Differences Based On Perpetrator Identity

Psychosocial And Behavioral Measures Among Female Adolescent Victims Of Sexual Abuse: Differences Based On Perpetrator Identity

Stephen Nagy

Abstract: Adolescent females in a rural southern state who were sexually active were compared with adolescent females who met inclusion criteria for sexual abuse on seven behavioral measures. Victimization comparisons utilized the classification categories based upon the identity of the perpetrator; immediate family member (N=316), boyfriend (N=237), external family member (N=60), and other or stranger (N=143). Risky profiles utilizing odds ratios showed that the victimized group where perpetrators were immediate family members only differed from the sexually active group on risk for pregnancy. The other perpetrator groups showed increased risk for most of the seven measures. Implications for developing intake criteria to review for sexual abuse history using the presence of risk factors is discussed.

Due to the frequency of sexual abuse, its sequelae have been studied extensively. Studies have shown that a number of factors have been associated with the effects of sexual abuse in children and adolescents, including: truancy, depression, inappropriate sexual behavior, substance abuse, and pregnancy (Nagy, Adcock & Nagy 1994; Green, 1996; Finkelhor, 1994). Problematic in interpreting much of this research has been the methodologic limitations of sampling from clinical populations and the varing definition among researchers of sexual abuse terminology (Green 1996; Paradise, Rose, Sleeper & Nathanson, 1994). Methodological and sampling issues have provided an inconsistent framework from which clinical personnel can apply findings to medical, school and community settings. These issues make it difficult to clearly distinguish which factors may assist in the identification of youngsters who have been victims of sexual abuse. All victims of sexual abuse do not necessarily need psychotherapy however, it is generally agreed that victims would benefit from professional counseling to ensure appropriate coping from victimization.

Although research has supported a relationship between a history of sexual abuse abuse.

sexually abused samples have usually been compared with non-abused samples (Rind & Tromovitch, 1997). This presents methodological issues since sexually active adolescents display attributes that are frequently different from their non-experienced counterparts (Nagy, Adcock & Nagy, 1994; Nagy, DiClementi & Adcock, 1995). When comparing sexually abused, sexually active and sexually inactive groups, behaviors become progressively more deviant, with sexually inactive adolescents showing fewest deviant behaviors while sexually abused adolescents display significantly more deviant behaviors. The comparison of sexually abused groups with sexually active groups should assist in clarifying some of these methodological issues when examining possible sequelae of sexual abuse.

The present study was undertaken to clarify factors associated with sexually active non-abused adolescent females and sexually abused adolescent females. A sample of grade eight and grade ten students were surveyed using an anonymous self-administered questionnaire. The main objective was to learn whether a non-clinical sample of sexually abused females differed from non-abused sexually active females, and whether the perpetrator’s identity modified the identified associations with sexual abuse.


Data were collected using an instrument that retained a core of items from two previous surveys conducted in 1988 (Nagy & Adcock, 1990), and 1990 (Nagy & Adcock, 1992) and examined a variety of health behaviors and attitudes. Unique to the 1993 questionnaire, were items on Sexual abuse and forced sex experiences. Sexual abuse was broadly defined as “someone touching you in a place that you did not want to be touched, or did something to you sexually that you did not want.” Previous experiences with similar samples (Nagy & Adcock, 1990; Nagy & Adcock 1992) demonstrated that students had difficulties in distinguishing between forced sex and sexual abuse. Subsequently, students were also asked if they had ever experienced forced sex by asking “has someone ever forced you to have intercourse (sex)?” Students also responded to two questions on the identity of the perpetrator, which allowed one selection from nine categories. Due to the age range of the students, school personnel did not allow additional questions addressing oral and anal sex and same gender experiences.

The questionnaire was reviewed by a panel of experts and pilot-tested with eighth and tenth grade students. A correlation of .80 was obtained (N=69) for the great majority of items when using a test re-test method with a four week interim. Most students completed the survey within a 50-minute class period.

