Prior to the 1980s, adolescent sex offenses were not taken seriously; typically they were explained as normal experimentation or developmental curiosity. The problem was either minimized or denied. Although research has mushroomed in the last decade, it has been concerned primarily with individual characteristics of offenders and offenses or with the family background of offenders. Information is sparse about the family environments of these offenders, and it is also unclear how the family environments of adolescent offenders differ, if they do, from those of juvenile delinquents who have committed either violent or nonviolent offenses or from the environments of normal adolescents, since few studies compare these groups or include a representative sample of nonoffenders as a comparison group.
This study was designed to assess adolescent sex offenders' perceptions of their family environments. The perceptions of male adolescent sex offenders were compared with those of male juvenile delinquents who had committed violent nonsex offenses, male juvenile delinquents who had committed nonviolent nonsex offenses, and the perceptions of a normative adolescent sample. Family environment was examined, using the Family Environment Scale (FES) (Moos & Moos, 1986) which assesses relationship, personal growth, and system maintenance dimensions.
Families of adolescent offenders are frequently involved in treatment, and findings from this study are likely to enhance services to these families. In residential treatment facilities, adolescent sex offenders and other juvenile delinquents are frequently placed together, but most experts advocate offense-specific treatment for sex offenders (Knopp, 1985). Professionals in the sex-offender field claim that adolescent sex offenders are unique and distinct from other delinquent and nondelinquent adolescents (Knopp, 1985; O'Brien, 1985). Results from this study may help clarify whether, and in what ways, the family environments of adolescent sex offenders differ from those of juvenile delinquents and from a normative adolescent population. Findings may also promote development of theories on the etiology of sex offenses by adolescents, and help in the early identification of at-risk families.
The Uniform Crime Reports (Federal Bureau of Investigation, 1991) indicate that in 1991, 95,533 males were arrested for sex offenses, including forcible rape but not prostitution. Seventeen percent (15,760) of these offenders were under the age of eighteen. These figures are probably underestimates due to the high number of offenses that go unreported. Ryan (1988), using the data from the Uniform Data Collection System of the National Adolescent Perpetrator Network in which the majority of the offenders were categorized as first-time offenders, found that the average number of victims per offender was seven, indicating that many offenses were not reported. In addition to underreporting, only a small number of complaints ever result in an arrest (Groth & Loredo, 1981). Ageton (1983), in a general population study, found that 3-4% of adolescents aged 15-20 had committed a sex offense. She estimated that 500,000 offenses occur annually, a vast difference from the less than 16,000 arrests in 1991.
Studies of adult sex offenders (Abel, Mittleman, & Becker, 1984; Becker & Abel, 1985) indicate that about half of adult offenders report that their first sex offense occurred as an adolescent, and often offenses escalated in frequency and severity over time. These alarming findings have led to increased efforts to identify and treat adolescent sex offenders and to recognize this group as a distinct juvenile justice problem and clinical population.
A thorough examination of the various sociocultural contexts of adolescent sex offenders is beyond the scope of this study; however, one significant context that will be examined is the family environment. Areas of that environment which have been identified as indicators of family functioning include relationship, which assesses the degree of commitment and support family members provide one another, personal growth, which assesses the extent to which family members are able to make their own decisions, and system maintenance, which assesses the degree of importance of clear organization and structure in planning family activities and setting family rules (Moos & Moos, 1986).
Aside from descriptive studies, which primarily have addressed individual offender and offense characteristics and family background, research on adolescent sexual offenders is limited many by small sample size and geographical bias. Studies including normal control groups are sparse (Davis & Leitenberg, 1987). Although clinical impressions abound, little scientific research has been conducted to confirm these impressions. This is especially true of research on the family environments of these adolescents, though there is conjecture on how the family influences the commission of an offense. Reviewing the literature, Monastersky and Smith (1985) concluded that studies are virtually unanimous in identifying the family as a crucial influence in the development or elicitation of the offending behavior, but it is not clear how this occurs.
Knopp (1982), reporting on limited clinical impressions of an unspecified number of families in one adolescent sex-offender program, found that the families were either very rigid and enmeshed, or very chaotic with a great deal of role confusion. In a review of another adolescent sex-offender program serving very violent and dangerous offenders in a long-term locked facility, Knopp reported that staff used the word "chaotic" when describing the families of the majority of the adolescent sex offenders. It was not uncommon for one of the parents to have demonstrated deviant behavior very similar to that of the child.
