Chapter 2: Etiology and Typologies of Juveniles Who Have Committed Sexual Offenses

Introduction

This chapter addresses two topics: the etiology of sexual offending by juveniles and typologies for juveniles who have committed sexual offenses.

The etiological research reviewed in this chapter addresses the causes or origins of juvenile sexual offending and the pathways related to the development, onset and maintenance of sexually abusive behavior in this population. Knowledge about the etiology of sexual offending is important because it provides both conceptual frameworks and specific guidance that can be used to develop more effective prevention efforts across a broad continuum, from primary to tertiary.¹

The typological research reviewed in this chapter addresses classification schemes based on types or categories of offenders or victims and offense characteristics. Empirically based typologies provide important information for clinical intervention by identifying key constructs for assessment, possible etiological factors specific to each subtype or typology of juveniles and unique risks and needs for each subtype that should be targeted in treatment (Faniff & Kolko, 2012). (For more information on assessment, see Chapter 4, "Assessment of Risk for Sexual Reoffense in Juveniles Who Commit Sexual Offenses," in the Juvenile section.) Aebi and colleagues (2012, p. 268) add that a validated typology "shows a specific profile of an offender, victim and offense characteristics that reflect [sic] underlying psychological processes" of the youth that are relevant to etiology, maintenance, treatment and recidivism. (For more information on the "Effectiveness of Treatment for Juveniles Who Sexually Offend," see Chapter 5, and for more information on "Recidivism of Juveniles Who Commit Sexual Offenses," see Chapter 3, both in the Juvenile section.) The information gained from typology research provides the foundation for designing and implementing more effective and efficient treatment programming and supervision protocols that reflect individualized risks and needs.

Summary of Research Findings

Etiology

The research on etiological factors for sexual offending includes studies that focus on single factors and those that focus on multiple factors. There appears to be a consensus in the field that etiological factors typically both co-vary and interact with each other in the development and onset of sexual offending and nonsexual delinquency.

Sexual Victimization

Etiological factors typically both co-vary and interact with each other in the development and onset of sexual offending and nonsexual delinquency.

Veniziano, Veniziano and LeGrand (2000) gathered information from a sample of 68 juveniles who had committed sexual offenses and were court ordered to a residential treatment facility. All of the juveniles had experienced sexual victimization. Information was gathered about their prior sexual victimization, and the characteristics and behaviors of their perpetrators. These data were compared to the adolescent offenders' behaviors in the sample and the characteristics of their victims. The results of the study supported the hypothesis that the juveniles who had been sexually victimized were more likely to select sexual behaviors that reflected their own sexual victimization in regard to age and gender of the victim and the types of sexual behaviors perpetrated against the victims. However, the relationship between prior victimization and subsequent offending was not as strong with respect to whether victims were inside or outside the family. The researchers concluded that findings from the study offered support for the notion that the sexual offending of some adolescents represents a re-enactment of their own sexual victimization, or a reactive conditioned and/or learned behavior pattern.

Grabell and Knight (2009) studied 193 juveniles who had committed sexual offenses, sampled from different inpatient treatment facilities. The study sought to examine child sexual abuse patterns and sensitive periods in the lives of juveniles who had committed sexual offenses. The results suggest a relationship between childhood sexual abuse and sexual fantasy in sexually abusive adolescents that is moderated by the age at which the abuse occurred. More specifically, it was found that ages 3 to 7 may be a sensitive period when sexual abuse can do the most damage and place a youth at higher risk for engaging in sexually abusive behavior later in life. The researchers compared their findings related to discrete periods of sexual abuse with those from an earlier study conducted by Burton, Miller and Shill (2002), and concluded that continuous sexual abuse was more likely related to severe perpetration than were discrete periods. Grabell and Knight concluded that both age and the length of the sexual abuse contribute to attitudes and behaviors in juveniles who commit sexual offenses.

Berman and Knight (2015) assessed a sample of 178 juveniles who committed sexual offenses from an inpatient treatment center. Two-thirds of the participants reported having experienced sexual abuse as a child. The study explored the nature of the relationship between the perpetrator and the victim as it relates to the factors important to subsequent sexual offending. The study examined multiple levels of interpersonal closeness (stranger, relative and cohabitant) between the perpetrators of the sexual abuse and their victims. The results presented were for a three-month minimum of cohabitation. Berman and Knight (2015) found that being sexually abused by a cohabitant perpetrator was the best predictor of subsequent sexualization (sexual preoccupation, sexual compulsivity and hypersexuality), callousness/manipulativeness and higher impulsivity or disinhibition. Child sexual abuse occurring in the context of an attachment relationship was discussed as it relates to a model for future relationships that encourage social learning of sexually abusive behavior. The results of the study supported the author's hypothesis that, "in studying the outcomes of sexual abuse, it is essential to differentiate levels of relatedness between victim and perpetrator" (p. 601).

Relationship Between Sexual Victimization and Personality Variables

Hunter and Figueredo (2000) focused on delineating the relationship between sexual victimization and personality variables in predicting patterns of sexual offending against children. Data were collected on 235 adolescents, representing subsamples of sexually victimized and nonvictimized, perpetrating and nonperpetrating, and emotionally maladjusted and nonmaladjusted youth. A younger age at time of sexual victimization, a greater number of incidents, a longer period of waiting to report the abuse and a lower level of perceived family support after revelation of the abuse were found to be predictive of subsequent sexual perpetration. "See Typologies" in this chapter.

Burton (2008) conducted an exploratory study of the contribution of personality traits and childhood sexual victimization to the development of sexually abusive behavior, thereby testing a social learning/victim-to-victimizer hypothesis for the development of sexually abusive behavior. The study compared 74 incarcerated sexual abusers and 53 nonsexual abusers. Similar to the findings of many previous studies, Burton found that adolescent sexual abusers tend to have higher rates of sexual victimization than nonsexually abusive youth. In addition, sexually abusive youth who had been sexually victimized themselves were likely to repeat what was done to them in regard to the relationship with and gender of their victim(s), modus operandi and sexual behaviors. These results suggest that sexually abusive youth may have learned to be sexually abusive from their own sexual perpetrator(s). The personality traits that contributed significantly to the social learning model were "submissive" and "forceful." Burton suggested that those who scored higher on the submissive trait may believe that close relationships with others are required to feel comfortable and socially confident, and that those who scored higher on the forceful personality trait may derive pleasure from inflicting pain on their victims and may attain this pleasure via forceful acts.

Multiple Types of Child Maltreatment

Awad and Saunders (1991) compared 49 male adolescents who sexually offended against females their age or older to 45 adolescents who engaged in sexually abusive behavior toward younger children. The results showed that the majority of the adolescents who sexually offended against females their age or older came from a disturbed family background. The sexual victimization rate for the adolescents who sexually offended against children was much higher. The researchers concluded that, for some of these adolescents, sexual aggression was a learned behavior, modeled after what they observed at home. "See Typologies" in this chapter.

Kobayashi and colleagues (1995) tested a theoretical model of the etiology of deviant sexual aggression by adolescents that included several family factors: perceived parental deviance, child physical and sexual abuse history, and a child's bonding to parents. Study subjects consisted of 117 juvenile males who committed sexual offenses and who had been referred to a treatment clinic. Results indicated that paternal physical abuse and sexual abuse by males increased sexual aggression among adolescents and that mother-child bonding had the opposite effect. The results can be explained from a social learning and a parent-child attachment or social control perspective. In addition, the researchers suggested an alternative perspective from evolutionary psychology to explain the findings. Kobayashi and colleagues noted that the behavior developed by juveniles who sexually offend may be the result of social modeling and highlighted the ethological literature related to sexual imprinting in animals to support this perspective.

