Violence and Victims, Vol. 17, No.

1, 2002

Violence and Gender:

Reports From an Urban High School
Teresa Scherzer
Howard L. Pinderhughes
University of California, San Francisco

This article reports on the survey component of a study examining urban high school stu-
dents’ experiences with violence. The survey’s purpose was to collect information on stu-
dents’ experiences with violence, explore gender differences, and identify which factors
are associated with the self-reported use of violence. Two prominent risk factors for the
self-reported use of violence were found: having a close friend or family member injured
by violence, and gun possession. Young men and women did not differ significantly in
overall exposure, victimization, and perpetration. However, gender clearly informed the
types of violence reported. The findings offer practical strategies for addressing adoles-
cent violence, such as reducing gun availability and community-level violence, but future
research must further examine the role of gender in order to structure more effective pre-
vention and intervention approaches that target different kinds of violence.

Keywords: violence; youth violence; urban violence; gender; school violence; adolescent
violence

A
fter an alarming increase in rates of youth violence from the early 1980s to the
mid 1990s, rates of youth arrests for homicide and other violent crimes have lev-
eled off and even decreased since the mid 1990s (Snyder & Sickmund, 1999).
Despite this recent decline, youth violence remains a serious public health concern. While
homicide rates for all age groups have fallen, rates of youth homicide remain substantial-
ly higher than levels in the mid 1980s. Nationally, homicide is the second leading cause
of death for youth ages 15 to 24 (Snyder & Sickmund, 1999). In California, data on cause
of death for youth ages 13 to 19 years indicate that violence with and without firearms
resulted in approximately four in ten deaths (California Department of Health, 1997).
Research on adolescent violence has established an array of risk factors for adolescents’
use or avoidance of violence; these individual-level, family-level, and community-level risk
factors act cumulatively and synergistically to produce a tendency towards aggressive and
violent behavior (Bell & Jenkins, 1991, 1993; Buka & Earls, 1993; Dryfoos, 1990;
Dukarm, Byrd, Auinger, & Weitzman, 1996; DuRant, Treiber, Goodman, & Woods, 1996;
Edari & McManus, 1998; Gorman-Smith & Tolan, 1998; Grunbaum, Basen-Engquist, &
Pandey, 1998; Prothrow-Stith, 1991; Rachuba, Stanton, & Howard, 1995; Reiss, Jr., &
Roth, 1993; Rivera & Widom, 1990; Saner & Ellickson, 1996; Valois, McKeown, Garrison,
& Vincent, 1995; Widom, 1989). Individual-level factors include low birth weight and other
physical abnormalities, and personal victimization, witnessing violence, relationships with
delinquent peers, drug or alcohol abuse, low degree of academic achievement or school
attachment. Family-level factors include a history of family conflict, violence, divorce or
separation of parents, alcohol or drug abuse, and economic hardship. Community-level fac-
tors are grounded in poverty and its concomitant characteristic. These include a lack of

© 2002 Springer Publishing Company 57

58 Scherzer and Pinderhughes

community cohesion, quality educational or employment opportunities; high levels of

unemployment, residential segregation, overcrowding, gang activity, and interpersonal
crime; and easy access to drugs, alcohol, and firearms.
Protective factors have also been noted (Dryfoos, 1990; DuRant et al., 1996; Edari &
McManus, 1998; Howell & Hawkins, 1998). Individual level factors include: high level
of academic achievement, “religiosity” (the importance or presence of religion in one’s
life), internal locus of control, impulse control, sense of purpose or future, and a sense of
maturity or reflectiveness. Family level factors are based in having sufficient material and
emotional resources to provide a caring and supportive environment for children. These
factors include: parents’ ability to manage stress and conflict, high expectations for
achievement, and a stable financial base. Community-level factors also include caring and
support, as well as low rates of unemployment, crime, and high degree of community
resources and cohesion.
Studies have shown that youth living in inner-city communities report high exposure
rates for witnessing violence and violent victimization (Rachuba et al., 1995; Reiss, Jr., &
Roth, 1993). Latino and African American adolescent boys and girls experience higher
rates of assault and homicide than the general youth population (Kann et al., 1998; Lowry
et al., 1998; Snyder & Sickmund, 1999). Adolescent boys are twice as likely as girls to be
victims of assault and homicide, and adolescent girls experience rape or sexual assault
over ten times more than boys (Snyder & Sickmund, 1999). Adolescent boys are more
likely than girls to commit violence (Bell & Jenkins, 1993; Conseur, Rivara, & Emanuel,
1997; Dryfoos, 1990; DuRant et al., 1996; Fitzpatrick, 1997; Grunbaum et al., 1998;
Howell & Hawkins, 1998; Saner & Ellickson, 1996), to carry a weapon (Dukarm et al.,
1996; Grunbaum et al., 1998; Kann et al., 1998; Lowry et al., 1998; Snyder & Sickmund,
1999; Valois et al. 1995), and to commit gun violence (Anderson, 1994, 1999; Fagan &
Wilkinson, 1998a, 1998b).
This article utilizes data collected during an extensive research project initiated in 1996
that examines urban high school students’ experiences with violence. This study stands
out from other studies of urban youth in two major respects. First, previous research on
inner city youth has utilized samples that are constituted predominantly of males, African
Americans or Latinos, gang members, school drop-outs, or violent or incarcerated youth.
In contrast, this study developed a sample that is much more diverse in terms of gender,
race/ethnicity, and academic achievement and school attachment.
The second way our study diverges from previous research on inner city youth is by its
focus on gender dynamics and violence. Most studies of inner city youth have not
explored young women’s experiences with violence and their differences, if any, com-
pared to their male peers. Studies about young women and violence have focused on
female gang members and others who share the “at-risk” and hardened profile (Belknap,
Holsinger, & Dunn, 1997; Campbell, 1991; Chesney-Lind, 1997; Chesney-Lind &
Shelden, 1998; Harris, 1988).
While a small minority of the youth in our sample reflect the profile of “at-risk” and
hardened inner city youth of previous studies, most of the other students do not. All, how-
ever, have to negotiate and live in an unpredictable and dangerous environment. By look-
ing at a group of young people who do not fit the profile of “violent” youth, who come
from a range of racial-ethnic backgrounds, who are divided approximately equally
between males and females, and who are by and large academically oriented, we hope to
see what young men and women living in the inner city experience in terms of violence
and which factors may lead these youth to commit violence themselves.
Violence and Gender 59