Thus, families of adolescent sex offenders have been described in the literature as either rigid and enmeshed or chaotic. However, the authors typically have based their conclusions solely on their clinical experiences.
Three studies have addressed adaptability and cohesion of adolescent sex-offender families formally, using the Family Adaptability and Cohesion Evaluation Scales developed by Olson and colleagues and revised in 1986 (FACES-II and III). In one of those studies Bera (1985) found no significant difference between the family systems of adolescent sex offenders and those of normative adolescents.
While Bera used only the adolescents' perception of their family systems, Smith and Monastersky (1987) gathered perspectives from adolescent sex offenders (n = 66) and their mothers (n = 71) and fathers (n = 51). Their sample included offenders treated in an outpatient program, a majority of whom had committed less aggressive offenses. They found that the families of adolescent sex offenders were more likely than the general population to be characterized as rigid in response to changes (i.e., low adaptability) and emotionally disengaged (i.e., low cohesiveness). In their study, degree of violence of the offense was related to the family system; the more rigid and disengaged the family (according to the parents' perceptions) the more violent the offense. These findings are consistent with correlations McGaha and Fournier (1988) found between level of rigidity and disengagement in the family system and seriousness of offense within the general juvenile delinquent population. Furthermore, Smith & Monastersky (1987) suggest that future research compare the family systems of adolescent sex offenders and nonsexual offending juvenile delinquents in order to identify distinguishing characteristics of the groups. Neither of these studies (Bera, 1985; Smith & Monastersky 1987) has been published in the professional literature.
One recent study used the newer version of FACES-III (Olson, 1986) to assess adolescents' perception of family adaptability and cohesion in order to determine whether and in what ways the family systems of adolescent sex offenders (n = 39) differed from violent juvenile delinquents (n = 25), nonviolent juvenile delinquents (n = 41), and normal families (normative data) (Bischof, Stith, & Wilson 1992). The data revealed that families of sex offenders are characterized by greater family cohesion when compared with other delinquents, but that sex offenders perceive their families as less cohesive than do members of normal families. No significant differences between the groups were found for family adaptability. Therefore, this study indicated that while adolescent sex offenders perceived their families as more helpful and supportive as compared to violent delinquents, they view their families as less helpful and supportive as compared to adolescents from normal families.
While the Family Environment Scale (FES) has not been used to assess families of adolescent sex offenders, it has been used to assess distressed families with adolescents. A variety of studies have been conducted using the FES to compare distressed and nondistressed families. For example, studies have compared delinquents and nondelinquents (LeFlore, 1988), examined adolescent adjustment (Kleinman et al., 1989), studied adolescent depression and suicidal ideation (Friedrich, Reams, & Jacobs, 1982), compared achievers and underachievers in high school (Wood, Chapin, & Hannah, 1988), studied females with various types of eating disorders (Stern et al., 1989), examined the family environments of uncontrollable adolescents (Kirst-Ashman, 1983), and female college students who had been sexually abused during childhood (Long & Jackson, 1991).
Comparing delinquent and nondelinquent adolescents, LeFlore (1988) found that delinquents (n = 68) perceived their families as having less active-recreational orientation, less cohesiveness, and less expressiveness than a matched group of nondelinquents (n = 130). The average age of the delinquents was 15.4 years, they were considered serious and repeat offenders, and came from families of a lower to lower-middle socioeconomic level.
Another study which used the FES (Kleinman et al., 1989) found the following family factors related to distress in males: low levels of cohesion, active-recreational orientation and expressiveness, and high levels of conflict. Similarly, Friedrich et al. (1982) studied depression and suicidal ideation in adolescents and found that depression is associated with a family environment that is less cohesive, less active-recreationally oriented, and more conflicted. In addition, severity of suicidal ideation was related to less cohesiveness, independence, and organization and more achievement orientation.