Subjects were selected from schools utilized in previous surveys and consisted of two metropolitan school districts, four rural school districts and three districts in semi-urban settings. State Health Department data on adult mortality and morbidity rates and teenage pregnancy rates show these districts to be similar to surrounding counties. All students present on the day of testing were included. A week before the survey was administered, students were provided with letters to take home informing parents about the study. If parents elected not to involve their child in the study they were required to return a signed letter indicating their decision. Fewer than 1% of the students returned signed letters.

Data were collected from all students in the rural districts. In larger districts, schools were randomly selected and at least 250 students from each grade level were included from randomly selected classes. Occasionally, several intact classes were assembled in a cafeteria, gymnasium, or auditorium and the instrument was administered to all the classes. Students usually were seated to allow for privacy and to respond unobserved by other students or members of the research team. They were provided with a questionnaire, a computer answer sheet, a pencil and a separate identification (referral) form for students wishing to speak to counselors about topics addressed in the survey. A member of the research team then read the instructions and indicated that students could elect not to participate or could discontinue at any time. When completed, students deposited their answer sheet and referral form in a large, enclosed collection box. Before departing from the school premises, researchers provided the referral forms to the principal. Approximately 3% of the students requested counselor contact. Follow up on the referral forms was the responsibility of each individual school and its staff.


Data were collected from 6288 students. Among this sample, there were 3124 females (54%) who comprised the group for analysis. Demographic data are shown in Table 1. The sample was predominantly 14, 15 or 16 (76%) years of age. More than half were white (59%)with 3.8% African-American and 4% of other races. With respect to familial living arrangement, half (51%) resided with their mother and father, 21% with their mother only and 28% resided in some other arrangement. Parental education levels were relatively high with about one-third possessing a college degree. Since the sample was predominantly white and African-American (together they accounted for 96% of the sample), the following analyses are limited to the two ethnic groups.

Table 1 Demographic Information (Answers in percent)


13 or younger 16

14 23

15 25

16 28

17 or older 8


White 58

Black 38

Hispanic 1

Asian 1

Indian 1

Other 1

Living Arrangement

Mother & Father 51

Mother Only 21

Mother & Relatives 2

Father & Relatives 5

Grandparents 1

Father & Stepmother 4

Mother & Stepfather 3

Other 13

Parents’ Education


Below high school 17/14

Graduated high school 28/29

Some college 22/18

College or more 33/39


Students who indicated that they had not been abused and were not sexually experienced comprised 45% of the sample and were not included in the analysis. Comparisons of these students with other students indicated that they were proportionately younger, resided more in two parent households and were proportionately more white (p [is greater than] .001). A number of additional adolescents were also not included in the analysis. These students consisted of adolescents who responded to sexual abuse items but did not indicate experiencing an episode of forced sex and other adolescents who were non-respondents to the sexual abuse items. This group accounted for 10% of the sample, and did not show meaningful differences on any of the demographic and behavioral measures when compared with sexually active and sexually abused students.

Possible interpretation issues of the respondents that surrounded the terms of “sexually abused” and “forced sex” encouraged a conservative definition of sexual abuse. When asked if they had ever been sexually abused (“someone touching you or doing something to you sexually that you did not want”), 22% of the sample responded positively. This group was comprised of students who had been sexually abused and not forced to have sex (13%) and those who had been sexually abused and forced to have sex (9%). The question on forced sex, however, indicated that 24% had been forced to have intercourse (“sex”). Clearly, some students who had been forced to have sex did not define themselves as sexually abused. By using the criteria of positive responses to the item on having experienced forced sex and a positive response to the item that identified the perpetrator of an episode(s) of forced sex, 13% of the adolescents were conservatively defined as having been sexually abused.

Understandably, there are limitations in this definition since a wide variety of sexual acts such as sodomy, and oral sex are serious forms of sexual abuse that may result in trauma. Unfortunately, the lack of ability to differentiate between major traumatic events and more minor events (i.e. touched on the breast while fully clothed) encouraged that this group not be included in the analyses.

The definition of having been sexually abused resulted in the identification of 756 female victims. A review of the entire victimized group showed four general perpetrator types which included boyfriend (N=237), immediate family member (N=316), external family member (N=60) and others (N=143). Additional characteristics of the perpetrator were not requested on the survey. The second group for the analyses was comprised of students who indicated that they had not experienced sexual abuse or forced sex episodes, but responded positively to being sexually experienced (32%).


