes and adolescent compliance with rules.
Randal Day (1992) analyzed data from the National Longitudinal Survey of
Market Experience of Youth to examine how proximate influences, such as attitudes,
beliefs, and family, and distal influences, such as community and culture affect the
transition to first sexual intercourse among a large sample of 14-to 21-year-old white,
black, and Hispanic males and females. While there were race and ethnicity
differences, the general findings were that women were most affected by religion,
locus-of-control, occupational desire, and self-esteem, while men were most affected
by family and community. The effect of fathers on male sexual activity had some
unexpected twists in this study, and was heavily mediated by the age and ethnicity of
the adolescent. Younger teen Chicano and Latino males were less likely to initiate
sexual activity when they lived with biological father, while older teens in this group
were unaffected. Older Chicano males were affected by the presence of a stepfather,
delaying their transition to sexual intercourse. Black males and older white males
were more likely to have sex if they lived with a biological father, while younger
white males followed the young Chicano and Latino pattern and were less likely to
have sex if they lived with a biological father. In addition, Chicano and white males
likelihood of having sex increased as their fathers; education increased, and black
males were similarly affected by mothers education.
An older study by Floyd and South (1972) supported the notion that family
influence lessens as adolescent age. In their study of white sixth, eight, tenth, and
twelfth graders, Floyd and South found that both females and males became
increasingly oriented to their peers, rather than their parents, until the tenth grade,
when males leveled off and females became less peer oriented.
In 1999, Whitbeck, Yoder, Hoyt, and Conger completed a developmental
study of the predictors of early adolescent sexual activity among white, rural 8th, 9th,
and 10th graders. They found that the best predictors were age, opportunity through
steady dating, sexually permissive attitudes, association with delinquent peers, and
alcohol use. Mothers warmth and supportiveness had no effect, while mother
rejection increased the likelihood of early sexual activity. Similar to Day (1992) and
Floyd and South (1972), Whitbeck et al. noted that one family variable, mother
monitoring, changed in effect and quality as the adolescents aged; monitoring
decreased intercourse for 8th and 9th graders, and increased the likelihood of
intercourse for older participants.
Fathers are Important
Several researchers have studied adolescent females engaging in sexual
activity and relationship between mothers influence and adolescent females
sexuality. However, there is a lack of research on male adolescent sexual behavior,
especially, regarding fathers influence on adolescent males sexual activity.
Moreover, research on fatherhood has only received scholars attention in recent
years. Studies on fathers used to be only on divorced fathers; therefore, there is luck
of information of fathers influence on male adolescent sexuality in intact families.
According to Hanson and Bozett (1987), Fathers, fathering, and fatherhood
have emerged since the 1970s as a topic of major interest to parents, practitioners,
educators, and policymakers (p333). Along the same lines: As women become
increasingly committed to their work careers, men may find it harder to avoid a larger
role at home and may be required to participate more in parental responsibilities
(Barnett and Baruch, 1988 cited in Harris and Morgan, 1991:532).
Most research has emphasized the roles of mothers within their families
because mothers are understood as the primary socializers and they are more likely to
have custody of their children after a divorce (Seltzer, 1991).
Some studies show that fathers with more children spend more time with
children (Barnett and Baruch, 1987; Elder and Bowerman, 1963; Nock and Kingston,
1988). For example, Elder and Bowerman (1963) studied the effects of family size
and gender composition on paternal involvement in rule setting and discipline. They
reported that fathers have more active roles in large families that included one or
more boys. Yet, they found that the lowest level of father involvement occurred in
families with all female children.
Cooksey and Fondell (1996) highlighted that Much less attention has been
paid to the involvement of men with children, over and above their traditionally
ascribed role as financial provider. Yet fathers, too, have important contributions to
make in childrens lives. Most children still live with a biological father or a
stepfather for an at least part of their childhood (Mott, 1990), and the presence or
absence of a father is only a single dimension along which a father-child interaction
can be measured (p693).
Mott (1990) found that African American fathers more likely to be absent
from the home than White fathers; however, nonresident African American fathers
more likely to live near and visit their children than White nonresident fathers. Lamb
(1986) found that regardless of where fathers live, children seem better off in terms of
cognitive, sex-role, and psychosocial development when they have a close and
supportive relationship with their fathers. Seltzer (1994) reported that after divorce
fathers and children have weaker relationship and Goldscheider, 1990; Spitze and
Miner, 1992 have predicted that divorced fathers will have weaker relations with their
adult children than will continuously married fathers.
Living with fathers is not enough for children, but also quantity and quality of
spending time is important for children. According to Montemajor (1982),
Typically, fathers spend little time with their adolescent children since they are tired
at the end of the work day and the children are busy with their peers, homework, or
other independent projects. What time is spending together is usually spent in passive
parallel activities such as television viewing (Montemajor, 1982; cited in Hanson
Harris and Morgan (1991) focused on intact families and examined cross-
sectional differences in fathers involvement with their adolescent children. Using
National Survey of Children, they examined both a behavioral and affective
dimension of father involvement and asked, What individual or family factors
promote greater father involvement in parenting? They found that fathers are more
involved with sons than with daughters and daughters with brothers are advantaged
relative to other girls.
Amato and Gilbreth (1999) employed meta-analyses of 63 studies dealing
with nonresident fathers and childrens well-being. They found positive association
between childrens well-being and fathers payment of child support and authoritative
parenting; however, the frequency of contact with nonresident fathers was not
associated with child outcomes in general. They also reported that feeling of
closeness and authoritative parentings were positively related to childrens academic
success, yet negatively related to childrens externalizing and internalizing problems.