In summary, the family environment is a significant context for adolescents. A review of nine studies which directly assessed distressed adolescents' perceptions of their family environments (Fox et al., 1983; Friedrich et al., 1982; Kirst-Ashman, 1983, 1983; Kleinman et al., 1989; LeFlore, 1988; Long & Jackson, 1991; Moos & Moos, 1986; Stern et al., 1989; Wood et al., 1988) revealed several consistent differences between distressed and nondistressed families of adolescents. In nine of ten comparisons, the distressed families were reported as less cohesive or supportive. Seven in ten comparisons showed that distressed families were also less expressive of their feelings. Six in ten showed less participation in social and recreational activities. Half the comparisons showed that the distressed families reported more conflict among family members. Inconsistent results were reported with regard to level of independence among family members, level of achievement orientation, level of intellectual-cultural orientation, active-recreational orientation, moral-religious emphasis, organization, and control between groups.
However, no study could be located which directly examined the family environments of adolescent sex offenders and compared these environments with those reported by other delinquents and/or by normative adolescent samples. Therefore, this study was designed to address these gaps in the literature.
Questionnaires were distributed to 109 adolescent males in various outpatient and residential programs; 105 (96.3%) questionnaires were returned. Participants were adolescent males, aged 12-18, who were grouped as follows: (a) sex offenders (n = 39) who self-reported having committed child sexual molestation or who were involved in treatment programs designated for identified adolescent sex offenders; (b) nonsex offenders (n = 25) who self-reported committing violent offenses (e.g., homicide, manslaughter, robbery, aggravated assault); and (c) nonsex offenders (n = 41) who self-reported committing nonviolent offenses (e.g., against persons and property, status offenses, substance abuse violations). A nondelinquent control group was not included in this study, but FES has been normed to the general population of adolescent families and these norms were compared to scores by the groups involved. Subjects were not paid, but had been encouraged to participate as a way to help other young people and society in general.
Family Environment Scale (FES) Form-R (Moos & Moos, 1986). The FES is a true-false self-report instrument designed to measure the social-environmental attributes of various types of families. Form-R measures respondents' perceptions of their nuclear family environment. There are 10 subscales with nine questions each, which assess relationship, personal growth, and system maintenance dimensions. Definitions for the subscales are shown in Table 1. Test-retest reliability for the subscales were all acceptable and ranged from a low of 0.68 for independence to a high of 0.86 for cohesion. The authors report a satisfactory level of internal consistency. The FES has been used extensively with many types of families and in several studies comparing distressed with normal families.
Delinquency and sex offense self-report. Participants were provided a list of several categories of delinquent acts and sex offenses, and were asked to indicate whether they had committed an offense from each category and whether they had been held by the police or convicted for any of the offenses. The seven categories included: nonviolent offenses against persons; general sex offenses; child molestation; offenses against public order and drug-abuse violations; and status offenses.
A paper-and-pencil self-report survey was administered to the participants by a research investigator or treatment professional. Persons participating in programs/facilities of the study signed a statement indicating their understanding of the purpose and procedures. Both the youth and the parents, guardians, or custodians gave their written permission and informed consent for the youth to participate.
Table 1 FES Subscales and Dimension Descriptions RELATIONSHIP DIMENSIONS Cohesion the degree of commitment, help, and support family members provide for each other Expressiveness the extent to which family members are encouraged to act openly and to express their feelings directly Conflict the amount of openly expressed anger, aggression, and conflict among family members PERSONAL GROWTH DIMENSION Independence the extent to which family members are assertive, are self- sufficient, and make their own decisions Achievement Orientation the extent to which activities are cast into an achievement-oriented or competitive framework Intellectual-Cultural the degree of interest in Orientation political, social, intellectual, and cultural activities Active-Recreational the extent of participation in Orientation social and recreational activities Moral-Religious Emphasis the degree of emphasis on ethical and religious issues and values SYSTEM MAINTENANCE DIMENSIONS Organization the degree of importance of clear organization and structure in planning family activities and responsibilities Control the extent to which set rules and procedures are used to run family life (Moos & Moos, 1986, p. 2)
Analyses of Variance (ANOVAs) were executed among the three groups on demographic factors. There was a significant difference in age among the three groups (F = 6.41, p = .002); mean age for the adolescent sex-offender group was 15.39, whereas mean age for violent and nonviolent juvenile delinquents was 16.16 and 16.34, respectively. While difference in age was statistically significant, the groups fall within the same developmental range of middle adolescence. Results from a chi-square analysis showed no significant differences in race among the three groups (73% white, 16% black, 3% Latino, 3% Asian, 5% other for entire sample).