METHODOLOGY

Sample Selection
The study was conducted in an urban high school in the San Francisco Bay Area. The
investigators constructed a purposive sample of 167 students in 16 classes from grades 10
through 12. Students were recruited to answer the survey and participate in the interviews
through the social studies classes. Although all classes were eligible to participate, class-
es were selected to maintain a balance of students across academic skills levels. (Two of
the classes were Spanish-speaking ESL classes, the rest were English-speaking. Two of
the classes were advanced social studies classes.) Ninety six percent of the students pres-
ent in the classes chose to participate. This represented 58% of the students enrolled in
those classes and 19% of the total student population in the school.

Participants
The sample was nonrandom, but nonetheless reflected the gender, racial-ethnic, academ-
ic, and socioeconomic composition of the school’s population (see Table 1). Respondents
were approximately evenly divided between males and females with a mean age of 16.5
years (SD ± .90). The vast majority (95%) were people of color, and of these, most were
Asian/Asian Americans and Latinos. Attempts to estimate respondents’ socioeconomic
status using the variables “mother’s occupation” and “father’s occupation” were unsuc-
cessful due to the lack of responses. However, zip codes of residence offered some infor-
mation, as all but three students reported zip codes for lower socioeconomic districts.

Procedures
The research goals and procedures were described to the students in class and each stu-
dent who agreed to participate was given one week to obtain parental consent. One week
later, a written survey questionnaire (in English or Spanish) was administered to those par-
ticipating students. Students filled out their own survey guided by a researcher who read
through each question to assure the comprehension of students with reading difficulties.

Instrument
A 52-item self-administered anonymous questionnaire asked respondents about social and
demographic characteristics and direct and indirect experiences with violence. Dependent
variables were the use of violence, use of sexual violence, and use of violence in a dating
relationship. Independent variables included risk and protective factors studied in previous
research. Questions on exposure and use of violence were adapted from the Youth Risk
Behavior Survey (YRBS). The YRBS has been found to be a reliable instrument for self-
reports of adolescent health risk behaviors (Brener et al., 1995). The survey instrument was
field tested for reliability and revised accordingly before its administration to the students.
Social Demographic Factors as “Risk” and “Protective” Factors. Students were
asked a range of demographic questions, including age, sex, grade, race and ethnicity,
household composition, parents’ occupations, and zip code of residence. Several risk fac-
tors and protective factors are different aspects of the same concept or variable, such as
gender, socioeconomic status, household composition, school attachment, academic
achievement, and religiosity. Socioeconomic status was measured by parents’ occupations
and zip code of residence. Household composition (“Who do you live with?”) was a nom-
inal variable with five choices, ranging from “live with both parents” to “live on my own.”
60 Scherzer and Pinderhughes

TABLE 1. Selected Characteristics of the Sample

(N = 150)
n (%)
Gender
male 83 (55%)
female 67 (45%)
Age
15 18 (12%)
16 62 (41%)
17 49 (33%)
18 17 (11%)
19 2 (1%)
Race/Ethnicity
African American 17 (12%)
Asian 60 (41%)
Latino 51 (34%)
White 7 (5%)
Other group 12 (8%)
Household Composition
both parents 75 (50%)
one parent 56 (37%)
other guardian 16 (11%)
live alone 2 (1%)
Grade Average
A 27 (18%)
B 57 (38%)
C 46 (31%)
D 12 (8%)
F 6 (4%)
Considered Dropping School 56 (37%)
Attendance, Religious Services
at least monthly 67 (45%)
1-3 times/year 43 (29%)
never 38 (25%)
Spent Time in Jail 9 (6%)