Rob Palkovitz (2002) wrote in his book titled Involved Fathering and Mens
Adult Development about how fatherhood change, develop, and contribute their
lives and relationship with others including social and work domain. Using
qualitative methods, particularly, using interview schedule, he interviewed 40 fathers
between the ages of 20 and 45 in an in-depth fashion and he analyzed the interviews
to understand the relationship between parenting and adult development. He reported
that stories of 40 different fathers indicated, Fathering had changed them-mostly for
the better (p264). He concluded his book by emphasizing For some years now, it
has been surmised that good fathering is good for child development outcomes, good
for the mothers of children, and good for communities where involved fathers live.
Now there is qualitative data to support theoretical and anecdotal accounts so that it is
possible to say with confidence that good fathering is perceived by fathers to be good
for mens adult development (p265).
Catherine Tamis-Le Monda and Natasha Cabrera (2002) wrote in a book titled
Handbook of Father Involvement about father involvement from all aspects
including sociological, anthropological, evolutionary, economic perspectives.
According to Tamis-Le Monda and Cabrera, the most important determinant of father
involvement is father residency and because resident fathers generally spend more
time with their children, father residency considers as the quality and quantity of
father involvement (p604). They reported, Father involvement affects infant-father
attachment, childrens peer relations and friendships, academic achievement, and
social development (p601). Moreover, they demonstrated that father involvement
was complex and it related to other factors such as cultural background, social status,
education, income, and family configuration.
Using data from a nationally representative sample of families and households
in the United States, Cooksey and Fondell (1996) investigated how fathers spent time
with children, the effect of family structure on paternal time, and whether or not
family structure and the time that fathers spend impacts their childrens well being.
Single fathers are quite involved in their childrens lives and are
more likely to engage in a variety of activities (not related to sharing
meals) than fathers in more traditional family settings. Fathers in
father-stepmother families also report spending more time with their
children than fathers in families with a father and a biological mother
They also reported stepfathers are significantly less likely to undertake
activities with their stepchildren than biological fathers are with theirs (p706).
Even though this studys focus is not on teen fathers, this topic is still related
to adolescent males sexual behavior.
Thomberry, Smith, and Howard (1997) studied early risk factors for the
likelihood of becoming a teen father. They found that teen fatherhood related to
social class, educational performance, precocious sexual activity, and drug use,
especially, they reported that teen fatherhood was strongly related to the accumulation
of risk factors across many domains.
Using data from the High School and Beyond Survey, Hanson, Morrison, and
Ginsburg found (1989) that being black, going steady, and having unorthodox views
about parenting outside of marriage increased risk of teen fatherhood.
Marsiglio (1987) examined data from the National Longitudinal Survey of
Youth and found that teenage fathers were more likely to be black, to be
fundamentalist, to have lived in a single-parent household, and to have parents with
low education levels.
Lerman (1986) examined the National Longitudinal Survey of Youth and
found that males who living in high unemployment levels counties were more likely
to become married teen fathers. He also reported that higher family income was
related to a decreased risk of teen fatherhood. Lerman (1993) found that
unemployment is associated with modest increases in unwed teen fatherhood.
Using data from the National Child Development Study of Great Britain, Hale
and Blankson (1994) examined social risk factors for becoming a father at a young
age. They found that family financial hardship influence on risk for early fatherhood.
They also reported that boys who older siblings were at risk for fathering a child at
In the past literature, researchers studied background variables (Ku,
Sonenstein, Pleck, 1993; Lauritsen, 1994 and etc), individual variables (Miller,
Christen, and Olson, 1987; Miller and Olson, 1988; Luster and Small, 1994 and etc),
and even some of the family structure and process variables (Newcomer and Udry,
1987; Hogan and Kitagawa, 1985, Fleweling and Bauman, 1990, Turner, Irwin, and
Millstein, 1993; Miller, Mccoy, Olson, and Wallace, 1986; Amato and Rivera, 1999
and etc) quite unsystematically. In this study, I first developed hypotheses, based on
previous literature, to test the background variables. Being informed by the literature,
I then developed hypotheses to test the individual level variables, and family level
variables. Even though some researchers emphasized the importance of family
structure (Small and Luster, 1994; Miller, Forehand, and Kotchick, 1999; Salem,
Zimmerman and Notaro, 1998 and etc), this is the first study that emphasizes family
structure and family process variables by including variables that indicated the nature
of the father-son relationship and the communication among family members.
The aim of this research is to determine which factors best predict adolescent
males engaging sexual behavior. In order to determine this, the following hypotheses
will be tested:
I. Structural variables such as the respondents education, the respondents
age, whether or not the respondent had worked in past year, respondents family
income, and respondents race will have an effect on the 1. respondents age at first
date, 2. respondents age at first intercourse, and 3. whether or not respondent had
ever made anyone pregnant.
II. Individual level variables such as respondents self-esteem and gender role
attitudes, and whether respondent would feel upset if he ever made someone pregnant
(only in analyses in which the dependent variable is ever made someone pregnant)
will have an effect on the 1. respondents age at first date, 2. respondents age at first
intercourse, and 3. whether or not respondent had ever made anyone pregnant.