Participants were predominantly from the Washington, D.C. metropolitan area. Level of family income differed significantly among the three groups (F = 4.44, p = .01). Mean family income was $38,400, $41,000, and $52,000 for adolescent sex offenders, violent juvenile delinquents, and nonviolent delinquents, respectively. The majority of fathers (74%) and mothers (71%) were employed full-time. Mothers most frequently worked in service occupations (33%) or in technical or clerical work (32%). Fathers worked primarily in service or military occupations (26%) and in administration, engineering or scientific endeavors (19%). There were no significant differences among the groups for parents' employment level or occupation.
In summary, sex offenders were slightly younger than both violent and nonviolent delinquents, but there were no significant differences in race, parental employment or occupation. Nonviolent delinquents in this study tended to come from families with higher incomes.
The results of this study point to significant differences among the four groups - adolescent sex offenders, violent juvenile delinquents, nonviolent juvenile delinquents, and normal adolescents - on six of the ten dimensions assessed by the Family Environment Scale (Table 2). There were no significant differences among the groups on level of conflict in the family, achievement orientation, moral-religious emphasis, or the amount of organization displayed by the family. There were differences among the groups on the other six subscales, as evidenced by the post hoc Duncan tests.
Each delinquent group differed from the normal group in perception of the amount of cohesion in their family (F = 15.58, df = 3, p [less than] .001). Each delinquent group considered their family to be considerably less cohesive than did the normal adolescent group. There were no significant differences among the groups of offenders on level of cohesion.
Each delinquent group also differed from the normal adolescents in perception of the amount of expressiveness in their family (F = 7.41, df = 3, p [less than] .001). Each delinquent group considered their family to be considerably less expressive than did the normal adolescent group. There were no significant differences among the groups of offenders on level of expressiveness.
[TABULAR DATA FOR TABLE 2 OMITTED]
There were also differences among groups on level of independence in the family (F = 6.34, df = 3, p [less than] .001). Each group of delinquents perceived their family as having a lower level of independence among family members than did the normal families. There were no significant differences among the groups of offenders on level of independence.
Significant differences also existed among the groups on level of intellectual-cultural orientation (F = 6.45, df = 3, p [less than] .001). Normal families (M = 5.23) were perceived as having significantly higher levels of intellectual-cultural orientation than did the families of violent and nonviolent delinquents. Families of sex offenders did not differ from any of the other groups on this measure.
A significant difference was also found among the groups on level of active-recreational orientation (F = 2.84, df = 3, p [less than] .05). Normal families (M = 5.75) were perceived as having higher levels of active-recreational orientation than did nonviolent delinquents (M = 4.97) while no other differences were found among the groups.
Finally, there were significant differences among the groups on level of control (F = 2.67, df = 3, p [less than] .05). Normal families (M = 4.87) were perceived as having significantly lower levels of control than did families of nonviolent delinquents (M = 5.68) while no other differences were found between the groups.
Thus, while there were differences between the perceptions of normal and delinquent adolescents on several aspects of family environment, there were no significant differences between the groups of delinquents.
Adolescent sex offenders' perceptions of their family environment were compared with the perceptions of violent and nonviolent delinquents. These data were compared with normative data for normal adolescents and their families. Ten aspects of family environment were considered. It was expected that the family environments of adolescent sex offenders would differ from violent and nonviolent delinquents and from normal adolescents, thus supporting recognition of adolescent sex offenders as a distinct juvenile justice and clinical treatment population.
The lack of support for this hypothesis suggests that if the families of adolescent sex offenders do indeed differ from those of other delinquents, they do so on dimensions of family life other than those assessed by the FES. In fact, from clinical experience, Smith and Monastersky (1987) proposed two primary distinguishing features of families of adolescent sexual offenders: denial of sexual tensions and a paucity of sexual knowledge or education. Further research is needed to verify these clinical impressions and to explore whether and how various types of sex offenders differ on important dimensions of family life.
Although the expected differences among groups of delinquent adolescents were not found, the findings of this study have a number of implications for intervention with families of adolescent sex offenders. Their family systems are similar to those of violent and nonviolent juvenile delinquents in most ways assessed by the FES. Therefore, it is likely that family interventions which have been demonstrated to be effective with juvenile delinquents in general are likely to be helpful with adolescent sex offenders. Clinical literature and experience with juvenile delinquents generally, are more advanced, and the lessons learned should be transferable to adolescent offenders and their families.