Cavanaugh, Pimenthal and Prentky (2008) studied a sample of 667 boys and 155 girls involved with social services, the vast majority of whom had engaged in hands-on sexualized behaviors. Almost all of the youth came from "highly dysfunctional" families and had experienced a high degree of physical, psychological and sexual abuse as well as neglect. The researchers found that 66.7 percent of the study subjects had attention-deficit/hyperactivity disorder (ADHD), 55.6 percent had posttraumatic stress disorder (PTSD) and 49.9 percent had a mood disorder. Approximately one-quarter used drugs and about one-fifth consumed alcohol. These findings highlight the importance of assessing and treating co-occurring issues, which can often be influential in sexual offending behaviors.

Seto and Lalumiere (2010) tested special and general explanations of male adolescent sexual offending by conducting a meta-analysis of 59 independent studies comparing male adolescents who committed sexual offenses with male adolescents who committed nonsexual offenses (n = 13,393) on theoretically derived variables reflecting general delinquency risk factors (antisocial tendencies), childhood abuse, exposure to violence, family problems, interpersonal problems, sexuality, psychopathology and cognitive abilities. The results did not support the notion that adolescent sexual offending can be explained as a simple manifestation of general antisocial tendencies. Adolescents who committed sexual offenses had much less extensive criminal histories, fewer antisocial peers and fewer substance abuse problems compared with nonsexual offenders. Special explanations for adolescent sexual offending suggested a role for sexual abuse history, exposure to sexual violence, other abuse or neglect, social isolation, early exposure to sex or pornography, atypical sexual interests, anxiety and low self-esteem. Explanations focusing on attitudes and beliefs about women or sexual offending, family communication problems or poor parent-child attachment, exposure to nonsexual violence, social incompetence, conventional sexual experience and low intelligence were not supported. Ranked by effect size, the largest group difference was obtained for atypical sexual interests, followed by sexual abuse history for adolescents who had committed sexual offenses and, in turn, criminal history, antisocial associations and substance abuse for nonsexual offenders.

Leibowitz, Burton and Howard (2012) collected data from 478 youth, comparing sexually victimized and nonsexually victimized adolescent sexual abusers with a group of nonsexually victimized delinquent youth. The researchers found that the sexually victimized sexual abusers had the highest mean scores on trauma and personality measures, followed by nonsexually victimized sexual offenders and general delinquent youth. The sexually victimized sexual abusers reported experiencing significantly greater levels of all five types of abuse (emotional abuse, emotional neglect, physical abuse, physical neglect and sexual victimization) than the other two groups. General delinquent youth had fewer behavioral and developmental problems than victimized and nonvictimized juveniles who commit sexual offenses. This difference between general delinquency youth and juveniles who commit sexual offenses has not been found in other studies (as summarized by Seto and Lalumière's 2010 meta-analysis).

Wanklyn et al. (2012) focused exclusively on childhood risk factors in comparing three groups of juvenile offenders on risk factors, including individual, family, peer and school domains:

• Violent non-sex offenders (VNSO): these offenders committed murder, assault, robbery and weapons offenses (n = 172)
• Pure sex offenders (PSO): these offenders committed violent sex offenses (n = 28)
• Versatile violent sex offenders (VVSO): these offenders committed violent sex and non-sex offenses (n = 24).

 

Findings from the study provide support for both the specialist and generalist models of sexual offending. Both subtypes of sex offending youth (VVSO and PSO) exhibited higher odds of sexually precocious behaviors. Extrafamilial sexual abuse and intrafamilial child maltreatment were found in higher frequency among the juvenile sexual offending groups. A number of findings suggested that VVSOs may encompass a more severe and high-risk subset of the juvenile sexual offending population. The percentage of VVSO youth engaging in antisocial behaviors was nearly double that of the other subgroups. VVSOs also had greater odds of growing up with criminal family members and experienced a higher incidence of a variety of childhood risk factors. This study provided a more robust understanding of different predictors of subgroups of violent juvenile offending, as well as the different developmental pathways experienced by violent sex and non-sex offenders. In discussing future research needs, Wanklyn et al. suggested that, "Further research should explore possible causal mechanisms that link distal risk factors to offending outcomes through more proximal mediating risk variables, such as substance use and coping styles … Understanding the unique childhood risk factors for each subgroup may assist in the early identification of high-risk children and inform the development of prevention strategies tailored to each group's specific risk factors" (p. 2140-2141).

Marini et al. (2014) studied the relationships among childhood victimization, substance use prior to the commission of a sexual offense and force used during a sexual offense in a sample of 406 residentially based, male juveniles who committed a sexual offense. Childhood victimization included sexual victimization, physical and emotional abuse, neglect and emotional neglect, as well as cumulative exposure to childhood victimization. Substance use was assessed in order to investigate the potential mechanism of disinhibition as it relates to the relationship between substance use and use of force. The authors found that substance use prior to committing a sexual offense partially accounted for the effects of multiple types of cumulative childhood victimization experiences on the force used during a sexual offense: Youth who used drugs or alcohol prior to a sexual offense used more force than those who did not. Contrary to expectations, the authors did not find that the relationship among victimization, substance use and severity of sexual offending was strongest for juveniles who committed sexual offenses and had experienced sexual abuse themselves. Proposed clinical implications included treating alcohol use as part of specialized treatment, as well as addressing the sequelae of complex trauma utilizing evidence-based interventions such as trauma-focused cognitive-behavioral therapy.

Felizzi (2015) studied 502 male adjudicated juvenile sexual and nonsexual offenders in a large Midwest state, exploring the effects of parent or caregiver instability and disrupted parental attachment on juvenile sex offending. Felizzi found that the sexually offending juveniles experienced significantly greater parent or caregiver instability, as characterized by "lots of moves or homelessness, followed by sexual abuse in the home; children placed out of the home; neglect by caregivers; observing hitting, punching or slapping; and physical abuse" (p. 650). The nonsignificant finding related to parental attachment appeared to contradict the findings of some previous research. The results of the Felizzi study support social learning as a powerful influence for the developing child and as an important explanation for juvenile sexual offending. Other factors that should be considered include television, access to internet pornography, drug and alcohol use and peer influences.

Relationship Between Multiple Types of Child Maltreatment and Personality Variables

Johnson and Knight (2000) studied 122 juveniles who committed sexual offenses and were in inpatient treatment centers. The researchers explored developmental pathways possibly conducive to adolescent sexually abusive behavior, measuring the extent to which the sample experienced childhood trauma, engaged in adolescent delinquency and exhibited particular personality dispositions and cognitive biases. The results suggest that sexual compulsivity and hypermasculinity, through misogynistic fantasy behavior, significantly discriminate verbally and physically coercive juveniles who commit sexual offenses from those who do not report using force in their offenses. Sexual victimization directly and indirectly (via sexual compulsivity) affected sexual coercion. The study's results also suggest that alcohol abuse may play a more salient role in the expression of juvenile sexual coercion than previously hypothesized. Physical abuse had an indirect effect on sexual coercion and was found to be predictive of delinquent behaviors such as peer aggression and adolescent alcohol abuse.