School attachment was a dichotomous variable (“Have you ever considered dropping out
of school?”). Academic achievement and religiosity were five point ordinal variables.
School achievement was operationalized as grade average (“A” to “F”). Religiosity was
operationalized as frequency of “church attendance” with five potential choices from
“more than once a week” to “never.”
Other risk factors were operationalized by dichotomous yes/no variables, including
gun possession, gun ownership, experience of having a close friend or family member
killed or injured by violence, previous victimization, and previous use of sexual or rela-
tionship violence. The youth were asked about lifetime and recent possession of guns,
including possession within the 30 days preceding the survey.
Experiences of Violence. Youth were also asked about their experiences with violence,
ranging from indirect exposure to personal experience as victim or perpetrator. Violence
was defined as the use of physical force which could result in injury. To evaluate violent
Violence and Gender 61

victimization, respondents were asked “Have you ever been a victim of violence?” To
evaluate the self-reported use of violence, respondents were asked “Have you ever used
violence against someone?” Both questions had yes/no choices. We were also interested
in analyzing differences in the kinds of violence experienced by the young people, and
asked more detailed questions beyond these initial, general ones. Respondents were asked
to describe the types of violence they experienced, and to identify the relationship, if any,
with the perpetrators or victims.
Respondents were asked specifically about victimization by or perpetration of specific
behaviors that constitute sexual violence and relationship violence. Sexual violence was
operationalized by asking respondents if they were ever forced (or had ever forced anoth-
er person) “to do sexual things.” Relationship violence was operationalized by asking
respondents if they had ever (or had ever been) “threatened, hit, slapped, or kicked [by] a
boy- or girlfriend.” All four questions had yes/no choices.
To evaluate respondents’ other exposures to violence, respondents were asked if they
had a close friend or family member who had been killed or seriously injured by violence.
Respondents were then asked to write in the number of close friends or family members
killed or seriously injured. In addition, to capture information on the youths’ recent expo-
sures to violence, respondents were asked about experiences of violence in the 12 months
preceding the survey: the number of fights, number of times injured from fights, injuries
from violence, and help-seeking for medical assistance.

Data Analysis
Prior to data analysis, questionnaires were examined for patterned responses (e.g., choos-
ing the same answer for all questions). Based on this examination, 17 were excluded,
resulting in a total of 150 usable surveys.
Descriptive statistics were used to capture the prevalence of lifetime and recent expe-
riences with violence, and lifetime and recent possession of firearms. The survey was also
structured to capture the types of violent experiences based on the perpetrators or victims
and the type of violence identified by the respondents. Chi square analyses explored dif-
ferences of experience based on gender and associations between risk factors and protec-
tive factors and the outcome variable of self-reported use of violence. Those factors found
to be significantly associated with the self-reported use of violence were entered into a
logistic regression.
Due to low cell frequencies in the chi square tests, non-dichotomous variables were
transformed into variables with fewer categories or into dichotomous variables. For exam-
ple, the ordinal variable of grade average was transformed into a dichotomous variable
passing and failing grades (“A-C,” “D-F”). Religiosity was collapsed into three categories:
high religious attendance, moderate religious attendance, and lack of religious attendance
(“at least once a month,” “one to three times a year,” and “never”). Household composi-
tion was collapsed into a dichotomous variable of “two-parents” and “other.”

RESULTS

Exposure to Violence
Table 2 provides a detailed picture of the rates of victimization, perpetration, and expo-
sure to violence, including lifetime and recent rates. More than two out of every five
youths (44%) reported being a victim of violence at some time, and 37% reported using
62 Scherzer and Pinderhughes

violence against another person at some time. In the 12 months preceding the survey, one
in three youth reported being in at least one physical fight (33%), and nearly one out of
five youth reported sustaining an injury from violence (19%). Forty-four percent of youth
reported having a close friend or family member injured from violence, and one in three
(33%) reported having someone close killed by violence.
Differences between males’ and females’ experiences were not statistically signifi-
cant, except for perpetration of sexual violence. In general, the young men reported high-
er rates than the young women of victimization and use of violence, recent fighting and
injury, and lifetime and recent gun possession. Only young men reported committing
sexual violence, and this was the only statistically significant gender difference. Women
reported higher rates than men of sexual and relationship victimization and higher rates
of using violence against someone they were dating—but these differences, like the oth-
ers, were not statistically significant. Slightly more men than women reported someone
close being injured by violence (54% vs. 50%), while a greater difference existed for
more women having someone close killed by violence (41% of the women vs. 29% of
the men).

Gender and the Types of Violence Reported

Although gender and the self-reported use of violence were in few statistically significant
relationships in this sample, gender was significantly associated with different types of
violence, as illustrated in Table 3. First, young men reported many more incidents than
young women of victimization or perpetration of “gang” and “criminal/random” violence.
In terms of “fighting/interpersonal” violence, young men and women reported practically
equal numbers of perpetrating this kind of violence, but young men reported twice as
many incidents of victimization.
Second, when asked to describe the types of violence experienced, only a few young
women—and none of the young men—reported victimization by or perpetration of vio-
lence involving a family member or friend of the family, as well as sexual violence or rela-
tionship violence. This does not correspond to the numbers of responses to the specific
questions about the behaviors operationalizing sexual or relationship violence. For exam-
ple, far more respondents (of both genders) reported sexual and relationship violence by
answering the survey questions that asked about specific behaviors, than the number of
respondents who described the violence they experienced. Also, even though the young
men did not describe the violence they experienced as “family violence,” “sexual vio-
lence,” or “dating violence,” their responses were quite different to the specific questions
about the behaviors operationalizing sexual or relationship violence.
As illustrated in Table 2, no statistically significant gender differences were found regard-
ing sexual victimization, although more young women than men reported having been
“forced to do sexual things” (19% vs. 13%; χ2 = 1.026, df = 1, p = .356). No statistically sig-
nificant gender differences were found regarding victimization by relationship violence,
although a higher percentage of young women than men reported being “threatened, hit,
slapped, or kicked” by their boy- or girlfriends (14% vs. 10%) (χ2 = .564, df = 1, p = .604).
However, a much higher percentage of young women reported committing these behaviors
(25% vs. 10%), a difference that approached significance (χ2 = 4.516, df = 1, p = .053).
Regarding the specific responses about violence, the only statistically significant gender dif-
ference was in terms of perpetrating sexual assault, as only young men reported having
“forced someone to do sexual things” (χ2 = 5.279, df = 1, p = .031); however, the cell fre-
quency is low (n = 6).
TABLE 2. Reported Rates of Victimization by, Commission of, and Exposure to Violence
Sample (N = 150) Males (n = 83) Females (n = 67)
Variable n %1 n %2 n %3 χ2 df p
Victimization
general 66 44% 41 49% 25 40% 1.365 1 0.314
Violence and Gender

