A Comparison of Incest Offenders Based on Victim Age
http://www.100md.com
《美国精神病学与法律学报》
Dr. Firestone is Professor, School of Psychology and Department of Psychiatry, and Mr. Dixon was a student in Psychology, University of Ottawa, Ottawa, Ontario, Canada. Dr. Nunes is Researcher, Programming Research Division, the Correctional Service of Canada, Ottawa, Ontario, Canada. Dr. Bradford is Director, Forensic Service, Royal Ottawa Hospital, Ottawa, Ontario, Canada.
Abstract
The purpose of the present study was to compare incest offenders (IOs) whose victims include infants or toddlers to IOs with adolescent victims on several variables commonly examined in the sexual offender literature. Participants were 48 men whose youngest victim was less than 6 years of age (younger-victim incest offenders; YVs); and 71 men whose youngest victim was 12 to 16 years of age (older-victim incest offenders (OVs). In general, YVs showed more emotional disturbance and pathology than OVs. Compared with OVs, YVs had a greater history of substance abuse and more current problems with alcohol. In addition, YVs reported significantly poorer sexual functioning and were significantly more psychiatrically disturbed. YVs were also more likely to have a male victim, to have victimized a nephew/niece or grandson/granddaughter, and to have denied their offense(s). It was evident that both the YVs and OVs demonstrated clinically significant difficulty with normal sexual functioning and exhibited deviant sexual arousal.
While sexual victimization of any age group is unacceptable, the current admonitions surrounding such offenses against adolescents could be viewed as largely attributable to societal/cultural norms and mores and to investigations regarding the physical and psychological damage endured by a young person who has been abused. That is, in an effort to protect young people, most modern societies have designated those under the age of roughly 16 as psychologically unprepared for sexual interactions (particularly with adults), irrespective of physical traits or reproductive competencies present that may suggest otherwise. For research and management purposes—if not in everyday parlance—it is primarily this arbitrary victim age demarcation that has formed the basis for labeling a sexual offender as a child molester. However, perhaps under the blanket designation of child molester there are important subgroups based on the victim’s age. Sexual abuse of an infant or toddler not only represents the same breach of society’s values and laws that a similar crime against an adolescent does, but it also contravenes any semblance of adaptive sexual behavior, biologically or otherwise. With this difference in mind, it is reasonable to predict that a person who offends against a very young child would differ, on one or many levels, from a person whose sexual assault victims appear to be limited to adolescents. Supporting this line of thought, Kalichman1 has reported that child molesters who abuse prepubescent children demonstrate significantly more psychopathology and emotional disturbance than those with older victims.
Although incest offenders are often treated in the literature as a discrete category, it has been noted that no universal characteristics or single profiles of such offenders have been found.2 Relatively common characteristics include passive personalities, dependent personalities, physical and emotional maltreatment during childhood, marital dissatisfaction, sexual dissatisfaction, and disturbances in empathy and attachment.2 Studer et al.3 observe that because their offending is limited to family members, it is commonly held that incest offenders pose a limited threat to society as a whole, and as a result of this perception, they are often given less intensive treatment. This characterization may be partially sourced from the results of comparisons between incestuous and nonincestuous child molesters that suggest that the arousal patterns of incest offenders are less sexually deviant.4,5 However, some studies have failed to uncover differences in this regard.6,7 We contend that the inconsistency in sexual preference findings in incest offenders and the difficulties experienced by researchers in developing a reliable profile of this population stem from a failure to attend to a possible heterogeneity within this category, based on the victim’s age.