Knight and Sims-Knight (2004) studied 218 juveniles who were adjudicated for sexual offenses and resided in inpatient specialized treatment facilities. As part of the study, the researchers presented a three-path model intended to serve as a framework for understanding sexually abusive behavior toward women. Knight and Sims-Knight emphasized that an alternative model should be developed for sexually abusive behavior toward children. The three latent traits that identified the paths are sexual drive/preoccupation, antisocial behavior/impulsivity and callous/unemotional trait. The paths predicted sexual coercion against women among juveniles who have committed sexual offenses. The researchers found that early traumatic physical and sexual abuse play an important etiological role, increasing the likelihood of sexually abusive behavior either directly by themselves or indirectly through the three intervening paths. The authors assert that these traits play a critical role across the life span, are critical in assessing risk of recidivism and should be targets of treatment. See "Typologies" in this chapter.

In contrast to the above study that presented a model for sexually abusive behavior toward women, Daversa and Knight (2007) focused on an etiological model for sexual offending behavior toward younger victims. Data were gathered on 329 juveniles from inpatient treatment facilities in four states, all of whom had committed a sexual offense. The results provided evidence that various developmental and early childhood maltreatment experiences and specific, mediating personality traits contribute significantly to predicting adolescent sexual offending against younger victims. Four significant paths emerged in the model (Daversa & Knight, 2007):

1. From emotional and physical abuse, through psychopathy and sexual fantasy, to child fantasy and child victimization
2. From emotional and physical abuse; through sexual inadequacy, sexual fantasy and child fantasy; to child victimization
3. From emotional and physical abuse, through sexual inadequacy, to child fantasy and child victimization
4. From sexual abuse directly to child victimization

The direct path from a history of sexual abuse to the sexual victimization of children is consistent with the finding that a disproportionate number of sexually abusive adolescents also report being victimized sexually. The researchers suggest that a subset of these sexually victimized offenders may select victim(s) specific to a particular age group that is consistent with the age at which they were victimized, indirectly supporting the victim-to-victimizer theory of adolescent sexual offending. The authors assert that this study provides data for the preliminary design of a dimensional model of adolescent sexual abusive behavior against younger children. See "Typologies" in this chapter.

Zakireh, Ronis and Knight (2008) examined the individual beliefs and attitudes, and victimization histories, of 100 male youth aged 13–19. The youth were divided equally into four demographically similar groups: 1) sexual offenders in residential placement, 2) sexual offenders in outpatient treatment, 3) nonsexual offenders in residential placement and 4) nonsexual offenders in outpatient treatment. The sexually offending youth included those who had exclusively offended against peer age and adult victims, those who had exclusively offended against children younger than age 12 and those who offended against mixed-age victims. The authors found that three categories of risk factors — greater hypersexuality or sexual deviance, more violent behavior or fantasies and an increased history of victimization — are consistent with path models that predict sexually abusive behavior toward peers and adults. The authors asserted that their findings were consistent with hypotheses about the significant role that the domains of callousness, unemotionality and antisocial behavior play in sexually abusive behavior against peers and the limited etiological role they play in sexually abusive behavior toward children. The study's findings are consistent with past evidence regarding the role that sexual victimization plays in subsequent sexual offending behavior. See "Typologies" in this chapter.

Zakireh, Ronis and Knight (2008) examined the individual beliefs and attitudes, and victimization histories, of 100 male youth aged 13–19. The youth were divided equally into four demographically similar groups: 1) sexual offenders in residential placement, 2) sexual offenders in outpatient treatment, 3) nonsexual offenders in residential placement and 4) nonsexual offenders in outpatient treatment. The sexually offending youth included those who had exclusively offended against peer age and adult victims, those who had exclusively offended against children younger than age 12 and those who offended against mixed-age victims. The authors found that three categories of risk factors — greater hypersexuality or sexual deviance, more violent behavior or fantasies and an increased history of victimization — are consistent with path models that predict sexually abusive behavior toward peers and adults. The authors asserted that their findings were consistent with hypotheses about the significant role that the domains of callousness, unemotionality and antisocial behavior play in sexually abusive behavior against peers and the limited etiological role they play in sexually abusive behavior toward children. The study's findings are consistent with past evidence regarding the role that sexual victimization plays in subsequent sexual offending behavior. See "Typologies" in this chapter.

Additional Etiological Factors

Burton, Leibowitz and Howard (2010) compared pornography exposure between male adolescents who sexually abuse and male nonsexual offending delinquent youth. Although previous literature indicates that pornography use for adult males at risk for aggression may result in sexually aggressive behavior, very little research has been reported on exposure to pornography on the part of juveniles who commit sexual abuse. The juveniles who had engaged in sexually abusive behavior reported more exposure to pornography when they were both younger and older than age 10 than nonsexual abusers. However, their exposure was not correlated to the age at which their sexually abusive behavior started, to the reported number of victims or to sexual offense severity. The exposure subscale before age 10 was not related to the number of children the group sexually abused, and the forceful exposure subscale was not correlated with either arousal to rape or degree of force used by the youth. Finally, exposure was significantly correlated with all of the nonsexual crime scores in the study. The researchers characterized this study as exploratory in nature and stated that no clear conclusions can be drawn regarding prohibitions or control of pornography for adolescents who sexually abuse and who are in treatment or on parole or probation.

Yoder, Leibowitz and Peterson (2016) explored the differences in characteristics of parental and peer attachments between youth who had committed sexual offenses (n = 355) and nonsexual offending youth (n = 150). The authors found that, compared to nonsexual offending youth, those who had committed sexual offenses had higher alienation as well as lower trust and communication with mothers. Youth who had committed sexual offenses also had higher alienation, and lower trust and overall attachment with fathers. There were no statistically significant differences between the groups on peer attachment. The study findings are consistent with previous research demonstrating an association between unmet attachment and socio-emotional needs with sexual offending, and they underscore the importance of parent-child relationships as a factor in the etiology of juvenile sexual offending. The study findings related to nonsexual offending youth also are important given the robust empirical evidence that juveniles who commit sexual offenses are more likely to reoffend nonsexually. The authors stated that there is a need for additional research focused on the interactions among social isolation, family dynamics, peer attachment and severity of offending. Replication of the current study's findings would lend support for treatment approaches that target family-level attachment characteristics through multimodal intervention approaches.

Boonman et al. (2015) conducted a meta-analysis involving 21 studies and a total sample size of 2,951 juveniles who sexually offended and 18,688 nonsexual delinquents to determine the prevalence of mental disorders in juvenile sex offenders. Differences in mental disorders between sexually offending juveniles and nonsexual delinquents were also assessed. The meta-analysis found that mental disorders are highly prevalent in juvenile sex offenders, as 69 percent had at least one mental disorder and 44 percent had at least two. Further, 51 percent were reported to have conduct disorder; 18 percent had anxiety disorder (PTSD = 8 percent); 14 percent had ADHD; 9 percent had affective disorder; 30 percent had at least one substance use disorder (SUD); and 42 percent had paraphilia.

Nonsexual delinquents were more likely than juvenile sex offenders to have externalizing disorders and to meet the criteria for one or more mental disorders and were more often diagnosed with ADHD, disruptive behavior disorder and SUD. Related to juvenile sex offenders, the prevalence of conduct disorder and alcohol and drug abuse/dependence disorders increased with age while PTSD and oppositional defiant disorder decreased with age. Substance use disorder increased overall when the percentage of females included in the studies increased. Related to this finding, the authors recommended the underlying causes of SUD, such as traumatic experiences and PTSD, should receive focus in the assessment and treatment of females who have committed sexual offenses.