sexual 22 15% 10 13% 12 19% 1.026 1 0.356

dating 17 11% 8 10% 9 14% 0.564 1 0.604
Perpetration
general 55 37% 35 43% 20 31% 2.001 1 0.172
sexual 6 4% 6 8% 0 0% 5.279 1 0.031*
dating 21 14% 7 10% 14 25% 4.516 1 0.053
Friend/family member
injured by violence 66 44% 38 54% 28 50% 0.229 1 0.72
Friend/family member
killed by violence 50 33% 24 29% 26 41% 2.207 1 0.162
In last 12 months . . .
had at least one fight 50 33% 32 42% 18 29% 2.526 1 0.154
injured from fight 15 10% 11 14% 4 7% 1.969 1 0.181
injured by violence 29 19% 21 26% 8 13% 3.881 1 0.06
sought medical help 17 11% 13 37% 4 21% 1.478 1 0.358
Gun possession
lifetime 48 32% 32 39% 16 24% 3.646 1 .077
in last 30 days 12 8% 10 31% 2 12% 2.279 1 .175
*p < .05. **p < .01.
1percent of entire sample’s responses.
2percent of males’ responses.
3percent of females responses.
63
64 Scherzer and Pinderhughes

TABLE 3. Types of Violence Reported, Frequencies, by Gender

Perpetration Victimization
Type of Violence Females Males Females Males
Gang 7 21 4 11
Fighting/interpersonal 9 10 12 23
Criminal/random 12 23 8 15
Other 0 1 1 0
Family (non-parent) 1 0 3 0
Parent—child 2 0 0 0
Sexual 1 0 0 0
Friend of family 1 0 1 0
Dating 0 0 2 0

These apparent inconsistencies may be explained by data collected in another component

of the study. After the survey was administered, focus group interviews were conducted with
the participating students in single-sex groups with a facilitator of the same sex; the qualita-
tive data are analyzed and discussed elsewhere (Pinderhughes and Scherzer unpublished
observations). These data may shed some light on the apparent reporting inconsistency and
on an apparent inconsistency specifically concerning the survey data about family violence.
First, the qualitative interview data illustrate that there are differences among the youth about
what constitutes violence, namely, that certain behaviors may or may not be “violent.” These
differences exist despite the specific definition of violence provided by the researchers for
students to use to complete the survey. This may help explain certain apparent inconsisten-
cies of the youths’ responses to different questions about violence, as for example, how none
of the males described violence as “sexual violence” while also reporting having been forced
(or having forced someone) “to do sexual things.” Second, regarding exposure to “family
violence,” the young women systematically described episodes of violence between them-
selves and a parent, other family member, or friend of the family. The young men talked pri-
marily about experiencing excessive physical discipline from a family member, and
witnessing violence against their mothers by lovers or family members. However, a few
young men talked about physically intervening in defense of their mothers or sisters, an
action that was also discussed at length by several of the young women.

Risk and Protective Factors

Chi square tests of association were conducted to explore which risk or protective factors
were significantly related to the self-reported use of violence. As detailed in Table 4, we
found that five out of the 14 factors were not significantly associated with the use of vio-
lence. These factors included social demographic and exposure variables—gender, house-
hold composition, religiosity, sexual victimization, and sexually assaulting another person.
Of the nine factors significantly associated with the use of violence, all but two had to
do with exposure to violence and firearms. The other two factors were the variables that
measured school achievement and school attachment—grade average, and whether
respondents had considered dropping out of school.

About Race
Chi square analyses were also conducted to explore differences between racial-ethnic
groups and the self-reported use of violence and prominent risk factors for the self-report-
ed use of violence. The results revealed that African Americans, while comprising only
Violence and Gender 65

12.5% of the sample (compared to Asian/Asian Americans at 43% and Latinos at 36%),
reported statistically significant higher rates of exposure to violence and the risk factors
associated with the self-reported use of violence. They reported much lower degrees than
Asians or Latinos of academic achievement (GPA) (χ2 = 9.248, df = 3, p = .026) and
school attachment (considered dropping school) (χ2 = 12.442, df = 3, p = .006), and much
higher rates of having a close friend or family member killed or injured by violence (χ2 =
11.909, df = 3, p = .008; χ2 = 14.403, df = 3, p = .002), of gun possession and ownership
(χ2 = 17.395, df = 3, .001; χ2 = 11.533, df = 3, p = .009), and of using violence (χ2 =
17.584, df = 3, p = .001).
We urge great caution about these statistically significant findings of “racial differ-
ences” for several reasons. First, our inability to reliably measure class in the sample pre-
vents us from analyzing the effect of class on the different racial groups. Second, although
we were unable to reliably measure class in the sample, based on our knowledge of the
students at this particular high school, there is less variation in socioeconomic back-
grounds of African American students and virtually all come from neighborhoods charac-
terized by entrenched poverty. Asian and Latino youth, in comparison, come from large
populations with more variation in terms of socioeconomic class and immigration status.
Therefore, we argue that any statistically significant findings about “race” are inextrica-
bly bound to class, and cannot be taken at face value. Based on these factors, we decided
to not control for race and omitted it from the logistic regression.