The study variables consisted of five categories of behavioral measures including: (1) sexual behavior which consisted of the presence of multiple sex partners, young age of sexual initiation, and pregnancy status; (2) lifetime use of illegal substances; (3) whether the student experienced an assault episode during the past year and; (4) mental health items that examined whether the student had made plans to commit suicide during the past year, and whether the student had experienced greater than average frequency of sadness during the past month. Adjusted odds ratios (OR) and confidence limits (CI) were calculated to assess the magnitude of association between the independent measures and group membership (sexually abused versus sexually active).

A review of the group comparisons (table 2) shows differences based upon the identity of the perpetrator and different perpetrator groups presented some different profiles. Victims who identified a family member as their perpetrator were the largest group (N=316) and seemed to display the least disparity from sexually active students. This group was more than two times more likely to have been pregnant but otherwise did not show any other differences on the independent measures.

Table 2 Group Comparisons on Measures by Identity of Perpetrator


Immediate Family Member N=316

Significance Level Odds Ratio Confidence


Multiple Partners .190

Pregnant .001 2.24 1.41-3.97

Assaulted .356

Depressed .867

Illegal Drug Use .515

Suicide Plans .948

Young Sex .968


Boyfriend N=237

Multiple Partners .269

Pregnant .001 2.39 1.44-3.97

Assaulted .001 4.98 3.11-8.84

Depressed .650

Illegal Drug Use .029 1.77 1.06-2.96

Suicide Plans .019 1.25 1.04-1.52

Young Sex .380


External Family Member N=60

Multiple Partners .842

Pregnant .651

Assaulted .013 3.25 1.36-8.25

Depressed .026 2.17 1.10-4.31

Illegal Drug Use .883

Suicide Plans .671

Young Sex .025 2.30 1.11-4.76


Other/Stranger N=143

Multiple Partners .005 2.08 1.48-3.45

Pregnant .310

Assaulted .004 2.60 1.39-4.93

Depressed .001 2.06 1.33-3.19

Illegal Drug Use .112

Suicide Plans .007 1.36 1.09-3.43

Young Sex .035 1.67 1.04-2.54


A second perpetrator group consisted of females who had been abused by boyfriends. This group was also more than twice as likely to become pregnancy and in addition, displayed much higher likelihood of experiencing assault (five times more likely), with greater risk for illegal drug use (OR=1.77) and suicidal plans (OR=1.25).

The third perpetrator group, external family members, showed a somewhat riskier profile than immediate family members but seemed to show a profile that was not as deviant as the boyfriend perpetrator group. Victims of the external family member group were more likely to be assaulted (OR=3.25), depressed (OR=2.17) and to initiate sexual intercourse at a younger age (OR=2.3).

The group comparison with the largest number of significant differences between the two comparison groups was the perpetrator category of other/stranger. This group was not different from the sexually active group on comparisons of pregnancy and illegal drug use and displayed greater risk profiles on the measures of multiple sex partners (OR=2.08), assault (OR=2.6), depression (OR=2.06), suicide plans (OR=1.36), and younger sexual intercourse initiation (OR=1.67).


Previous research addressing factors such as truancy, mental health characteristics, sexual behavior, substance use and pregnancy have frequently compared sexually abused samples with non-abused samples, without recognizing that sexually active non-abused adolescents display more risky behaviors than their non-sexually active counterparts, and may therefore confound associations. Furthermore, much of this work has focused on clinical samples and utilized a broad definition of sexual abuse. A broad definition of sexual abuse generally includes incidences of fondling, genital manipulation, vaginal sex, anal sex and oral sex and may fail to distinguish between the form of abuse and the impact that these different forms may have upon victims. This study has utilized a large non-clinical sample from a predominantly rural state, and limited its focus to defining sexual abuse within the context of forced vaginal sex; a form of sexual abuse which may not be the most harmful form of sexual abuse (Wyatt, Newcomb & Riederle, 1993). However, 13% of the adolescents confirmed being victimized in this manner by identifying their perpetrators.