Summary on Etiology

Knight and Sims-Knight (2004, p. 49) provide an excellent synthesis on the importance of etiology regarding the treatment, supervision and policy response to juvenile sexual offending:

Identifying the developmental antecedents of sexual aggression not only informs treatment planning (i.e., tertiary intervention), but also will ultimately be the basis for identifying at-risk groups for primary and secondary interventions. Having a validated model of the etiology of sexual aggression is the cornerstone of any public health approach to sexual aggression and a necessary prerequisite for implementation of a primary prevention perspective.

The research cited above describes both single- and multiple-factor etiological theories. There is strong evidence that sexual victimization plays a disproportionate role in the development of sexually abusive behavior in adolescents. A number of studies have described a direct path from sexual victimization to sexually abusive behavior, and others have described an indirect path that is affected by personality variables. Overall, the empirical evidence supports the notion that sexual abuse should not be examined in isolation as it clearly co-varies with other developmental risk factors. Much of the research has described multiple-factor theories in which early childhood maltreatment (traumatic physical and sexual abuse, neglect and chaotic family environments) increases the likelihood of sexually abusive behavior, either directly or indirectly, in relationship with more proximal mediating variables, including personality variables, coping styles, substance use, etc. See table 1 for a summary of the etiological research.

 

Table 1. Summary of Etiology Research

Study	Focus	Findings
Sexual Victimization
Veniziano, Veneziano & LeGrand (2000)	Sexual victimization and subsequent sexual offending	Sexual offending of some adolescents represents a reenactment of their own sexual victimization or a reactive conditioned and/or learned behavior pattern.
Berman & Knight (2015)	Nature of relationship between a perpetrator and the victim as it relates to the factors important to subsequent sexual offending	Being sexually abused by a cohabitant perpetrator was the best predictor of subsequent sexualization, callousness/manipulativeness and higher impulsivity or disinhibition as it relates to juvenile sexual offending.
Grabell & Knight (2009)	Child sexual abuse patterns and sensitive periods in juveniles who had committed sexual offenses	Ages 3–7 may be a sensitive period during which sexual abuse can do the most damage and place a youth at higher risk for engaging in sexually abusive behavior later in life.
Sexual Victimization and Personality
Hunter & Figueredo (2000)	Delineating the relationship between sexual victimization and personality variables in the prediction of patterns of sexual offending against children	The following factors predict subsequent sexual offending in sexually victimized offenders: a younger age at time of sexual victimization, a greater number of incidents, a longer period of waiting to report the abuse and a lower level of perceived family support after revelation of the abuse.
Burton (2008)	Contribution of personality and childhood sexual victimization to a social learning victim-to-victimizer hypothesis for the development of sexually abusive behavior	Sexually abusive youth who had been sexually victimized were likely to repeat what was done to them in regard to the relationship with and gender of their victim(s), modus operandi and sexual behaviors. Suggests that sexually abusive youth may have learned to be sexually abusive from their own sexual perpetrator(s).
Multiple Types of Child Maltreatment
Awad & Saunders (1991)	Compared male adolescents who sexually offended females their age or older to juvenile delinquents and adolescents who engaged in sexually abusive behavior toward younger children	A majority of the adolescents who sexually offended against females their age or older came from a disturbed family background. The rate of sexual victimization for the adolescents who sexually offended against children was much higher and suggested that in some of these adolescents their sexual aggression was a learned behavior, modeled after what they observed at home.
Kobayashi et al. (1995)	Tested a theoretical model of the etiology of deviant sexual aggression by adolescents that included several family factors: perceived parental deviance, child physical and sexual abuse history and a child's bonding to parents	Physical abuse by the father and sexual abuse by males increased sexual aggression by adolescents. Also, a child's bonding to his mother was found to decrease his sexual aggression. The results can be explained from a social learning and a parent-child attachment or social control perspective. Alternative perspectives of evolutionary psychology are also considered.
Cavanaugh, Pimenthal & Prentky (2008)	Co-occurring issues that can often be influential in sexual offending behaviors	Almost all of the youth came from "highly dysfunctional" families and had experienced a high degree of physical, psychological and sexual abuse and neglect. A total of 66.7% had ADHD, 55.6% had PTSD and 49.9% had a mood disorder. Approximately a quarter used drugs and about one-fifth consumed alcohol.
Seto & Lalumiere (2010)	Tested special and general explanations of male adolescent sexual offending	Results did not support the notion that adolescent sexual offending can be explained as a simple manifestation of general antisocial tendencies.
Leibowitz, Burton & Howard (2012)	Compared sexually victimized and nonsexually victimized adolescent sexual abusers with a group of nonsexually victimized delinquent youth on trauma and personality measures	Sexually victimized sexual abusers reported experiencing significantly greater levels of all five types of abuse than the other two groups (emotional abuse, emotional neglect, physical abuse, physical neglect and sexual victimization). General delinquent youth had fewer behavioral and developmental problems than victimized and nonvictimized juveniles who commit sexual offenses.
Wanklyn et al (2012)	Compared three groups of juvenile offenders on childhood risk factors, including individual, family, peer and school domains	Extrafamilial sexual abuse and intrafamilial child maltreatment were found in higher frequency among the juvenile sexual offending groups. Versatile violent offenders may encompass a more severe and high risk subset of the juvenile sexual offending.
Marini, et al (2014)	Relationships among childhood victimization, substance use prior to and force used during a sexual offense	Substance use prior to committing a sexual offense partially accounted for the effects of multiple types of cumulative childhood victimization experiences on the force used during a sexual offense.
Felizzi (2015)	Effects of parent or caregiver instability and disrupted parental attachment on juvenile sex offending	Sexually offending juveniles experienced significantly greater parent or caregiver instability. Supports social learning as a powerful influence and as an important explanation for juvenile sexual offending.
Multiple Types of Child Maltreatment and Personality
Johnson & Knight (2000)	Explored developmental pathways possibly conducive to adolescent sexually abusive behavior, measuring the extent to which the sample experienced childhood trauma, engaged in adolescent delinquency and exhibited particular personality dispositions and cognitive biases	Results suggest that sexual compulsivity and hypermasculinity significantly discriminate verbally and physically coercive sexually abusive juveniles from those who do not report using force. Sexual victimization directly and indirectly (via sexual compulsivity) affected sexual coercion. Alcohol abuse may play a more salient role. Physical abuse had an indirect effect on sexual coercion and was found to be predictive of delinquent behaviors.
Knight & Sims-Knight (2004)	Three-path model intended as a model for sexually abusive behavior toward women	Early traumatic physical and sexual abuse play an important etiological role, increasing the likelihood of sexually abusive behavior either indirectly through the three intervening paths or directly.
Daversa & Knight (2007)	Etiological model for sexual offending behavior toward younger victims	Various developmental and early childhood maltreatment experiences and specific, mediating personality traits contribute significantly to predicting adolescent sexual offending against younger victims. A subset of sexually victimized sexual abusers may select victim(s) specific to a particular age group that is consistent with the age at which they were victimized, indirectly supporting the victim-to-victimizer theory of adolescent sexual offending.
Zakireh, Ronis & Knight (2008)	Individual beliefs and attitudes, and victimization histories of four groups of sexual and nonsexual offenders	Three categories of risk factors — greater hypersexuality or sexual deviance, more violent behavior or fantasies and an increased history of victimization — are consistent with path models that predict sexually abusive behavior toward peers and adults. Findings were consistent with past evidence regarding the role that sexual victimization plays in subsequent sexual offending behavior.
Additional Etiological Factors
Leibowitz, Burton & Howard (2010)	Compared pornography exposure between male adolescents who sexually abuse and male nonsexual offending delinquent youth	The juveniles who had engaged in sexually abusive behavior reported more exposure to pornography when they were both younger and older than age 10 than nonsexual abusers. Exposure was significantly correlated with all of the nonsexual crime scores in the study.
Yoder, Leibowitz & Peterson (2016)	Differences in characteristics of parental and peer attachments between youth who had committed sexual offenses and nonsexual offending youth	Sexual offending youth had higher alienation and lower trust and communication with mothers, and higher alienation and lower trust and overall attachment with fathers. Important in lieu of previous research that associate unmet attachment and socio-emotional needs with sexual offending
Boonman et al (2015)	Conducted a meta-analysis to determine the prevalence of mental disorders in juveniles who sexually offended.	Mental disorders are highly prevalent in juvenile sex offenders with the following percentages: 69% at least one mental disorder; 44% at least two mental disorders; 51% conduct disorder; 18% anxiety disorder (PTSD = 8%); 14% ADHD; 9% affective disorder; 30% at least one substance use disorder; and 42% paraphilia