El. Family level variables such as strictness of the household rules, whether
respondent grew up with both parents, the level of involvement of respondents
father, how close respondent felt to his father, whether respondents parents would
feel upset if the respondent ever made someone pregnant (only in analyses dealing
with ever made someone pregnant), whether respondent had communication with his
parents about birth control, pregnancy, STDs, and AIDS will have an effect on the 1.
respondents age at first date, 2. respondents age at first intercourse, and 3. whether
or not respondent had ever made anyone pregnant.
Data and Sampling
Data for this study come from the 1995 National Survey of Adolescent Males
(NSAM) collected by Sonenstein, Ku, Pleck, and Turner (1998). NSAM is a
nationally representative survey of adolescent men that included a total of 890
variables and 1,729 cases. In this survey that is designed to increase knowledge and
understanding of factors that determine adolescent male contraceptive use, sexual
activity and related risk behaviors.
Data were collected from 15-19 year-old males in the United States. For the
final sample the investigators screened 54,265 housing units and selected 56,199 of
them. Out of 2,2240 young men available, 1,729 have completed the interview
(77%). In this sample, Blacks and Hispanics were oversampled and parental consent
was required minors under 18.
In this survey, there were three major parts:
a. a face-to-face interview that lasted around 75 minutes
b. a self-administered questionnaire that 1,689 (98%) of the 1,729 who
completed the interview had also completed
c. a urine specimen that was requested from those 18 or older to be used for
Variables and Measurement
In this study, there are three dependent variables. The first one, AGEINTER,
is a continuous variable indicating the age at first intercourse. The second one,
AGEDATE, is also a continuous variable. This dependent variable indicates the age
at first dating. The third dependent variable, EVERPREG is a dichotomous
categorical variable. This variable indicates whether the respondent ever had made
someone pregnant; yes is coded 1 and no is coded 0.
Independent variables are age of respondent (AGER), total family income
(INCOME), respondent lived with both parents at age 14 (BOTHPAR), respondents
highest grade of school completed (EDUCATION), respondents self esteem
(ESTEEM), respondents gender role attitudes (GENDER), respondents race
(RACE), respondent worked for pay last year (WORKED), importance of religion to
respondent (RELIGION), respondents communication with parents regarding birth
control (BIRTCON), respondents communication with parents regarding sexually
transmitted diseases (STD), respondents communication with parents regarding
pregnancy (PREGNANCY), respondents communication with parents regarding
AIDS (AIDS), how involved was respondents father in upbringing (FATHERINV),
how close does respondent feel to father (FATHERCLO), strictness of home rules at
age 14 (HOMERULES), respondent would feel upset if got someone pregnant
(RUPSET), and respondents parents would feel upset if respondent got someone
Age (AGER) is a continuous variable, total family income (INCOME) has
$10,000 increments, respondent lived with both parents at home (BOTHPARENT) is
a dichotomous categorical variable where yes is 1 and no is 0, respondents highest
grade completed (EDUCATION) is treated as a continuous variable, guided by factor
analysis results, respondents self esteem (ESTEEM) is created as a scale variable
using the following variables from the original data set (I am satisfied with myself, I
am person of worth equal with others, always felt life would work out the way I want,
willing to admit when I make mistake, can do anything I set my mind to, able to do
things as well as others, I am always a good listener, little control over what happens
to me). The last variable, little control over what happens to me, has been recoded to
make the direction of the response categories consistent with the other variables. The
same procedure was used to create GENDER as a scale variable. GENDER was
created from the following variables: essential for male to get respect from others,
male always deserves respect of wife and children, male loses respect if he talks
about problems, bothers R when male acts like female, male should not have to do
housework, males are always ready for sex, male who doesnt fight back loses
respect, male shows love when fights for girlfriend, in dating female is out to take
advantage of male, females get pleasure putting male down, females respect males
who lay down law to her, females should be more concerned with being wife and
mother). Respondents race (RACE) was originally coded as 1.Black, 2.Hispanic,
3.White, and 4.0ther (check this coding). For analyses in this study, this variable has
been transformed into 3 dichotomous categorical variables as Black, Hispanic, and
White with other being the left-out one. For all of these variables yes is 1 and no is
0. Respondent worked (WORKED) for pay last year is a dichotomous categorical
variable (yes = 1 and no = 0). The importance of religion to respondent (RELIGION)
is coded with 1. very important, 2. fairly important, 3. fairly unimportant, 4. not
important at all. Respondents communication with parents is composed of 4
variables (1. ever talk with parents about AIDS, 2. ever talk with parents about birth
control methods, 3. ever talk with parents about pregnancy, 4. ever talk with parents
about STDs). Variables BIRTHCON, PREGNANCY, STD, and AIDS were coded
yes=l and no=0. The original response categories for each of these variables were
yes = 1 and no = 2. First, these individual variables had been recoded where yes = 1
and no = 0 for each of them. Then, these individual variables were combined. How
involved was respondents father in upbringing (FATHERINV) has response
categories as 1. very involved, 2. somewhat involved, 3. not too involved, 4. not at all
involved. How close does respondent feel to father (FATHERCLO) has response
categories 1. very close, 2. somewhat close, 3. not too close, and 4. not at all close.
Strictness of home rules at age 14 (HOMERULES) has response categories as 1. very
strict, 2. somewhat strict, 3. not strict at all, and 4. no rules. The last two variables,
respondent would feel upset if respondent got someone pregnant (RUPSET), and
respondents parents would feel upset if respondent got someone pregnant (PUPSET),
were originally coded the same way: 1. very upset, 2. a little upset, 3. a little pleased,
4. very pleased, 5. no different. Both variables were recoded into response
categories: 1. very upset, 2. a little upset, 3. no different, 4. a little pleased, and 5.