The focus of this study has been on examining adolescent females who identify themselves as having been forced to have vaginal sex and whether this group displays different markers based upon the identity of their perpetrator. Different profiles would assist physicians, psychologists, school counselors and other health care providers in directing initial communications about a history of sexual abuse. The most outstanding feature of these analyses is that there is no clear pattern that would confirm a hypothesis that certain deviant forms of behavior are a consequence of the emotional harm caused by victimization. It is noticeable that victim groups whose perpetrator was in the boyfriend, external family member or other/stranger group had considerably more risky profiles than the group whose perpetrator was an immediate family member. However, there were inconsistencies among the profiles across the four perpetrator groups.

Profiles that would help to identify sexually abused adolescent females include positive responses regarding multiple partners, a relatively young age for initiation of sexual behavior, or pregnancy and are all factors that individually increase the likelihood of sexual abuse by about twofold. A positive response to all three of these sexual behavior items clearly identifies an adolescent who is at greater risk of having experienced sexual abuse. If the adolescent also acknowledges frequent sadness, or suicidal thoughts, and the adolescent indicates that they have experienced assault then health care providers should suspect that the adolescent has been sexually abused. It is important to recognize that these characteristics are not likely to identify someone who has been victimized by an immediate family member.

Since many health care and school health providers interact with children and adolescents in a caregiver role, they have a unique opportunity to address issues that children and adolescents may be reluctant to discuss. It is well established that some victims of sexual abuse cope through denial and it is unlikely that health care providers will easily confirm abuse in such cases. However, the discussion of behaviors and factors related to sexual abuse provide professionals with additional insights that may assist in the detection of sexual abuse. Physicians, psychologists, counselors and other health care providers should carefully consider the development of a protocol questioning adolescent clients about their behaviors as one of their approaches in assessing the child’s sexual abuse status. This protocol should include a sexual behavior component, an illegal drug component, a mood component, and one that addresses possible assault. These insights should assist in confirming the possibility of sexual abuse.

Additional research needs to be conducted confirming the role of different types of perpetrators and how this places the victims at greater risk for deviant behaviors that can be used as profile markers.


Finkelhor, D., (1994). Immediate and long-term impacts of child sexual abuse. The Future of Children, 4(2): 54-69.

Green, A.H., (1996). Overview of child sexual abuse. In S. J. Kaplan (Ed.), Family violence: A clinical and legal guide. (pp. 73-104). Washington, D.C., American Psychiatric Press, Inc.

Nagy, S., DiClemente, R., & Adcock, A. G. (1995). Adverse factors associated with forced sex among southern, adolescent girls. Pediatrics, 96: 944-946.

Nagy, S., Adcock, A. G., & Nagy, M. G. (1994). A comparison of risky health behaviors of sexually active, sexually abused and abstaining adolescents. Pediatrics, 93(4): 570-575.

Nagy, S., & Adcock, A. G. (1992). Summary Report II, The Alabama Adolescent Health Survey: Health Knowledge and Behaviors. ERIC Resources in Education 338 608.

Nagy, S. & Adcock, A. G. (1990). Summary Report: The Alabama adolescent health survey: Health knowledge and behaviors of grade 8 and 10 students. ERIC Resources in Education Document: SP 032 094.

Paradise, J.E., Rose, L., Sleeper, L.A., & Nathanson, M., (1994). Behavior, family function, school performance, and predictors of persistent disturbances in sexually abused children. Pediatrics, 93(3): 452-459.

Rind, B., & Tromovitch, P., (1997). A meta-analytic review of findings from national samples on psychological correlates of child sexual abuse. Journal of Sexual Research, 34: 237-255.

Wyatt, G.E., Newcomb, M.D., & Riederle, M.H., (1993). Sexual abuse and consensual sex, Sage, Newbury Park, CA.

Stephen Nagy, Ph.D., The University of Alabama, P.O. Box 870311, Tuscaloosa, AL. 35487, Snagy@ches.ua.edu.

COPYRIGHT 2000 University of Alabama, Department of Health Sciences

COPYRIGHT 2001 Gale Group