Sexual abuse of infants and toddlers is not uncommon, with cases involving victims as young as two to three months having been reported.8,9 Accurate estimates of the number of very young children who are sexually abused are difficult to obtain, as the limited verbal and abstraction skills of the victims, along with the damaging effects of the abuse itself, present significant impediments to full or even reasonable disclosure.10 Finkelhor11 found that of 270 cases of sexual abuse of children in daycare, 6 percent occurred against children younger than two years and 60 percent involved children less than four years of age. Schetky12 surveyed Child Protection Services serving three small counties in Maine and found 53 substantiated cases of sexual abuse involving children younger than four years between the years of 1985 and 1988. According to Schetky, the perpetrators of sexual abuse against infants and toddlers tend to be family members, babysitters, or daycare workers who capitalize on the access to defenseless victims furnished by their caregiving role, an opportunity not available to a stranger.
Although the authors are unaware of any published research in which incest offenders are compared as a function of victim age, some comparisons of extrafamilial child molesters have been conducted. Greenberg et al.,10 in a comparison of extrafamilial child molesters who had offended against victims younger than five years with those who had offended against victims aged 8 to 12 years, found few differences on measures of victim characteristics, offense characteristics, or psychological characteristics of the offender. However, offenders who victimized younger children were significantly younger than those who victimized older children. It is possible that the failure to find further statistically significant differences was due to the small sample size. A follow-up study was conducted by Greenberg et al.,13 and, as in the previous study, the offenders against younger children were observed to have features similar to those whose offenses were against older children, with the exception, once again, of age. In addition, offenders against young victims were significantly more likely to deny their offenses and to have a higher score on Factor 2 (antisocial lifestyle) of the Psychopathy Checklist-Revised (PCL-R14). Although this study involved 192 offenders against older children, there were only 18 whose victims were younger children, which may have again failed to provide adequate statistical power to detect real differences that may have existed between the two groups. In light of Schetky’s12 observation of the preponderance of family members involved in cases of infant and toddler sexual abuse and given the importance of frequent and isolated access to the victim in this form of sexual offending, it could be expected that this phenomenon is more likely to be witnessed in incest offenders. The studies by Greenberg and his colleagues13 involved samples of extrafamilial child molesters only. The exclusion of incest offenders may have contributed to the small sample sizes of the groups with victims younger than five years.
In the current study, we investigated whether incest offenders who have offended against very young children are distinguishable from those whose youngest victims were pubertal (as far as criminal and psychiatric files indicate). There is some similarity in the current study to the methodology and focus of the earlier studies of sexual abusers of infants and toddlers by Greenberg et al.10,13 Greenberg and his colleagues are also affiliated with the Sexual Behavior Clinic at the Royal Ottawa Hospital and used data from a database similar to the one we have used. However, Greenberg et al. studied extrafamilial child molesters, while our investigation focused on a sample of incest offenders, and so there is no overlap in participants. To our knowledge, there have been no other studies of child molesters who have offended against children under the age of five in such a population of offenders. A potential limitation of the studies by Greenberg et al. concerns the fact that the group with victims 5 years of age or younger was compared with a group with victims as old as 16 but as young as 6 years. This may have masked differences that existed between offenders in the two groups whose victims’ ages were more dissimilar. With this in mind, another key difference in the protocol of this study is the use of a comparison group consisting of men who have exclusively offended against post-pubertal victims in their index offenses. This decision was made to maximize the contrast between the ages, and thus the physical and emotional characteristics of the victims of the two groups of offenders. Twelve years was chosen as the lower limit for the group with older victims based on the fact that the average age for the onset of menses in girls in the United States is 12.5 years,15 and the average age at which pubic hair develops in boys is 13.16
It is our contention that a man who sexually abuses an adolescent who presumably displays at least some semblance of sexual maturity, while no less criminal in his actions, presents a less disturbed person than one who offends similarly against society’s most vulnerable and least sexually provocative demographic: infants and toddlers. The current study is preliminary and descriptive in nature, and the existing literature does not lend itself to many specific hypotheses, due to the current study’s differing population (i.e., incest offenders) and methodology. However, several differences are expected to emerge and to be congruent with the central prediction that the group with victims younger than six years of age are more generally disordered and pathological. The groups were compared on demographic and historical variables, psychological test scores, phallometric indices, index offense and victim characteristics, and criminal histories. Preliminary comparisons of this nature can contribute to the elucidation of the traits of those in the incest offender category. Furthermore, evidence supporting heterogeneity within this population may hold significance in treatment and management of these offenders.