Logistic Regression Model

Based on the results of the chi square tests and on the conceptual and empirical impor-
tance of gender in the use of violence, a logistic regression model addressing risk and pro-
tective factors in the self-reported use of violence was constructed (Table 5). The model
focused on lifetime exposure to violence and gun possession and eliminated redundancy
among the nine independent variables. The outcome variable was the self-reported use of
violence. This model had 120 cases with eight independent variables and was a good fit
overall to the data.
The results of the logistic regression emphasize the prominence of two risk factors for
the self-reported use of violence: having a close friend or family member injured by vio-
lence and gun possession. These risk factors offered unique contributions to the model,
above and beyond the effects of the other variables in the regression. Respondents who
reported either of these two experiences were over six times more likely to report com-
mitting violence than youth who did not report these risk factors. Gender, while adding to
the model, was not found to be a statistically significant contributor on its own.

DISCUSSION

Young people’s self-reports about their experiences with violence clearly indicate that
violence remains a serious problem. However, the sample’s rates of exposure to vio-
lence are comparable to data from the national Youth Risk Behavior Survey and are, in
fact, lower than other studies of youth in inner-city communities. Data from the 1997
YRBS indicate in the 12 months preceding the survey, 36.6% of the respondents were
in at least one physical fight and 3.5% received medical treatment by a doctor or nurse
because of injuries sustained by fighting; in the 30 days preceding the survey, nearly 6%
carried a gun. Males and youth of color were more likely than females and white youth
to report these behaviors (Kann et al., 1998). In recent studies of inner-city youth,
TABLE 4. Risk and Protective Factors and Self-Reported Use of Violence
Used Did Not Use
Violence Violence
Independent Variable [n (%)]1 [n (%)] χ2 p
Gender
female 20 (31%) 44 (69%) 2.001 .172
male 35 (43%) 47 (57%)
Household composition
two-parent 23 (31%) 52 (69%) 3.482 .086
other 32 (46%) 38 (54%)
Religious attendance
at least once a month 27 (42%) 38 (59%) 1.148 .563
one to three times a year 14 (33%) 28 (67%)
never 12 (32%) 25 (68%)
Grade average
A-C/100-70 44 (35%) 82 (65%) 4.577 .032
D-F/< 69 11 (61%) 7 (39%)
Considered dropping school
no 25 (27%) 67 (73%) 12.371 .001
yes 30 (57%) 23 (43%)
Close friend/family member
killed by violence
no 27 (28%) 69 (72%) 11.543 .001
yes 27 (57%) 23 (43%)
Close friend/family member
injured by violence
no 7 (12%) 53 (88%) 30.486 .000
yes 38 (59%) 26 (41%)
Personally victimized
no 19 (24%) 60 (76%) 14.831 .000
yes 36 (55%) 29 (45%)
Sexually victimized
no 41 (35%) 75 (65%) 2.882 .10
yes 12 (55%) 10 (46%)
Victimized by boy- or girl-friend
no 41 (33%) 83 (67%) 11.92 .001
yes 13 (77%) 4 (24%)
Sexually assaulted another
no 48 (37%) 83 (63%) 0.438 .671
yes 3 (50%) 3 (50%)2
Used violence against boy-
or girl-friend
no 27 (27%) 75 (74%) 26.345 .000
yes 18 (86%) 3 (14%)3
Own gun
no 50 (36%) 90 (64%) 5.556 .029
yes 5 (83%) 1 (17%)
Ever possessed gun
no 20 (21%) 77 (79%) 37.303 .000
yes 35 (73%) 13 (27%)
1Frequency/percent of each value of Independent variable.
2Percentages suggest that of the persons reporting sexual offending, half considered this
“violent” and half did not.
3These figures suggest that of the persons reporting use of violence against a boy- or
girl-friend, approximately one out of seven did not consider this behavior “violent.”
Violence and Gender 67

between one-half and three-fourths of the sample had been victimized or had witnessed
violence (Bell & Jenkins, 1993; Fitzpatrick, 1997; Gorman-Smith & Tolan, 1998). For
example, Gorman-Smith & Tolan (1998) found that youth reported that in the year pre-
ceding the survey, 6% had been victimized by violent crime and 65% had witnessed
some type of violence, with 50% witnessing more than one event. Over one in three
respondents reported having a family member robbed or attacked (33%), and nearly one
in ten (9%) reported having a close friend killed.
The data in this study generally support the findings of the extant research in terms
demonstrating that exposure to violence and possession of firearms are prominent risk
factors for the self-reported use of violence. Consistent with the literature, previous vic-
timization by violence, having a close friend or family member killed or injured by vio-
lence, and academic orientation were all significantly related to the self-reported use of
violence. However, several of the social demographic risk and protective factors previ-
ously found to be significantly associated with the self-reported use of violence were not
found to be so in this study. In fact, one of the most surprising findings was that in
marked contrast to the literature, gender was not significantly related to the self-report-
ed use of violence.
If gender is not significantly associated with the self-reported use of violence, this
strongly suggests that exposure to violence crosses gender lines and that young men and
women in the inner-city youth share many risks for and experiences of violence. This is
underscored by our findings that not only was gender not related to the self-reported use
of violence, but gender was not significantly related to other markers of exposure and risk
for violence—personal victimization, victimization of a close friend or family member,
recent fighting and injury, and lifetime and recent gun possession. These findings are con-
sistent with studies that have shown higher rates of exposure and offending among inner
city young African American women (Chesney-Lind & Shelden, 1998; Sommers &
Baskin, 1992). However, those studies still found significant differences between males
and females. One potential explanation is that these previous studies examined victimiza-
tion, arrest and offending rates for male and female juvenile delinquents. Our study exam-
ines the rates of exposure in a school-based population of youth. The relationship of
gender to violence is shaped and altered somewhat by the inner city context, with its high
rates of poverty, social disorganization, and high levels of violence.