The first two dependent variables AGEINTER and AGED ATE were analyzed
on a subsample that was created from the original sample after missing values were
taken out for the dependent variables. Out of the listed independent variables,
RUPSET and PUPSET were not used in data analyses. In this subsample, there were
921 adolescent males out of which 343 were Black, 264 were white, and 296 were
Hispanic. The rest (18 respondents) feel into the other category.
Table 4.1 displays the descriptive statistics for the dependent and independent
variables in this sample.
Descriptive Statistics for Model 1 and Model 2
VARIABLES Mean Median Mode Std. Dev. Range N
AGEINTER 14.53 15.00 14 1.928 14 921
AGEDATE 13.73 14.00 14 1.767 12 855
AGER 17.18 17.00 18 1.339 4 921
INCOME 3.58 .3.00 2 1.825 6 875
EDUCATION 10.34 10.00 11 1.533 20 918
WORKED 1.25 1.00 1 .435 1 918
RELIGION 1.89 2.00 2 .899 3 919
BLACK .372 .00 .00 .484 1 921
HISPANIC .321 .00 .00 .467 1 921
WHITE .287 .00 .00 .452 1 921
ESTEEM 13.18 13.00 13.00 3.069 20 921
GENDERATT 29.97 30.00 28.00 5.189 31 921
FATHERIN 1.76 1.00 1 .968 3 921
FATHERCLO 1.86 2.00 1 .962 3 921
BIRTHCON .48 .00 0 .500 1 921
PREGNANCY .68 1.00 1 .467 1 920
STD .51 1.00 1 .500 1 921
AIDS .58 1.00 1 .494 1 921
BOTHPAR .56 1.00 1 .497 1 921
HOMERULE 1.76 2.00 2 .692 3 921
To determine the effects of the independent variables on the first dependent
variable, AGEDATE, a multiple regression analysis was used. In this analysis, first
the background variables (age, education, race, importance of religion, occupation
status, and family income were entered. In the second step, the variables were
defined as the individual factors (self esteem and gender role attitudes) were entered.
In the last step, the full model, finally variables were defined as family factors (father
involvement, being close to father, strictness of home rules, whether respondent lived
with both parents at the age of 14, whether the respondent had talked with parents
about birth control, sexually transmitted diseases, AIDS, and pregnancy) were entered
into the equation. The coefficients of the first two models and of the last full model
are displayed in Tables 4.2,4.3, and 4.4, respectively.
Regression Coefficients for Demographic Variables with Regard to Age of First Date
Model 1A Unstandardized Coefficients B Standardized Coefficients Beta Significance Level
Constant 5.849 .000
AGER .439 .337 .000
INCOME -.070 -.074 .043
EDUCATION .054 .048 .328
WORKED -.045 -.011 .746
RELIGION -.083 -.043 .204
HISPANIC .137 .037 .740
WHITE .089 .024 .830
BLACK .555 .155 .177
Regression Coefficients for Attitudinal Variables with Regard to Age of First Date
Model IB Unstandardized Coefficients B Standardized Coefficients Beta Significance Level
Constant 5.286 .000
AGER .436 .335 .000
INCOME -.072 -.077 .037
EDUCATION .049 .043 .374
WORKED -.043 -.011 .761
RELIGION -.088 -.045 .180
HISPANIC -.157 -.043 .703
WHITE .083 .022 .842
BLACK .592 .165 .151
ESTEEM .011 .019 .579
GENDER .017 .051 .132
Regression Coefficients for Family Variables with Regard to Age of First Date
FuH Model 1C Unstandardized Coefficients B Standardized Coefficients Beta Significance Level
Constant 6.134 .000
AGER .424 .326 .000
INCOME -.076 -.080 .030
EDUCATION .031 .028 .574
WORKED -.052 -.013 .709
RELIGION -.066 -.034 .321
HISPANIC .229 .062 .580
WHITE .197 .052 .639
BLACK .692 .193 .094
ESTEEM .009 .015 .653
GENDER .017 .051 .132
BOTHPAR .131 .038 .287
FATHERINV -.019 -.011 .815
FATHERCLO .041 '.022 .615
HOMERULES -.309 -.124 .000
BIRTHCON -.002 -.001 .985
STD -.036 -.010 .805
PREGNANCY .048 .013 .718
AIDS -.194 -.055 .192
R2 values were .148 for Model 1A, .151 for ModellB, and .170 for the full Model 1C.
Interpretations of the Coefficients
Multiple regressions were performed for the dependent variable AGEDATE-
age at first dating. The findings are shown in Table 4.4. In this full model, for the
dependent variable-age at first dating, age of respondent was statistically significant
at .000 with a positive effect. The unstandardized coefficient was .424, which means
that one-year increase in respondents age results in .424-year increase in age of first
date. Total family income was significant at .030 and had a negative standardized
coefficient of -.080. The unstandardized coefficient was -.076, which means that one
unit increase in income will result in -.076 year delayed age of first date. Strictness of
home rules at age 14 was significant at .000 with a negative standardized coefficient
(-.124). The unstandardized coefficient was -.309, which means that one unit
increase in strictness of home rules will result in -.31 year delay in age of first date.