 

Grabell and Knight (2009) suggest that in addition to having risk factors, it is likely that juveniles who have committed sexual offenses lack protective factors — such as emotional support and social competence — to buffer against risk in early experience. Future research should consider the complex relationships between these risk and protective factors in the development of sexually abusive behavior.

Typologies

Research supports a multifactorial explanatory theory regarding etiological pathways.

Typology research to date has primarily differentiated subtypes of juveniles who have committed sexual offenses based on victim age, delinquent history (differentiating sex-only vs. sex-plus offenders) and personality characteristics. This section focuses on research as it relates to these dimensions.

Subtypes Based on Victim Age

Awad and Saunders (1991) found that the majority of adolescents who sexually offended against females their age or older were recidivists, had a history of antisocial behavior predating and coinciding with their sexual offenses and came from a disturbed family background. Adolescents who sexually assaulted peer or older females were less likely to be socially isolated than those who offended against children and more likely to socialize with older peers than a comparison group of delinquents. Alcohol and drugs were not found to play a prominent role in the adolescents who sexually victimized peer or older females. Sexually deviant impulses and antisocial traits were found to be motivating factors for the majority of these youth.

Hunter and Figueredo (2000) found that juveniles who offended against children were more likely to be pessimistic and less likely to be self-sufficient than nonoffending youth. These findings appear to be consistent with a conceptualization of juveniles who sexually offend against children as youth who are lacking in social competencies and who are competitively disadvantaged relative to their peers. The researchers proposed that the sexual acting out of these juveniles may be more reflective of compensatory behavior than psychopathy and more reflective of arrested sexual development than paraphilic interest.

Hunter, Hazelwood and Slesinger (2000) conducted a study comparing 62 adolescents who offended against children to 64 adolescents who offended against peers and adults. The findings suggest that a meaningful differentiation can be made between youth who sexually offend against younger children (5 or more years younger) and those who target peers and adults. According to the study, adolescents who targeted peers and adults were more likely to have —

• Selected a female victim who was either a stranger or acquaintance;
• Committed their offense in a public area, and acted in a group with others;
• Committed the sex crime in association with other criminal activity and have been more aggressive and violent in commission of the offense; and
• Used a weapon.

Hunter, Hazelwood and Slesinger (2000) suggested that the differences in victim age (peer/adult vs. younger children) represent unique populations of sexually aggressive youth. In general, peer/adult adolescents who commit sexual offenses display behaviors that suggest they have greater antisocial tendencies and are more prone to violence in the commission of their sexual offenses than adolescents who molest children.

In a follow-up study, Hunter and colleagues (2003) contrasted adolescent males who sexually offended against prepubescent children with those who targeted pubescent and postpubescent females. Table 2 identifies the differences found between the two groups.

Table 2. Differentiation Between Categories of Adolescent Males Who Sexually Offend

Those Who Target Prepubescent Children	Those Who Target Pubescent and Postpubescent Females
Greater deficits in psychosocial functioning Use less aggression in their sexual offending More likely to offend against relatives More likely to meet criteria for clinical intervention for depression and anxiety	Employ more force in the commission of their sexual offense More likely to use a weapon and to be under the influence of alcohol or drugs at the time of the offense Less likely to be related to their victim Less likely to commit the offense in the victim's home or in their own residence More likely to have a prior arrest history for a nonsexual crime Demonstrate less anxiety and depression, and less pronounced social self-esteem deficits
Source: Hunter et al. (2003).

 

Deficits in psychosocial functioning were found to affect the influence of childhood exposure to violence against females on adolescent perpetration of sexual and nonsexual offenses. Childhood physical abuse by a father or stepfather and exposure to violence against females were found to be associated with higher levels of comorbid anxiety and depression. Noncoercive childhood sexual victimization by a male nonrelative was found to be associated with sexual offending against a male child.

Knight and Sims-Knight's (2004) three-path model for sexual coercion against women found that juvenile rapists evidenced more antisocial behavior and a higher use of alcohol. Additionally, juvenile rapists were more likely to come from more disturbed families and to have experienced more caregiver instability. The researchers found that these juveniles had committed more violent offenses than offenders who victimized younger children and that they evidenced borderline intellectual functioning more frequently.

Daversa and Knight's (2007, pp. 1326–1327) dimensional model of adolescent sexually abusive behavior against younger children indicates that "different models of developmental antecedents and core traits are involved in adolescent sexually abusive behavior against peer-aged girls or women and younger children and that identifiable paths to offending are evident in each model." The researchers proposed that their results suggest that a typology based on victim age and developmental trajectory is possible. Further, their findings challenged those from prior research that suggested all adolescents who offend against children are submissive, dependent, socially isolated and less aggressive in their sexual offending. Daversa and Knight suggested that a subgroup of adolescent child molesters may be impulsive and aggressive in their offense planning, entertain sadistic fantasies and demonstrate a high degree of sexual arousal toward young children.

Hunter (2009) reported on a study of a national sample of 256 adolescent males who committed sexual offenses and were receiving treatment in an institutional or community-based setting. Initial results suggest the presence of five subgroups and associated characteristics:

• Life Course Persistent — Antisocial
  • Has the highest arrest rate for nonsexual crimes and the highest reported rate of childhood exposure to violence
  • Evidences lengthy childhood histories of exposure to violence and early developmental onset of pornography viewing and drug/alcohol use
• Adolescent Onset — Experimental
  • Is more inclined to sexually offend against peer and adult females
  • Appears less psychosocially and psychosexually disturbed than other subgroups, and reports less childhood exposure to violence and less preadolescent pornography/substance use
  • Appears to have the lowest average number of victims of the five subgroups
• Socially Impaired — Anxious and Depressed
  • Predominantly sexually offends against children
• Pedophilic Interests — Antisocial
• Pedophilic Interests — Non-Antisocial
  • Evidences lengthy childhood histories of exposure to violence and early developmental onset of pornography viewing and drug/alcohol use

Zakireh, Ronis and Knight (2008) found that greater hypersexuality or sexual deviance, more violent behavior or fantasies and an increased history of victimization are consistent with path models that predict sexually abusive behavior toward peers and adults. Additionally, they found that the domains of callousness, unemotionality and antisocial behavior play a significant role in sexually abusive behavior against peers and a limited etiological role in sexually abusive behavior toward children.