Method
Participants
The present sample consisted of two groups of adult men convicted of hands-on sexual offenses against family members (biological child, stepchild, nephew, niece, grandchild, or sibling) who were 16 years old or younger. All participants were assessed at the Royal Ottawa Hospital, Sexual Behaviors Clinic. In one group were 48 men whose youngest victims were less than six years of age (younger-victim incest offenders; YVs). In the other group were 71 men whose youngest victims were at least 12 years of age (older-victim incest offenders; OVs). Ideally, the YV group would have consisted of men whose oldest victims were also less than six years of age. However, this degree of purity could not be furnished by the sample. Signed informed consent was obtained from all participants. The Research Ethics Committee of the Royal Ottawa Hospital approved the research protocol. Although the offenders’ names appeared on both criminal and psychiatric records, confidentiality was maintained by numerically labeling each case in the computer database.
Measures
Sexual Functioning
The Derogatis Sexual Functioning Inventory (DSFI) is designed to assess general and specific dimensions of sexual functioning.17,18 It collects information by using numerous items at once to grasp "the fundamental components judged essential to effective sexual behavior" (Ref. 18, p 117). The 10 subscales are Information, Experience, Sexual Drive, Sexual Attitude, Psychological Symptoms (also known apart from the DSFI as the Brief Symptom Inventory), Affect, Gender Role Definition, Sexual Fantasy, Body Image, and Satisfaction. The Sexual Functioning Index (SFI) is a global measure derived by adding the 10 subtest scores. It provides an overall measure of an individual’s level of sexual functioning. The DSFI has been used with large nonforensic samples, but its use with sexual offenders is limited. There is some suggestion that sexual offenders show high levels of sexual dissatisfaction.19
Psychiatric Disturbance
The Brief Psychiatric Rating Scale (BPRS)20 is an inventory designed to provide clinician ratings on psychiatric symptomatology in 16 domains, such as anxiety, emotional withdrawal, somatic concerns, and other symptomatology. For each domain, the clinician is required to rate the degree to which symptomatology is present in the subject. Ratings may range from a low score of zero (not present) to a high of six (extremely severe). In addition to providing information about specific symptomatology, a total score provides an overall indication of psychiatric disturbance. An acceptable internal consistency coefficient of .63 has been reported.21
Alcohol Abuse
The Michigan Alcoholism Screening Test (MAST), a 24-item self report inventory, is used to identify behavior indicative of alcohol abuse.22 The internal consistency is adequate, with a reported overall coefficient of .87, and the test is relatively unaffected by age or by denial of socially unacceptable characteristics.23,24 Scores of five or six are considered suggestive of alcohol problems, and a score of seven or more is considered strongly indicative of alcohol abuse.25 The MAST has been found to correlate with DSM-III-R criteria for alcohol dependence.24 The MAST has been extensively used as a screening tool for alcoholism, and many studies have used samples of sex offenders.25–28
Psychopathy
The Psychopathy Checklist-Revised consists of 20 clinical rating scales designed to assess behavior and personality characteristics considered fundamental to psychopathy.14 Rigorous testing has indicated that the PCL-R is a psychometrically sound instrument.29 Valid PCL-R ratings can be made on the basis of high-quality archival information.30,31 The existence of two factors has been replicated using various samples: (1) the degree of personality, interpersonal, and affective traits deemed significant to the construct of psychopathy and (2) the degree of antisocial behavior and an unstable, corrupted lifestyle.32 Previous studies have shown the interrater reliability and internal consistency of both factors to be high, despite the small number of items per factor.14,32 In the present investigation, the PCL-R was completed by two research assistants who examined descriptive material contained in institutional files. These files included criminal records, court reports, and psychiatric records. A random sample of 100 clinic files was independently rated by each researcher, resulting in satisfactory interrater reliability (r = 0.88; p < .0001).