To determine the effects of the same independent variables on the second
dependent variable, AGEINTER, the same procedure was used; that is in Model 2 A
(Table 4.5) the demographic variables, in Model 2 B (Table 4.6) the individual
variables were regressed on the dependent variable. The full model (Table 4.7) again
included the two prior models and family factor variables.
Regression Coefficients for Demographic Variables with Regard to Age of First
Model 2 A Unstandardized Coefficients B Standardized Coefficients Beta Significance Level
Constant 6.730 .000
AGER .399 .280 .000
INCOME -.039 -.038 .267
EDUCATION .170 .137 .003
WORKED .096 .022 .491
RELIGION -.127 -.059 .054
HISPANIC -.417 -.103 .332
WHITE .142 .034 .744
BLACK -1.147 -.292 .007
Regression Coefficients for Altitudinal Variables with Regard to Age of First
Model 2 B Unstandardized Coefficients B Standardized Coefficients Beta Significance Level
Constant 6.032 .000
AGER .398 .279 .000
INCOME -.046 -.044 .193
EDUCATION .158 .127 .005
WORKED .100 .023 .473
RELIGION -.130 -.060 .051
HISPANIC -.383 -.094 .373
WHITE .130 .031 .764
BLACK -1.098 -.280 .010
ESTEEM .002 .003 .911
GENDER .027 .073 .020
Regression Coefficients for Family Variables with Regard to Age of First Intercourse
Full Model 2C Unstandardized Coefficients B Standardized Coefficients Beta Significance Level
Constant 6.535 .000
AGER .389 .273 .000
INCOME -.063 -.060 .077
EDUCATION .140 .112 .013
WORKED .080 .018 .564
RELIGION -.097 -.045 .148
HISPANIC -.298 -.073 .487
WHITE .193 .046 .656
BLACK -.949 -.242 .027
ESTEEM .000 .001 .983
GENDER .027 .073 .019
BOTHPAR .370 .097 .003
FATHERINV -.093 -.047 .249
FATHERCLO .080 .040 .312
HOMERULES -.204 -.075 .014
BIRTHCON .124 .033 .352
STD -.272 -.071 .066
PREGNANCY -.033 -.008 .803
AIDS .120 .031 .424
R2 values were .237 for Model 2A, .242 for Model 2B, and .261 for the full Model
Interpretations of the Coefficients
Multiple regressions were performed for the dependent variable AGEINTER-
age at first intercourse. The findings are shown in Table 4.7. For the dependent
variable-age at first intercourse, age of respondent, respondent's highest grade of
school completed, being black, respondents gender role attitudes, respondent lived
with both parents at age 14, and strictness of home rules at age 14 were significant
variables. The age of respondent was significant at .000 with a positive standardized
coefficient (.273). The unstandardized coefficient was .389, which means that one
year increase in respondents age results in .389 year increase in age of first
intercourse. Respondent's highest grade of school completed was significant at .013
with a positive standardized beta coefficient (.112). The unstandardized beta
coefficient was .140, which means that one grade/year increase in education results in
.140 years increase in age of first intercourse. Being black was significant at .027 and
had a negative standardized coefficient of -.242. The unstandardized coefficient was
-.949, which means that being black will lower the age of first intercourse by .949
years. Respondents gender role attitudes was significant at .019 with a positive
standardized coefficient (.073). The unstandardized coefficient was .027, which
means that being conservative on gender role attitudes results in .027 years delay in
age of first intercourse. Respondent living with both parents at age 14 was significant
at .003 with a positive standardized coefficient (.097). The unstandardized coefficient
was .370, which means that living with both parents at age 14 results in .370 year
delay in age of first intercourse. Strictness of home rules at age 14 was significant at
.014 with a negative standardized coefficient. The unstandardized coefficient was
-.204, which means that one unit increase in strictness of home rules will results in
.204 year delay in age of first intercourse.
To test the effects of the independent variables on the dichotomous dependent
variable EVERPREG which indicated whether respondent ever made anyone
pregnant, a new sample was created, again after having cleaned the data for missing
values (including dont know and refused to answer categories). In this sample,
there were 642 respondents.
To determine the effects of the same independent variables on the third
dependent variable, EVERPREG, the same procedure was used; that is in Model 3 A
(Table 4.9) the demographic variables, in Model 3 B (Table 4.10) the individual
variables were regressed on the dependent variable. The full model 3C (Table 4.11)
again included the two prior models and family factor variables.
Table 4.8 displays the descriptive statistics for the dependent and independent
variables in this sample.