Kemper and Kistner (2010) gathered archival data on 296 male adolescents who were committed to a residential high-risk facility for serious and/or chronic offenders aged 12–19. The study examined the strength of the relationship between victim-age-based subgroup membership and personal, criminal history and offense history variables. Consistent with previous research, juveniles who offended against children tended to victimize male and female relatives while peer offenders tended to victimize female acquaintances. Child and mixed-victim-age offenders were more likely to have been victims of sexual abuse. Peer offenders had a more extensive nonsexual delinquent history. Few associations were found between subgroup membership and measures of physical abuse, social skills or impulsivity. Kemper and Kistner (2010) argued that victim age is more likely a proxy for pertinent factors associated with sexual offending and that these include the physical and emotional development of the victim. They proposed that when information related to the victim is used in classification, the combination method of using both victim age and offender-victim age discrepancy is preferable.

Miner and colleagues (2010) explored the relationship between sexual abuse perpetration and insecure attachment and adolescent social development. The researchers compared three samples of 13- to 18-year-old adolescent males: adolescents who committed sexual offenses against child victims, adolescents who committed sexual offenses against peer/adult victims and nonsexual delinquent youth. The results indicated that attachment style had an indirect effect on sexual abuse perpetration. Attachment anxiety affected involvement with peers and interpersonal adequacy. Feelings of interpersonal inadequacy, combined with oversexualization and positive attitudes toward others distinguished adolescents who committed sexual offenses against child victims from nonsexual delinquents and from adolescents who committed sexual offenses against peer/adult victims. Attachment anxiety with a lack of misanthropic attitudes toward others appears to lead to isolation from peers and feelings of interpersonal inadequacy. The researchers proposed that individuals with this constellation of factors may turn to children to meet their exaggerated intimacy and sexual needs. The data suggest that youth who assault peers or adults are not substantially different from other delinquent youth on most measures.

Faniff and Kolko (2012) studied a sample of 176 males adjudicated for a sexual offense who were considered low risk and court ordered to participate in outpatient treatment. Participants were classified into one of three groups based on their victim's age: child victims (at least four years younger than the offender), peer/adult victims or both types of victims (referred to as "mixed"). Regardless of victim type, the researchers found more similarities than differences across the groups in regard to maltreatment experiences, antisocial tendencies, mental health functioning, family functioning and recidivism risk. In contrast to much of the research discussed above, Faniff and Kolko concluded that it is not clear that the selection of a particular type of victim is indicative of unique risks and needs, and suggested that subtyping based on criminal history or personality measures may prove more meaningful. They acknowledged, however, that there is a pattern across studies suggesting greater anxiety and internalizing problems in juveniles with child victims. Similarly, juveniles with peer/adult victims had higher general rearrest rates, consistent with the hypothesis that juveniles who select peer/adult victims are more generally antisocial than those who select child victims. The current study was not able to draw any firm conclusions about mixed offenders given how few were present in the sample.

Miner et al. (2014) followed up on a previous study and further explored the association between insecure attachment to parents, social isolation and interpersonal adequacy, as they relate to adolescent sexual offending against children. The study compared adolescents who committed offenses against children (n = 140) and those who committed offenses against peer or adults (n = 92), with a sample of similarly aged males in treatment for mental health or substance use issues (n = 93). The results of the current study replicated prior Miner et al. (2010) findings that anxious attachment had an indirect effect on child sexual abuse perpetration through social isolation and interpersonal inadequacy. The authors hypothesize, "that when an adolescent experiences isolation from peers, this anxious attachment could contribute to feelings of further isolation and alienation, and is related to feelings of inadequacy, especially in the masculine role… Sexual offending may be an attempt to compensate for this inadequacy" (p. 147). The study found that sexual offending against peer/adults was related to sexual behavioral control, as opposed to interpersonal functioning as it relates to attachment, social involvement or masculine adequacy. Sexual compulsivity, or perceived lack of control over sexual behavior, distinguished both sexual offending groups of adolescents from the adolescent group with mental health or substance abuse problems.

Zeng et al. (2014) investigated if there were differences in the risk and criminogenic needs of 167 Singaporean youth who sexually offended based on two typologies. The study compared the risk and need profiles of a) youth who offended only sexually and youth who offended sexually and nonsexually (see discussion below in Delinquent History) and b) youth who sexually offended against child victims and youth who sexually offended against nonchild victims. The study was retrospective in nature, utilizing clinical file reviews and administering the Youth Level of Service/Case Management Inventory (YLS/CMI) (Hoge & Andrews, 2010) and the Estimated Risk of Adolescent Sexual Offense Recidivism (ERASOR) (Worling & Curwen, 2001) ratings. Regarding sexual reoffending risk factors, a greater number of previous sexual assaults was the only factor that differentiated youth who sexually offended against child victims from youth who offended against nonchild victims. A possible explanation for this difference could be a greater availability of child victims, and a lower risk of a child victim being able to resist an older and stronger perpetrator. Youth who offended against child victims did not differ significantly from youth who offended against nonchild victims in terms of level of general criminogenic needs.

Leroux et al. (2016) examined a Canadian sample (n = 162) comparing adolescents who had sexually offended against children, adolescents who have offended against peers or adults and adolescents who had victims in both age groups. Participants were compared on several factors including childhood sexual abuse, atypical sexual interests, sexual experience, social competence, psychiatric history and general delinquency. The authors concluded that the study results provide further support for the validity of distinguishing adolescents who commit sexual offenses by victim age, as synthesized in the following table:

Adolescents with child victims

Adolescents with peer or adult victims

Less sexually experienced, which suggests they are more likely to offend against children secondary to lacking the sexual and social maturity to form intimate relationships with peers Most sexually atypical Lower on measures of general delinquency Results "partially support the hypothesis that adolescent sexual offending against children is better explained by special explanation factors rather than general delinquency factors" (p. 91)

More likely to have inflicted more physical harm on their victims More likely to have been under influence of substances at time of offense Most severe behavioral problems include higher rates of general substance use, prior conduct disorder, and disruptive behavior disorder at time of assessment Results more consistent with the general delinquency explanation

 

The results suggest that adolescents with mixed victim age (both child and peer/adult victims) constitute a distinct group of offenders and that both general delinquency and special explanation factors contribute to the onset of the mixed victim age type of sexual offending. It further implies that this subtype of juvenile who has committed a sexual offense has the most treatment needs.

Joyal, Carpentier and Martin (2016) assessed the value of subgrouping juveniles who have committed sexual offenses based on victim age and secondarily, based on sibling incest. Retrospective data was collected on 351 adolescent males from the archives of a legal forensic center in Montreal. The results of the study "further confirms the usefulness of distinguishing [juvenile sexual offenders] of children versus [juvenile sexual offenders] of peers/adults" (p. 18).

JSOs with child victims	JSOs with peer/adult victims
Socially isolated or rejected by their peers Low social competence Atypical sexual interests	Present antisocial profile: diverse criminality, criminal activity with peers, using drugs and alcohol and conduct disorder diagnosis Have more peer-aged friends Previous consensual sex with peer-age teen Targeted an unknown victim

 

Sibling incest juvenile sex offenders were more likely to have been sexually victimized during their own childhood, suggesting that victimization is more strongly related to family relation than victim age. Compared to juvenile sex offenders with extrafamilial child victims, sibling incest juvenile sex offenders were also less likely to have much younger friends and to be socially isolated or rejected by peers and more likely to have received a conduct disorder diagnosis. The authors argued that these findings indicate that sibling incest offenders have more varied and complex treatment needs, stating that the overall findings of the study "stress the importance of considering subtypes of juvenile sex offending in treatment plan elaboration" (p. 20).