TABLE 5. Logistic Regression: Factors for the Self-Reported Use of Violence

Variable Adjusted OR 95% CI p
Grade average 1.40 .34, 5.74 .6385
Considered dropping out
of school 2.47 .83, 7.33 .1034
Close friend or family member
killed by violence 1.21 .38, 3.83 .7443
Close friend or family member
injured by violence 6.61 1.87, 23.39 .0034*
Previous victimization 2.57 .91, 7.20 .0735
Own a gun 3.41 .23, 51.80 .3761
Ever possessed gun 6.34 2.16, 18.58 .0008**
Gender 1.75 .58, 5.27 .3195
Note. Dependent Variable: use violence against another person. N = 120.
Overall Hosmer/Lemeshow Goodness-of-Fit test: χ2 = 10.4814, df = 7, p = .1661.
*p < .01. **p < .001.
68 Scherzer and Pinderhughes

Our data illustrate an important dynamic. While young men and young women did not
differ significantly in terms of overall exposure, victimization, and perpetration, gender
clearly informed the types of violence the youth reported. There appears to be less of a dif-
ference in reporting violence experienced outside one’s family and more of a difference in
reporting sexual and family violence. Only young women reported experiencing family vio-
lence as victim or perpetrator or victimization by sexual violence—a finding consistent with
Chesney-Lind’s and colleagues’ (1997, 1998) studies of young female violent offenders.
Only young men reported having “forced someone to do sexual things,” although in
responding to earlier questions about using violence on the survey, none of them named vio-
lence they reported as “sexual violence,” and only half (n = 3) also reported ever having used
violence against another person. This strongly suggests that sexual coercion or assault is not
always considered to be “violent” compared to, for example, an attack of a non-sexual nature
There is an interesting paradox in the gender differences in the reported use of vio-
lence in dating relationships. Although the preponderance of intimate partner violence is
committed by men against women, studies of dating violence found that men and women
generally report similar rates of dating violence, with women reporting slightly higher
rates of using violence (Arias, Samios, & O’Leary, 1987; Foo & Margolin, 1995;
Hamberger & Ambuel, 1998; Riggs, 1993). In Avery-Leaf and colleagues’ (1997) study
of high school students, young women and men did not differ significantly in terms of
self-reported victimization, but did for self-reported use of violence, as 53% of the young
women reported using violence against a boyfriend, while 21% of the young men report-
ed using violence against a girlfriend.
Our finding that young women reported higher rates than young men of using violence
in dating relationships may be due to a number of factors. First, young women may be
more likely than men to remember, and admit, acting violently against a boy- or girlfriend,
as suggested by a recent study of intimate partner violence (Schafer, Caetano, & Clark,
1998). However, women may be more likely to commit lower-degree violence (e.g., a
slap) than men (Arias et al., 1987; Foo & Margolin, 1995; Hamberger & Ambuel, 1998;
Makepeace, 1983; Schafer et al., 1998). In addition, the qualitative interview data suggest
that many of the young women were in relationships with older men (whose experience
of dating violence would be excluded from the survey) and so the young women’s and
young men’s reports of relationship violence cannot be “matched” or compared with one
another. In the young men’s focus groups, when the topic of dating violence was brought
up, they rarely talked about it beyond generally acknowledging that it was “wrong” to hit
one’s girlfriends.

CONCLUSION

This study’s data offer the opportunity to look at the relevance of specific “risk” or “pro-
tective” factors in relation to the types of violence experienced for inner-city high school
students. For these youth, it appears that direct and indirect experiences of violence and
guns were much more important influences than social demographic factors regarding the
self-reported use of violence. The interaction of gender with other factors indicates that
exposure to violence and the use of violence are not clearly delineated along gender lines.
Gender is an important factor because it informs the type of violence reported. That is, our
survey and interview data suggest that young men and young women experience similar
rates of violence and different kinds of violence; moreover, there is no one agreed-upon
definition of violence (Pinderhughes and Scherzer unpublished observations).
Violence and Gender 69

This study demonstrates that certain risk and protective factors for the self-reported use
of violence are relevant for a sample of inner-city students, while also acknowledging that
violence is a sweeping phenomenon which occurs in multiple domains. A key point is that
gender accounts for differential distribution of types of violence and different under-
standings of what constitutes “violence.”