While similar to other delinquents' families, differences between the delinquent and the normative families did exist. The finding that delinquents view their families as less supportive than do normative families coincides with the findings of LeFlore (1988) and can be beneficial to clinicians. Of note is that similar results were found frequently in other studies of distressed adolescents (Fox et al., 1983; Friedrich et al., 1982; Kirst-Ashman, 1983; Kleinman et al., 1989; Leflore, 1988; Long & Jackson, 1991; Moos & Moos, 1986; Wood et al., 1988). Understanding adolescents' perspective can be helpful in creating strategies that fit with their world view. Utilization of interventions designed to enhance emotional support and bonding among family members would seem appropriate for most families of delinquents.
Delinquents in this study and those in LeFlore's 1988 study reported that their families were less expressive than normative families. This finding was also typical of other studies with distressed adolescents (Kirst-Ashman 1983; Kleinman et al., 1989; Long & Jackson, 1991; Moos & Moos, 1986; Wood et al., 1988). Clinicians may find it helpful to teach these families how to express their feelings more directly.
We also found that delinquent adolescents believed that their families encouraged independence or self-sufficiency less than did normal families. Therapists working with these families will need to determine if this difference is the result of living with a delinquent adolescent or indicates a lack of encouragement of independence which may in itself be a factor in the development of delinquency. In any event, the level of independence seems to be an important factor which differentiates delinquent from nondelinquent families.
Further, delinquent families seem to place less emphasis on intellectual-cultural as well as active-recreational activities than do nondelinquent families. LeFlore's (1988) findings are in agreement, that families of delinquents participate in fewer social and recreational activities than do families of nondelinquents. Research on other distressed adolescents also found this to be true (Fox et al., 1983; Friedrich et al., 1982; Kleinman et al., 1989; Moos & Moos, 1986; Wood et al., 1988). Both of these areas should be targeted for intervention in the treatment of these families.
Finally, we found that delinquent families were more bound to rules and procedures in structuring day-to-day family life. Again, it is unclear whether this is a result of delinquency or a factor in promoting delinquency. However, it may be important for therapists to examine this issue.
Limitations and Future Research Suggestions
Participants in this study were voluntary and self-selected, and therefore are not necessarily representative of the delinquent populations included here. Parental/guardian permission first had to be obtained, after which the adolescents decided if they would participate. Subjects were from programs/facilities that had agreed to cooperate with the study. Nonsex-offending delinquents were primarily from four residential programs; all but one was involved in a residential program. Adolescent sex offenders were closely divided between outpatients and residents. Placement in residential care often comes about only after other less restrictive alternatives have been exhausted, and frequently severe family dysfunction is a criterion for placement out of the home. Future studies should either include offenders from various levels of outpatient and residential treatment, or control for placement setting.
Information in this study was obtained by adolescent retrospective self-reports. Participants were asked to describe conditions at the time of their offense. In some cases, a long period of time had passed since the offense, inviting the possibility of treatment effect or inaccurate recall. For a more accurate and complete perspective of the family environment, information should be obtained early in the assessment or screening process and should include the perceptions of several family members. It is recommended that family instruments be included routinely in assessment, because families are frequently involved in treatment. A more comprehensive understanding of the family context would prove valuable in treatment planning and intervention design.
Detailed information about offenses and offense history was not obtained in this study. Future studies should attempt to differentiate family characteristics and environments among various types of adolescent sex offenders. Indeed, Bera (1985) found little difference between families of adolescent sex offenders and normal adolescent families in general, but significant differences emerged between various types of adolescent sexual offenders, classified according to offenses and offense patterns.
Abel, G., Mittleman, M., & Becker, J. (1984). Sexual offenders: Results of assessment and recommendations for treatment. Unpublished manuscript, available from NY State Psychiatric Institute, Sexual Behavior Clinic.
Ageton, S. S. (1983). Sexual assault among adolescents. Lexington, MA: Heath.