Descriptive Statistics for Model 3
VARIABLES Mean Median Mode Std. Dev. Range N
EVERPREG .15 .00 0 .362 1 742
AGER 17.22 17.00 18 1.34 4 742
INCOME 3.78 4.00 2 1,82 6 705
EDUCATION 10.44 11.00 11 1.54 7 742
WORKED 1.23 1.00 1 .42 1 739
RELIGION 1.91 2.00 2 .89 3 740
BLACK .34 .00 0 .47 1 742
HISPANIC .31 .00 0 .46 1 742
WHITE .32 .00 0 .47 1 742
ESTEEM 12.89 12.00 11 3.0 17 742
GENDERATT 30.33 30.00 30.00 5.11 31 742
FATHERIN 1.73 1.00 1 .94 3 742
FATHERCLO 1.85 2.00 1 .95 3 742
BIRTHCON .48 .00 0 .50 1 742
PREGNANCY .68 1.00 1 .46 1 741
STD .51 1.00 1 .50 1 742
AIDS .58 1.00 1 .49 1 742
BOTHPAR .57 1.00 1 .49 1 742
HOMERULE 1.78 2.00 2 .69 3 742
RUPSET 1.96 2.00 1 1.08 4 733
PUPSET 1.62 1.00 1 .79 4 735
Regression Coefficients for Demographic Variables with Regard to Whether the
Respondent Had Ever Made Someone Pregnant
Model 3A Unstandardized Coefficients B Significance Level
Constant -10.520 .000
AGER .617 .000
INCOME -.069 .308
EDUCATION -.140 .215
WORKED -1.021 .003
RELIGION -015 .910
HISPANIC 1.063 .325
WHITE .219 .842
BLACK 1.497 .165
Regression Coefficients for Attitudinal Variables with Regard to Whether the
Respondent Had Ever Made Someone Pregnant
Model 3 B Unstandardized Coefficients B Significance Level
Constant -9.616 .000
AGER .615 .000
INCOME -.061 .370
EDUCATION -.118 .301
WORKED -1.018 .004
RELIGION -.017 .894
HISPANIC 1.084 .317
WHITE .298 .786
BLACK 1.504 .164
ESTEEM .001 .973
GENDER -.039 .086
Regression Coefficients for Family Variables with Regard to Whether the Respondent
Had Ever Made Someone Pregnant
FuU Model 3 C Unstandardized Coefficients B Significance Level
Constant -9.774 .000
AGER .458 .001
INCOME -.028 .702
EDUCATION -.021 .850
WORKED -.995 .006
RELIGION -.108 .444
HISPANIC .945 .396
WHITE .234 .837
BLACK 1.169 .296
ESTEEM .003 .933
GENDER -.023 .344
BOTHPAR .072 .781
FATHERINV .207 .219
FATHERCLO -.051 .761
HOMERULES -.091 .595
BIRTHCON -.458 .158
STD .010 .973
PREGNANCY -.137 .620
AIDS .815 .011
RUPSET .323 .000
PUPSET .255 .036
Interpretations of the Coefficients
Logistic regression was performed for the dependent variable EVERPREG-
whether the respondent had ever made someone pregnant. The findings are shown in
Table 4.11. In this full model, age of respondent, respondent worked for pay last
year, respondent would feel upset if got someone pregnant, respondents parents
would feel upset if respondent got someone pregnant, and ever talk with their parents
about AIDS were significant variables. Age of respondent was statistically
significant at .001 with a positive standardized coefficient. The unstandardized
coefficient was .458, which means that one-year increase in respondents age results
in .458 unit increase in the ratio of the probability of making someone pregnant.
Respondent worked for pay last year was significant at .006. The unstandardized
coefficient was -.995, which means that respondent worked more last year result in a
lower ratio of the probability of making someone pregnant. Respondents feeling
upset if got someone pregnant was significant at .000. The unstandardized coefficient
was .323, which means that the more a respondent feels upset to make someone
pregnant, the less likely he has made someone pregnant. Respondents parents
feeling upset if respondent got someone pregnant was significant at .036. The
unstandardized coefficient was .255, which means that the more parents feel upset if
their son got someone pregnant, the less likely their son has made someone pregnant.
Ever talking with their parents about AIDS was significant at .011. The
unstandardized coefficient was .815, which means that respondent who has talked
with their parents about AIDS tends to have lower chance to make someone pregnant.
In this study, three hypotheses were tested and were partially supported.
From the structural level or background variables, respondents age and total
family income, and from family level variables, strictness of home rules were
significant predictors for the dependent variable the age at first dating. However,
background variables such as the respondents education, whether or not the
respondent had worked in past year, respondents race, and importance of religion to
respondent were no effect on the respondents age at first date. Moreover, both
individual level variables self-esteem and gender role attitudes, and rest of family
level variables such as whether respondent grew up with both parents, the level of
involvement of respondents father, how close respondent felt to his father, and
whether respondent had communication with his parents about birth control,
pregnancy, STDs, and AIDS had no effect on the respondents age at first date.
The respondents age, respondents education, being black, respondents
gender role attitudes, strictness of household rules, and whether respondent grew up
with both parents were significant predictors for the dependent variable the age of
first intercourse. However, other background, individual, and family level variables
including respondents family income, whether or not the respondent had worked in
past year, importance or religion, respondents self-esteem, the level of involvement
of respondents father, how close respondent felt to his father, and whether
respondent had communication with his parents about birth control, pregnancy,
STDs, and AIDS were not significant predictors for the age at first intercourse.
For the third dependent variable-whether the respondent had ever made
someone pregnant; respondents age, respondent had worked in past year, whether
respondent had communication with his parents about AIDS, respondent would feel
upset if got someone pregnant, and respondents parents would feel upset if their son
got someone pregnant were better predictors; while rest of the background,
individual, and family level variables including total family income, respondents
education, respondents race, importance of religion, respondents self-esteem,
respondents gender role attitudes, strictness of household rules, and whether
respondent grew up with both parents, the level of involvement of respondents
father, how close respondent felt to his father, and whether respondent had
communication with his parents about birth control, pregnancy, STDs, and AIDS had
no effect on the whether the respondent had ever made someone pregnant.