Limitations
Several limitations of the study should be noted. First, the study utilized a nonrandom
sample, and the findings cannot be generalized to the larger population of youth outside
the high school we selected. We acknowledge that because the sample is drawn from a
single school, there is limited variance of social and demographic factors. However, the
composition of the sample was similar to the school population in terms of gender, SES,
racial-ethnic composition, and distribution of grade average.
The consent process and the use of school classes as the recruitment site may bias the
sample by excluding youth who have dropped out of school, don’t attend school regular-
ly, or who skip class. This is suggested by the fact that despite the high participation rate
among students present in selected classes (96%), this represented less than 60% of the
students enrolled in these classes. This possible selection bias would likely produce an
underreporting of victimization and perpetration of violence, as more youth at risk for vio-
lence would not have participated in the study.
The survey instrument has several limitations. While the data can illustrate the differ-
ent types of violence reported, they are unable to suggest the motive or context, or sever-
ity of the act, especially since dichotomous (yes/no) categories were used. No conclusions
can be drawn, then, as to how many youth used violence in self-defense or in acts of
aggression, the degree of severity, and how these might be associated with gender differ-
ences. This is particularly relevant for the reported violence between boy- and girlfriends.
In terms of respondents’ reporting whether they had/had been “threatened, hit, slapped, or
kicked [by] a boy- or girlfriend,” the data tell us that a higher percentage of young women
report involvement in relationships that have these behaviors. We cannot conclude any-
thing further in order to compare the reports and experiences of young men and women.
Despite the clear definition of “violence” provided to the students as they completed
the survey, students may have retained their own and competing definitions, which may
explain some of the reporting inconsistencies discussed earlier. Regardless of students’
self-reports of the specific behaviors detailed in the survey (e.g., if they had/had been
“threatened, hit, slapped, or kicked [by] a boy- or girlfriend”) the students may or may not
have considered the specific incident “violence.” Moreover, differences in regarding
acceptable or appropriate behavior in family or dating relationships may result in under-
reporting of violence, especially among male respondents (Schafer et al., 1998).
The survey also did not measure exposure to delinquent peers or drug or alcohol use,
which are recognized risk factors for committing violence. Also, as described earlier,
attempts to reliably measure students’ class or socioeconomic status were not successful.
Therefore any findings of statistically significant “racial differences” in terms of the self-
reported use of violence and risk factors are based on the understanding that these differ-
ences are informed by class, based on our knowledge of the students. Lastly, issues of
social desirability and fear of legal ramifications may preclude respondents from report-
ing incidents of violence and from reporting ownership or possession of guns.
Despite these limitations, the data from this study offer the opportunity to further
explore young people’s shared and different experiences with violence, and moreover, the
opportunity to develop relevant and appropriate prevention and intervention approaches to
70 Scherzer and Pinderhughes

reduce violence among adolescents. The most prominent risk factors, exposure to violence
and firearm possession, point to the need for intervention programs and solutions that
focus on the whole environment and which do not treat violence as a problem of individ-
ual families or persons. The findings from the chi square tests and the logistic regressions
may offer a more practical direction for addressing adolescent violence, such as approach-
es that focus on reducing gun availability and community-level violence. Taken together,
the survey and interview data demonstrate researchers must maintain an awareness that
young men and women may have very different experiences of and ideas about violence.
This points to the need for researchers to actually listen to what young people are saying
about their experiences with violence. When we can hear how they talk about violence,
we can more accurately see the role of gender in experiences of victimization, perpetra-
tion of violence, coping strategies, and in structuring more effective prevention and inter-
vention approaches that target different kinds of violence.

REFERENCES

Anderson, E. (1994, May). The code of the streets. The Atlantic Monthly, 273(5), 81-94.
Anderson, E. (1999). Code of the street: Decency, violence, and the moral life of the inner city. New
York: W. W. Norton.
Arias, I., Samios, M., & O’Leary, K. D. (1987). Prevalence and correlates of physical aggression
during courtship. Journal of Interpersonal Violence, 2(1), 82-90.
Avery-Leaf, S., Cascardi, M., O’Leary, K. D., & Cano, A. (1997). Efficacy of a dating violence pre-
vention program on attitudes justifying aggression. Journal of Adolescent Health, 21(1), 11-17.
Belknap, J., Holsinger, K., & Dunn, M. (1997). Understanding incarcerated girls: The results of a
focus group study. Prison Journal, 77, 381-404.
Bell, C. C., & Jenkins, E. J. (1991). Traumatic stress and children. Journal of Health Care for the
Poor and Undeserved, 2(1), 175-185.
Bell, C. C., & Jenkins, E. J. (1993). Community violence and children on Chicago’s Southside.
Psychiatry, 56(1), 46-54.
Brener, N. D., Collins, J. L., Kann, L., Warren, C. W., Williams, B. I. (1995). Reliability of the Youth
Risk Behavior Survey questionnaire. American Journal of Epidemiology, 141, 575-580.
Buka S., & Earls, F. (1993). Early determinants of delinquency and violence. Health Affairs, 12(4),
46-64.
California Department of Health, Vital Records. (1997). Death by Age 13-19, Sex, and Cause of
Death, California Residence Death, California, 1997. Sacramento, CA: Author.
Campbell, A. (1991). The girls in the gang (2nd ed.). Cambridge, MA: Basil Blackwell.
Chesney-Lind, M. (1997). The female offender: Girls, women, and crime. Thousand Oaks, CA:
Sage.
Chesney-Lind, M., & Shelden, R. G. (1998). Girls, delinquency, and juvenile justice (2nd ed.).
Belmont, CA: Wadsworth.
Conseur, A., Rivara, F. P., & Emanuel, I. (1997). Juvenile delinquency and adolescent trauma: How
strong is the connection? Pediatrics, 99(3), E51-E57. Available: www.pediatrics.org/cgi/
content/full/99/3/e5
Dryfoos, J. G. (1990). Adolescents at risk: Prevalence and prevention. New York: Oxford University
Press.
Dukarm, C. P., Byrd, R. S., Auinger, P., & Weitzman, M. (1996). Illicit substance use, gender, and
the risk of violent behavior among adolescents. Archives of Pediatrics and Adolescent
Medicine, 150, 979-801.
DuRant, R. H., Treiber, F., Goodman, E., & Woods, E. R. (1996). Intentions to use violence among
young adolescents. Pediatrics, 98, 1104-1108.
Violence and Gender 71