Subtypes Based on Delinquent History

Butler and Seto (2002) sought to distinguish between adolescents who sexually offend based on nonsexual offense history. Based on their criminal records, 114 male adolescent offenders were divided into three groups: adolescents who commit sexual offenses, versatile offenders and nonaggressive offenders. The adolescents who committed sexual offenses were further classified as "sex only" if they had only been charged with sex offenses or as "sex plus" if they had ever been charged with a nonsexual offense. Youth were compared on measures of childhood conduct problems, current behavioral adjustment, antisocial attitudes and beliefs and risk for future delinquency. The researchers found that sex-only adolescents who committed sexual offenses had significantly fewer childhood conduct problems, better current adjustment, more prosocial attitudes and a lower risk for future delinquency than did the adolescents who committed nonsexual offenses. Sex-plus adolescents resembled criminally versatile offenders. Butler and Seto concluded that differences between sex-only and sex-plus adolescents who committed sexual offenses reflect a valid typological distinction and that this discrimination has implications for differential intervention. Sex-plus adolescents are at higher risk for general reoffending than are sex-only adolescents and are more likely to benefit from treatment targeting general delinquency factors. They may also be more likely to require multisystemic interventions that simultaneously address individual, family and social influences on antisocial behavior.

Zakireh, Ronis and Knight (2008) found that juveniles who have committed sexual offenses may share a number of common difficulties with general delinquents because many of these youth have similar patterns of criminal offending. Thus, sexual offending may be part of a broader pattern of serious antisocial behavior for a proportion of sexual offending juveniles.

As a follow-up to the above-noted Butler and Seto (2002), Pullman et al. (2014) compared sex-only (n = 71) and sex-plus (n = 87) sexually abusive adolescents on a range of antisocial and atypical sexual interest variables and variables associated with specialist models of sexual offending. The sex-plus adolescents had to have a sexual and nonsexual delinquent history and/or a history of moderate to severe conduct disorder before age 10. The sex-plus adolescents were further differentiated between early onset (moderate to severe conduct problems before age 10) and late onset. A synthesis of the findings differentiated the two groups as follows:

Sex-Only Offenders	Sex-Plus Offenders
More atypical sexual interests Greater deficits in romantic relationships More likely to have a child victim More likely to have a male victim	More likely to be diagnosed with antisocial personality traits More likely to use drugs Exhibit more psychiatric issues Greater deficits in general social skills Caused more physical injury to victim and more likely to use weapon More likely to have familial psychiatric and substance abuse history More likely to be living in single-parent home More likely to have been physically abused

 

Little support was found for the use of the early vs. late-onset classification scheme. The overall results provide further support for the validity of a sex-only vs. sex-plus distinction, as well as supporting the generalist/specialist literature as it relates to etiology. Generalists are more driven by general antisocial factors, which suggests that treatment focused on delinquency factors will be more effective. Sex-only adolescents appear to be more driven by specialist explanations and, as such, should receive specialized interventions, including addressing difficulty with romantic partnerships and atypical sexual interests.

McCuish, Lussier and Corrado (2015) utilized Loeber and Hay's (1994) three pathway model in assessing differences in antisocial pathways within juveniles who had committed sexual offenses and juveniles who had committed nonsexual delinquency. The final sample consisted of 51 juvenile sexual offenders and 94 juvenile nonsexual offenders incarcerated in open and secure custody facilities in British Columbia, Canada. The results indicated the presence of multiple antisocial behavioral pathways in juvenile sexual offenders that reflect those of nonsexual offenders. A Low Antisocial group was the most prevalent pathway, representing about 50 percent for both juvenile sexual and nonsexual offenders. The other half of the sample of both juvenile sexual and nonsexual offenders was almost evenly split between the Overt and Covert groups, and can be distinguished based on their pattern of predominantly covert or overt antisocial behavior.

Juvenile Sexual Offenders — Overt Group	Juvenile Sexual Offenders — Covert Group
May have engaged in sexual offending as part of an escalating and long-standing (before age 12) pattern of aggressive and violent behavior Sexual victims were more likely to be aware of their victimizer because the overt confrontational nature	Types of behaviors require actively avoiding detection Sexual offense may have reflected well-concealed and deceitful pattern that possibly included identifying vulnerable victims who were less likely to report (e.g., a mentally disabled child, children under their supervision while babysitting or severely intoxicated individuals)

 

The authors posit that the findings support the need for more precise assessments of the antisocial behavior profile in order to determine whether there is a link between the nature of the antisocial behavior pathway and the modus operandi of a juvenile sexual offender's offense. This should include assessment of criminogenic factors and risk for general recidivism, which research suggests is the most likely outcome of juvenile sexual offenders who reoffend.

As discussed above, Zeng et al. (2015) investigated if there were differences in the risk and criminogenic needs of 167 Singaporean youth who sexually offended based on two typologies, one of which was comparing youth who offended both sexually and nonsexually to youth who offended only sexually. Youth who offended both sexually and nonsexually were found to have higher risk and criminogenic needs as compared to youth who engaged in sexual offending only. This suggests that criminally diverse youth who sexually offended may be influenced by general criminogenic risk and needs factors. The study results also are consistent with prior research findings that youth who offended sexually and nonsexually bear similarities to nonsexual delinquent youth in their group characteristics. The authors suggested that antisocial interpersonal orientation, a lack of intimate peer relationships, interpersonal aggression, and problematic parent–child relationships represent particularly salient risk factors and needs for youth who offended sexually and nonsexually.

Subtypes Based on Victim Age and Delinquent History

Aebi and colleagues (2012) sampled 223 male children and adolescents aged 10–18 who had been convicted of sexual assaults against children, sexual assaults against adolescents and adults, coercive sexual behavior, exhibitionism and sexual harassment in Zurich, Switzerland. The research tested the validity of typing sexually abusive juveniles based on victim age, co-offender status and crime history. The best evidence was found for the victim-age-based subtype that differentiated juveniles who offended against children from those who had offended against adolescents and adults. Consistent with findings from previous research, Aebi and his colleagues found that offenders against children were younger at the time of offense, less likely to be of foreign nationality, more likely to have male victims and showed less aggression in their offenses. The researchers also found that sexual offense severity was higher among child offenders and included more intrusive behaviors relative to adolescents who had offended against adolescents and adults. The consideration of a distinct underlying psychological mechanism differentiating offenders based on victim age may be important for intervention planning.

Although there was some support for juveniles who offend against children as a separate type, Aebi and colleagues (2012) concluded that the limited validity and lack of independence found for the three types strongly suggest that a comprehensive typology is not feasible. The researchers suggested that a dimensional approach based on the following factors is more adequate for describing juveniles who have committed sexual offenses:

• Single offender with severe molestation of a related child
• Persistent general delinquent with migrant background
• Older offender with alcohol use and familial constraints
• Multiple and aggressive offender with social adversities
• Offender with unselected and multiple victims

Aebi and colleagues (2012, p. 283) concluded that these findings suggest "distinct dimensions of criminality implying different pathways that lead to sexual offending in youth" and proposed that the identification of criminality dimensions in terms of relevant patterns of sexual offending characteristics may be more useful in guiding treatment intervention.