Measurement of Sexual Arousal
Changes in penile circumference in response to audio stimuli were measured by means of an indium-gallium strain gauge.
Stimulus Presentation.
The order of stimulus presentation, held constant for all participants, was computer controlled. Participants were presented with one or more of three series of audiotapes, according to the nature of the participants’ sexual offense. The audiotapes consisted of 120-second vignettes that described activities varying in age and gender of the victim or partner, and degree of violence portrayed.33 After being instructed to allow normal arousal to occur, each participant was presented with at least one full set containing one vignette from each category. The categories used in the assessment of arousal in response to children that are relevant to the current study were: (1) a child initiating sexual contact, (2) the child participating mutually in sexual activity, (3) nonphysical coercion of a child, (4) physical coercion of a child, (5) sadistic sex with a child, (6) nonsexual assault of a child, and (7) consenting sex with an adult. The audiotape series used to identify sexual attraction to rape included two scenarios of two minute’s duration in each of three categories: (1) consenting sex with an adult female, (2) the rape of a female adult, and (3) the nonsexual assault of a female adult.
Scoring.
The Pedophile Index (PI) was computed by dividing the highest response to the child initiates or child participates mutually stimulus by the highest responses to a consenting adult stimulus. Adequate criterion validity for the PI (utilizing the same stimuli as the present study) has been demonstrated by its ability to differentiate child molesters and non-offenders significantly at levels of accuracy above chance.34
The Pedophile Assault Index (PAI) was computed by dividing the highest response to an assault stimulus involving a child victim (nonphysical coercion, physical coercion, sadistic sex, or nonsexual assault) by the highest response to the child initiates or child participates mutually stimulus. Adequate criterion validity for the PAI (utilizing the same stimuli as the present study) has been demonstrated in past studies that found significant differences between homicidal sex offenders and non-offenders.35,36
Procedure
Most of the data examined in the current study, such as demographic, historical, psychological test, phallometric, and index sexual offense information, were derived from the medical records or the participants’ assessments, which routinely involved numerous interviews and extensive testing. After the assessments, data on prior criminal charges were gathered from the Canadian Police Information Center (CPIC), which has compiled a national database of criminal arrests and convictions, including INTERPOL reports from the Royal Canadian Mounted Police.
Results
Before statistical tests were performed, the data were screened to ensure that the assumptions underlying the tests were not violated. In the very limited number of cases in which outlying values occurred, they were adjusted upward or downward according to the direction of the problem. This method is appropriate when case retention is desirable and when it does not unduly influence the group means.37 Student’s t tests were performed on continuous variables with = .05; two-tailed and dichotomous variables were analyzed by chi-square, cross-tabular, independent group tests with = .05.
Demographic and historical data for YVs and OVs are presented in Table 1. The groups did not differ in age or number of years of education. However, a statistical trend (p < .10) suggested that the YVs had lower full-scale IQ scores than the OVs. The YV group was also significantly more likely to have a history of drug abuse and alcohol dependency. In terms of family background, the group with younger victims was significantly more likely to have been raised in a family with a history of criminality, but there were no other significant differences.
The two groups of incest offenders were compared on a variety of psychological and phallometric measures (Table 2). The YVs had poorer overall sexual functioning than the OVs, as evidenced by significantly lower group mean scores on the DSFI. It is interesting to note that neither group scored above the fifth percentile on this sexual functioning inventory. In addition, scores on the MAST indicated that YVs are significantly more likely than OVs to be alcohol dependent; was adjusted for this comparison, because the assumption of homogeneity of variance was not met. Offenders against younger children also exhibited significantly more psychiatric disturbance than the other group on the BPRS. The PCL-R did not distinguish between the YV and OV groups. As indicated at the bottom of Table 2, phallometric analysis revealed that both groups showed clinically meaningful levels of deviant sexual preferences. Nevertheless, there was no significant difference between the YV and OV participants.