Edari, R., & McManus, P. (1998). Risk and resiliency factors for violence. Pediatric Clinics of North
America, 45, 293-305.
Fagan, J., & Wilkinson, D. (1998a). Guns, youth violence, and social identity in inner cities. In M.
Tonry (Ed.), Youth violence, crime and justice: A review of research (Vol. 24, pp. 105-188).
Chicago: University of Chicago Press.
Fagan, J., & Wilkinson, D. (1998b). Social contexts and functions of adolescent violence. In D. S.
Elliott & B. A. Hamburg (Eds.), Violence in American schools: A new perspective (pp. 55-93).
New York: Cambridge University Press.
Fitzpatrick, K. M. (1997). Aggression and environmental risk among low-income African-American
youth. Journal of Adolescent Health, 21(3), 172-178.
Foo, L., & Margolin, G. (1995). A multivariate investigation of dating aggression. Journal of Family
Violence, 10, 351-377.
Gorman-Smith, D., & Tolan, P. (1998). The role of exposure to community violence and develop-
mental problems among inner-city youth. Development and Psychopathology, 10(1), 101-116.
Grunbaum, J. A., Basen-Engquist, K., & Pandey, D. (1998). Association between violent behaviors
and substance abuse among Mexican-American and Non-Hispanic White high school students.
Journal of Adolescent Health, 23, 153-159.
Hamberger, L. K., & Ambuel, B. (1998). Dating Violence. Pediatric Clinics of North America, 45,
381-390.
Harris, M. G. (1988). Cholas: Latino girls and gangs. New York: AMS Press.
Howell, J. C., & Hawkins, J. D. (1998). Prevention of youth violence. In M. Tonry (Ed.), Youth vio-
lence, crime and justice: A review of research (Vol. 24, pp. 263-315). Chicago: University of
Chicago Press.
Kann, L., Kinchen, S. A., Williams, B. I., Ross, J. G., Lowry, R., Hill, C. V., Grunbaum, J. A.,
Blumson, P. S., Collins, J. L., & Kolbe, L. J. (1998). Youth risk behavior surveillance—United
States, 1997. Journal of School Health, 68, 355-369.
Lowry, R., Powell, K.E., Kann, L., Collins, J. L., & Kolbe, L. J. (1998). Weapon-carrying, physical
fighting, and fight-related injury among U.S. adolescents. American Journal of Preventive
Medicine, 14, 122-129.
Makepeace, J. M. (1983). Life events stress and courtship violence. Family Relations, 32(1), 101-109.
Pinderhughes, H. L., & Scherzer, T. Data analysis. Unpublished observations.
Prothrow-Stith, D. (1991). Deadly consequences: How violence is destroying our teenage popula-
tion and a plan to begin solving the problem. New York: HarperCollins.
Rachuba, L., Stanton, B., & Howard, D. (1995). Violent crime in the United States: An epidemio-
logic profile. Archives of Pediatrics and Adolescent Medicine, 149, 953-960.
Reiss, Jr., A. J., & Roth, J. A. (Eds.). (1993). Understanding and preventing violence. National
Research Council. Washington, DC: National Academy Press.
Riggs, D. S. (1993). Relationship problems and dating aggression: A potential treatment target.
Journal of Interpersonal Violence, 8(1), 18-35.
Rivera, B., & Widom, C. S. (1990). Childhood victimization and violent offending. Violence and
Victims, 5, 19-35.
Saner, H., & Ellickson, P. (1996). Concurrent risk factors for adolescent violence. Journal of
Adolescent Health, 19, 94-103.
Schafer, J., Caetano, R., & Clark, C. (1998). Rates of intimate partner violence in the United States.
American Journal of Public Health, 88, 1702-1704.
Snyder, H. N., & Sickmund, M. (1999, September). Juvenile offenders and victims: 1999 national
report. Washington, DC: Office of Juvenile Justice and Delinquency Prevention, National
Center for Juvenile Justice.
Sommers, I., & Baskin, D. (1992). Sex, race, age, and violent offending. Violence and Victims, 7,
191-201.
Valois, R. F., McKeown, R. E., Garrison, C. Z., & Vincent, M. L. (1995). Correlates of aggressive
and violent behaviors among public high school adolescents. Journal of Adolescent Health,
16(1), 26-34.
72 Scherzer and Pinderhughes

Widom, C. S. (1989). Child abuse, neglect, and adult behavior: Research design and findings on
criminality, violence, and child abuse. American Journal of Orthopsychiatry, 59, 355-367.

Acknowledgement. This project was supported by grant number HS 00086 from the Agency for
Heathcare Research and Quality.

Offprints. Requests for offprints should be directed to Teresa Scherzer, Department of Social and
Behavioral Sciences, Box 0612, University of California San Francisco, CA 94143-0612.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.