The results of this study indicate that some of the findings were similar to the
ones of previous researches. For example, being black (Ku, Sonenstein, and Pleck,
1993), gender role ideology (Pleck, Sonenstein, and Ku, 1993) were related to
adolescent sexual behavior. Moreover, Hogan and Kitagawa, 19S5; Miller,
Higginson, McCoy, and Olson, 1985; Mott, 1984; Thornton and Cambum, 1983;
Zelnik, Kantner, and Ford, 1981; Pleck, Sonenstein, and Ku, 1993 found that not
living with both biological parents is a predictor of early sexual intercourse among
adolescents. The findings of this study show that living with both parents at home
was associated with delaying age of first intercourse, which were same to the
previous researches. However, previous studies reported the importance of religion
(Davids, 1982; Notzer, Leuvan, Mashiach, and Soffer, 1984; Reiss, 1967; Reiss and
Miller, 1979, sack, Keller, and Hinkle, 1984; Young, 1982), self-esteem (Miller,
Christen, and Olson, 1987) were related to adolescent sexuality, which the study did
not support. In addition, this study did not find any effect of father involvement and
being close to father on adolescent male sexuality.
This study has several important points. First, this study systematically
investigated background, individual, and family structure and process variables. In
this analysis, first the background variables (respondents age, education, occupation
status, race, importance of religion, and family income) were entered to determine the
effects of the independent variables on the three dependent variables. In the second
level, the variables were defined as the individual factors (self-esteem, gender role
attitudes, and respondent would feel upset if got someone pregnant) were entered. In
the last level, the full model, family variables (father involvement, being close to
father, strictness of home rules, whether respondent lived with both parents at age 14,
whether the respondent had talked with parents about birth control, sexually
transmitted diseases, AIDS, and pregnancy, and respondents parents would feel
upset if respondent got someone pregnant) were entered into the equation. Second,
this study emphasizes family structure and process variables by including variables
that indicated the father-son relationship and the relationship among family members.
Third, the study used a nationally representative sample of adolescent men and
Blacks and Hispanics were oversampled in this sample.
However, there are some limitations of this study. First, this study was
limited to the questions asked by the primary data collection. For example, these data
had very little information regarding neighborhood, peers, sibling (especially sisters),
media, and respondents relationship with their mothers. Second, measurement could
be an issue because the data were collected based on self-reports and were not
validated by another source. Third, the data are limited by attitudinal and family
variables, which would increase knowledge and give better understanding of
predictors that determine adolescent males sexuality.
The findings suggest that education directed toward adolescent males
communicating with their parents regarding birth control, STDs, AIDS, and
pregnancy may not work as well as it has been emphasized. We need to examine
what kind of educational programs would have more effect on adolescent males
This study also suggests that future longitudinal research needs to examine
father-son relationship with different ages and ethnic background. It would be
difficult to reduce adolescent males engaging in early and unsafe sexual intercourse,
unless greater social policy, program, and research are directed toward this issue.
Amato, P. R. & Gilbreth, J. G. (1999). Nonresident fathers and childrens well being:
A meta-analysis. Journal of Marriage and the Family, 61, 557-573.
Amato, P. R., & Rivera, F. (1999). Paternal involvement and childrens behavior
problems. Journal of Marriage and the Family, 61, 375-384.
Bearman, P. (1997). Reducing the risk: Connections that make a difference in the
lives of youth. MD: Bethesda,14.
Browning, C. R., & Laumann, E. O. (1997). Sexual contact and adults: A life course
perspective. American Sociological Review, 62, 540-560.
Cooksey, E. C., & Fondell, M. M. (1996). Spending time with kids: Effects of family
structure on fathers and childrens lives. Journal of Marriage and the Family,
Davis, E. C. & Friel, L. V. (2001). Adolescent sexuality: Disentangling the effect of
family structure and family context. Journal of Marriage and Family, 63, 669-
Day, R. D. (1992). The transition to first intercourse among racially culturally diverse
youth. Journal of Marriage and the Family, 54, 749-762.
Dearden, K., Hale, C., & Blankson, M. (1994). Family structure, function, and the
early transition to fatherhood in Great Britain: Identifying antecedents using
longitudinal data. Journal of Marriage and the Family, 56, 844-852.
Doherty, W. J., Kouneski, E. F., & Erickson, M. F. (1998). Responsible fathering: An
overview and conceptual framework. Journal of Marriage and the Family, 60,
Floyd, H. H. & South, D. R. (1972). Dilemma of youth: The choice of parents or
peers as a frame of reference for behavior. Journal of Marriage and the
Family, 34, 627-634.
Grunbaum, J., et. al. (2002). Youth risk behavior surveillance (YRBS)-United States,
2001. Morbidity and Mortality Weekly Report, 51,1-64.
Hanson, S. M. H., & Bozett, F. W. (1987). Fatherhood: A review and resources.
Family Relation, 36, 333-340.
Harris, K. M., & Morgan, S. P. (1991). Fathers, sons, and daughters: Differential
parental involvement in parenting. Journal of Marriage and the Family, 53,
Hoff, T., et. al. (2003). National survey of adolescent and young adults: Sexual health
knowledge, attitudes, and experiences. Menlo Park, CA: Henry Kaiser Family
Hogan, D. P. & Kitagawa, E. M. (1985). The impact of social status, family structure,
and neighborhood on the fertility of black adolescents. The American Journal
of Sociology, 90, 825-855.
Hovell, M., Sipan, C., Blumberg, 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.
In Their Own Right: Addressing the Sexual and Reproductive Health Needs of
American Men. (2002). New York: The Alan Guttmacher Institute.