Subtypes Based on Victim Age and Personality Characteristic

Carpenter, Peed and Eastman (1995) compared the personality characteristics of adolescents who committed sexual offenses by examining the extent (if any) to which personality differences exist between adolescents who offend against their peers and adolescents who offend against younger children. The sample consisted of 36 adolescents who committed sexual offenses (16 peer offenders and 20 child offenders) and who were committed to Virginia's Department of Youth and Family Services. The researchers found that adolescents who molested children are more schizoid, avoidant and dependent than adolescents who offended against peers. The adolescents who offended against children frequently demonstrated a pattern of withdrawing from social encounters with peers and, as such, commonly experienced loneliness and isolation. In discussing these findings, Carpenter and his colleagues (1995, p. 196) stated that these results "may help explain why adolescent sexual offenders against children gravitate to their victims." Results also suggest that adolescents who offend against peers have an inflated self-image and are arrogant and interpersonally exploitative. Evidence suggests that the design and effectiveness of treatment programs may be enhanced if the personality differences between the type of victim (child or peer) can be taken into account.

Worling (2001) examined the California Psychological Inventory Scores from 112 males aged 12–19 who committed sexual offenses. A cluster analysis of the factor-derived scores revealed four personality-based subtypes and associated characteristics:

• Antisocial/impulsive youth are likely to have a propensity for rule violations. Their sexual offending, at least initially, is more a result of this factor than deviant sexual arousal. Descriptors of this subgroup may include anxious, unhappy and rebellious.
• Unusual/isolated youth are emotionally disturbed and insecure. They are characterized by a peculiar presentation and social isolation. Their awkward personality features may inhibit their ability to develop and maintain healthy and intimate relationships with consenting peers.
• Overcontrolled/reserved youth endorse prosocial attitudes, are cautious to interact with others and tend to keep their feelings to themselves.
• Confident/aggressive youth are confident, self-centered, outgoing, aggressive and sociable.

Significant differences were observed among the groups regarding history of physical abuse, parental marital status, residence of the juveniles and whether or not they received criminal charges for their initial sexual assaults. Membership in the subgroups was unrelated to victim age, victim gender and the juvenile's history of sexual victimization. The juveniles in the two most pathological groups — antisocial/impulsive and unusual/isolated — were most likely to be charged with a subsequent violent (sexual or nonsexual) or nonviolent offense. Worling reported that 39 percent of his sample was sexually victimized, with no between-group differences being found. Twice as many juveniles in the antisocial/impulsive group had a history of physical victimization compared to the other groups in the study. Worling asserted that these results provided evidence for heterogeneity in the presence and nature of psychopathology, personality characteristics and social functioning in adolescents who commit sexual offenses and of different etiological pathways and treatment needs.

Richardson and colleagues (2004) described a personality-based taxonomy based on an outpatient sample of 112 sexually abusive adolescents. Five subgroups were derived from cluster analytic procedures applied to personality pattern scales scores from the Millon Adolescent Clinical Inventory (MACI): normal, antisocial, submissive, dysthymic/inhibited and dysthymic/negativistic. The groups were also found to be differentiated on MACI's psychopathology scales, with mood disorders indicated in three of the five subgroups. The study results provide evidence of the heterogeneity of adolescents who sexually abuse in both personality characteristics and psychopathology. A comparison of the groups differentiated on the basis of victim characteristics did not indicate a relationship between personality and sexual offense. The broad clinical distinction between internalizing and externalizing disorders was found to be valid in this sample. The researchers suggested that it may be better clinical practice to facilitate treatment planning that is based on subgroup membership, as evidenced by personality type and clinical presentation.

Summary

The dearth of research on juveniles who have committed sexual offenses in the 1980s resulted in a "trickle-down" approach, in which an adult sexual offender model was used that supported a narrow and specialized one-size-fits-all treatment and supervision approach. The national experts who participated in the SOMAPI forum identified the importance of individualizing treatment for juveniles who commit sexual offenses. In addition, research has increasingly demonstrated the heterogeneity of the population of juveniles who have committed sexual offenses in regard to etiological pathways, offending patterns, delinquent history, personality characteristics and clinical presentation, and risk for sexual and nonsexual recidivism. The integration of findings from etiological and typology studies suggests differential risk and treatment and supervision needs.

The importance of using individualized treatment and supervision strategies was also acknowledged by the experts at the SOMAPI forum. Related to typologies, Knight and Prentky (1993, p. 77) provide a balanced assessment of the use of clinical labels:

Clinical labels have some negative consequences. If however, we refrain from applying labels because of fears about the possible negative consequences of the misapplication of such labels, we would also forfeit our chances of discerning causes, of designing intervention programs that address the more specific needs of subgroups, of identifying vulnerable individuals who might profit from primary prevention programs, and of improving our dispositional decisions about specific subgroups of offenders…. Thus, categorization yields multiple advantages, and must be pursued. We must also remain cognizant of the limits of our taxonomic models and continually challenge our constructs and scrutinize the empirical validity of the measures and types we generate.

Typology research dealing with juveniles who commit sexual offenses has focused primarily on the subtyping of juvenile offenders based on victim age, delinquent history and personality characteristics. This research has yielded substantial insights in regard to identifying differential etiological paths, typological characteristics and associated treatment targets. Although there is a great deal of diversity in juvenile sexual offenders, there is increasing research support for both victim-age based and sex-only vs. sex-plus theories for differentiating subtypes of juvenile sexual offenders. There appears to be a convergence of research that delineates subtypes differentiated by the presence of a nonsexual delinquent history and a general delinquency orientation (generalists) versus juveniles whose sexual offending is associated with issues related to anxious attachment, social and sexual immaturity and inadequacy, and social isolation (specialists). A small subset of the specialists manifests atypical sexual interests. Given the large number of potential influences and interactions of sexual offending characteristics, Aebi and colleagues (2012) argue that juvenile sexual offenders may be better described by using dimensional measures rather than assigning them to specific types. The research to date has provided very useful information regarding dimensions that include trauma and chaotic family environments, attachment, psychosocial adjustment, delinquent history and orientation, co-occurring mental health problems, sexual drive and preoccupation, and atypical sexual interests.

The evolving knowledge on etiological pathways and typologies is increasingly informing intervention practices, particularly the ability of sex offender management professionals to design intervention programs that address the specific risk and needs of subgroups of juveniles who commit sexual offenses. Empirical evidence concerning the prevalence of child maltreatment in early development offers support for continuing treatment aimed at victimization and trauma resolution with sexually abusive youth. Developmental models, which have included early childhood experiences and family functioning, should be broadened to include larger social variables such as exposure to sexually violent media and characteristics of social ecologies.

While research has documented the heterogeneity and differential treatment and supervision needs that exist within the juvenile offender population, policy responses tend to be designed with only the highest risk offenders in mind. Rather than using a one-size-fits-all approach, legislative initiatives should encourage risk assessment and the application of aggressive strategies and the most intensive interventions only for those offenders who require the greatest level of supervision, treatment and personal restriction. In this way, both community safety and the successful rehabilitation of youth who offend can be ensured.

 

Notes

1 Primary prevention approaches occur before sexual violence to stop initial victimization; tertiary prevention approaches occur after sexual victimization to address the consequences to the victim as well as the management of known sex offenders to minimize the possibility of reoffense (Association for the Treatment of Sexual Abusers, 2013).