Becker, J. V., & Abel, G. G. (1985). Methodological and ethical issues in evaluating and treating adolescent sex offenders. In E. M. Otey, & G. D. Ryan (Eds.), Adolescent sex offenders: Issues in research and treatment (pp. 109-129). Rockville, MD: USDHHS.
Bera, W. H. (1985). A preliminary investigation of a typology of adolescent sex offenders and their family systems. Unpublished master's thesis, University of Minnesota.
Bischof, G., Stith, S., & Wilson, S. (1992). A comparison of the family systems of adolescent sexual offenders and nonsexual offending delinquents. Family Relations, 41, 318-323.
Davis, G. E., & Leitenberg, H. (1987). Adolescent sex offenders. Psychological Bulletin, 191, 417-427.
Federal Bureau of Investigation. (1991. Uniform crime reports for the United States, 1991. Washington, DC: U.S. Government Printing Office.
Fox, R., Rotatori, A., Macklin, F., Green, H., & Fox, T. (1983). Socially maladjusted adolescents' perceptions of their families. Psychological Reports, 52, 831-834.
Friedrich, W., Reams, R., & Jacobs, J. (1982). Depression and suicidal ideation in early adolescents. Journal of Youth and Adolescence, 11(5), 403-407.
Groth, A.N., & Loredo, C. M. (1981). Juvenile sexual offenders: Guidelines for assessment. International Journal of Offender Therapy and Comparative Criminology, 25, 31-39.
Kirst-Ashman, K. (1983). Exploration of the family environment and problems of uncontrollable adolescents. Unpublished doctoral thesis, University of Illinois, Urbana, IL.
Kleinman, S. L., Handal, P. J., Enos, D., Searight, H. R., & Ross, M. J. (1989). Relationship between perceived family climate and adolescent adjustment. Journal of Clinical Child Psychology, 18(4), 351-359.
Knopp, F. H. (1982). Remedial intervention in adolescent sex offenses: Nine program descriptions. Orwell, VT: Safer Society Press.
Knopp, F. H. (1985). Recent developments in the treatment of adolescent sex offenders. In E. M. Otey, & G. D. Ryan (Eds.), Adolescent sex offenders: Issues in research and treatment (pp. 1-27). Rockville, MD: USDHHS.
LeFlore, L. (1988). Delinquent youths and family. Adolescence, 23 (91)629-642.
Long, P., & Jackson, J. (1991). Children sexually abused by multiple perpetrators. Journal of Interpersonal Violence, 6(2), 147-159.
McGaha, J. E., & Fournier, D. G. (1988). Juvenile justice and the family: A systems approach to family assessment. Marriage and Family Review, 12, 155-172.
Monastersky, C., & Smith, W. (1985). Juvenile sexual offenders: A family systems paradigm. In E. M. Otey, & G. D. Ryan (Eds.), Adolescent sex offenders: Issues in research and treatment (pp. 164-175). Rockville, MD: USDHHS.
Moos, R. H., & Moos, B. S. (1986). Family Environment Scale Manual (2nd ed.). Palo Alto, CA: Consulting Psychologists Press.
O'Brien, M. (1985). Adolescent sexual offenders: An outpatient program's perspective on research directions. In E. M. Otey, & G. D. Ryan (Eds.), Adolescent sex offenders: Issues in research and treatment (pp. 147-163). Rockville, MD: USDHHS.
Olson, D. H. (1986). Circumplex model VII: Validation studies and FACES III. Family Process, 25, 337-351.
Ryan, G. (1988) The juvenile sexual offender: A question of diagnosis. Uniform Data Collection System, Data on 1000 Juvenile Offenders, presented April 1988: National Symposium on Child Victimization, Anaheim, California. Cited in Ryan, G. D. & Lane, S. L., Eds. (1991), Juvenile Sexual Offending. Lexington, MA: Lexington Books.
Smith, W. R., & Monastersky, C. (1987). Strategies for studying the role of the family in the commission of sexual offenses by adolescents. Unpublished manuscript.
Stern, S. L., Dixon, K. N., Jones, D., Lake, M. Nemzer, E., & Sansone, R. (1989). Family environment in anorexia nervosa and bulimia. International Journal of Eating Disorders, 8(1), 25-31.
Wood, J., Chapin, K., & Hannah, E. (1988). Family environment and its relationship to underachievement. Adolescence, 23(90), 283-290.…