Jaccard, J., Dittus, P. J., & Gordon, V. V. (1996). Maternal correlates of adolescent
sexual and contraceptive behavior. Family Planning Perspectives, 28, 159-
Koblinsky, S., & Atkinson, J. (1982). Parental plans for childrens sex education.
Family Relations, 31, 29-35.
Ku, L., Sonenstein, F. L, & Pleck, J. H. (1983). Factors influencing first intercourse
for teenage men. Public Health Reports, 108, 680-694.
______. (1993). Neighborhood, family, and work: Influences on the premarital
behaviors of adolescent males. Social Forces, 72, 479-503.
Lauritsen, J. L. (1994). Explaining race and gender differences in adolescent sexual
behavior. Social Forces, 72, 859-883.
Luster, T., & Small, S. A. (1994). Factors associated with sexual risk-taking
behaviors among adolescents. Journal of Marriage and the Family, 56, 622-
Lye, D. N. (1996). Adult child-parent relationships. Annual Review of Sociology, 22,
Meschke, L. L., Bartholomae, S., & Zentall, S. R. (2000). Adolescent sexuality and
parent-adolescent processes: Promoting healthy teen choices. Family
Relations, 49, 143-154.
Miller, B.C. (2002). Family influences on adolescent sexual and contraceptive
behavior. The Journal of Sex Research, 39, 22.
Miller, B. C., Christensen, R. B., & Olson, T. D. (1987). Adolescent self-esteem in
relation to sexual attitudes and behavior. Youth & Society, 19, 93-111.
Mifler, 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.
Mifler, B. C., Norton, M. C., Curtis, T., Hill, E. J., Schaneveldt, P., & Young, M. H.
(1997). The timing of sexual intercourse among adolescents: Family, peer,
and other antecedents. Youth and Society, 29, 54.
MiUer, B. C., & Olson, T. D. (1988). Sexual attitudes and behavior of high school
students in relation to background and contextual factors. The Journal of Sex
Research, 24, 194-199.
Miller, K. S., Forehand, R., & Kotchick, B.A. (1999). Adolescent sexual behavior in
two ethnic minority samples: The role of family variables. Journal of
Marriage and the Family, 61, 85-98.
Newcomer, S., & Udry, J. R. (1987). Parental marital status effects on adolescent
sexual behavior. Journal of Marriage and the Family, 49, 235-240.
Palkovitz, R. (2002). Involved fathering and mens adult development. Provisional
balances. New Jersey: Lawrence Erlbaum Associates, Publishers Mahwah.
Perkins, D. F., Luster, T., Villaruel, F. A., & Small, S. (1998). An ecological, risk-
factor examination of adolescents sexual activity in three ethnic groups.
Journal of marriage and the Family, 60, 660-673.
Pleck, J. H., Sonenstein, F. L., & Ku, L. C. (1993). Masculinity ideology: Its impact
on adolescent males heterosexual relationships. Journal of Social Issues, 49,
Ramirez-Valles, J., Zimmerman, M. A., & Newcomb, M. D. (1998). Sexual risk
behavior among youth: Modeling the influence of prosocial activities and
socioeconomic factors. Journal of Health and Social Behavior, 39, 237-253.
Rodgers, K. B. (1999). Parenting processes related to sexual risk-taking behaviors of
adolescent males and females. Journal of Marriage and the Family, 61, 99-
Salem, D. A., Zimmerman, M. A., & Notaro, P. C. (1998). Effects of family structure,
family process, and father involvement on psychosocial outcomes among
African American adolescents. Family Relations, 47, 331-341.
SexSmarts: Decision-Making. (2000). Memo Park, CA: Henry Kaiser Family
Foundation and Seventeen Magazine.
Small, S. A & Luster, T. (1994). Adolescent sexual activity: An ecological, risk-
factor approach. Journal of Marriage and the Family, 56, 181-192.
Sonenstein, F. L., Pleck, J. H., & Ku, L. (1991). Levels of sexual activity among
adolescent males in the United States. Family Planning Perspectives, 23,
South, S. J. & Baumer, E. P. (2000). Deciphering community and race effect on
adolescent premarital childbearing. Social Forces, 78, 1379-1407.
Tamis-Le Monda, C. & Cabrera, N. (2002). Handbook of father involvement.
Multidisciplinary perspectives. New Jersey: Lawrence Erlbaum Associates,
Teitler, J. O. & Weiss, C. C. (2000). Effects of neighborhood and school
environments on transitions to first sexual intercourse. Sociology of
Education, 73, 112-132.
Thomberry, T. P., Smith, C. A., & Howard, G. J. (1997). Risk factors for teenage
fatherhood. Journal of Marriage and the Family, 59, 505-522.
Udry, J. R. & Billy, J. O. G. (1987). Initiation of coitus in early adolescence.
American Sociological review, 52, 841-855.
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, 920-933.
Werner-Wilson, R. J. (1998). Gender differences in adolescent sexual attitudes: The
influence of individual and family factors. Adolescence, 33, 519(1).
Whitebeck. L. B., Simons, R. L., & Meei-Ying Kao. (1994) The effects of divorced
mothers dating behaviors and sexual attitudes on the sexual attitudes and
behaviors of their adolescent children. Journal of Marriage and the Family,
Whitbeck, L. B., Yoder, K. A., & Conger, R. D. (1999). Early adolescent sexual
activity: A developmental Study. Journal of Marriage and the Family, 61,