Nonparametric analysis of data regarding the offender’s sexual offenses and the victim’s characteristics revealed that YVs were more likely to have multiple victims (Table 3). This variable was calculated with dichotomous data (i.e., one victim versus more than one victim). Examination of the frequencies, however, revealed that the majority of YVs (87.5%) had two or fewer victims, while the remaining 12.5 percent had between three and seven victims. In other words, it may be more appropriate to consider YVs more likely to have two victims than OVs.
The procedure by which the YV and OV groups were constructed in the current study created a potential threat to internal validity. Specifically, the inclusion criteria for the YV and OV groups could have been responsible for the higher victim counts of the YVs and, consequently, for the other differences that were found between the groups. The YVs could have any number of victims older than five years (but <16 years), so long as they had at least one victim aged five years or younger. Meanwhile, for the sake of maintaining purity in the OV group, an individual who offended against a 13-year-old but also had a 7-year-old victim would not be included. To determine whether the exclusion of offenders from the OV group, because they also had victims between 6 and 11 years old, artificially lowered the number of victims in the OV group, the medians for each group were examined. Both those included in and excluded from the OV group had a median of one victim. This suggests that the stringent criteria for the OV group did not unduly affect the results on this variable.
The YVs were more likely to have offended against a male victim than the OVs. The victim’s relationship to the offender was derived from a categorical variable that listed the following options: stepson/stepdaughter, biological son/daughter, sibling, and relative (grandson/granddaughter, niece/nephew). The only significant difference between YVs and OVs concerning the victims’ relationships to the offenders was that of relative. All other victim relationships to the offender were equally likely in both groups. Another way of describing this finding is that YVs were more likely than OVs to have been the grandfathers or uncles of their victims. It was also evident that the YVs were more likely than OVs to deny having committed their offenses. Finally, there were no significant differences in criminal history between the two groups (Table 3).
Discussion
In general, the results of the current study provide support for the notion that incest offenders with prepubescent victims are a group with more disordered personalities than men who have sexually assaulted only pubescent or postpubescent children. With one exception, whether statistically significant or not, those men convicted of sexual offenses against younger children scored more poorly on nearly all measures studied, with the only exception being previous sexual offenses.
Demographic and Historical Information
Unlike the previous findings of Greenberg et al.13,10 in extrafamilial child molesters, the group with the younger victims in this study did not differ in age from the group with older victims. However, the average age of all participants in this study (mean (M) = 40.1 years) is considerably older than the ages of participants in the report by Greenberg and his colleagues (M = 27.9 years). The difference in age may be due to the participants in this study being incest offenders. Such offenders are usually older than extrafamilial child molesters.38 The two groups of men did not differ in the number of years of education completed. There was a statistical trend that emerged in the data suggesting that sexual aggressors against prepubescents have lower intelligence than sexual aggressors against adolescents, as measured by IQ scores. The rates of marriage in the histories of the two offender groups did not differ.
Williams and Finkelhor2 have noted that many incest offenders rationalize their crimes by blaming their offending behavior on alcohol. In the present study, there were no differences between the two groups regarding whether they reported being influenced by drugs or alcohol during the commission of the offense. Nevertheless, sexual offenders with young victims were more likely to have reported drug and alcohol abuse, than were offenders against older victims. The offenders with younger victims scored well beyond the cutoff of seven on the MAST, which is strongly indicative of alcohol abuse.25 The effects of recreational drugs or alcohol may augment an offender’s libidinal drive, distorting his perceptions or lowering his inhibitions, any number of which may lead to impaired judgment regarding sexual interactions with a very young member of his family. It is also possible that the sexual offenses against prepubertal children are part of the paraphilic attraction that these men have. It was certainly evident that both groups of offenders in the present investigation experienced clinically meaningful levels of deviant sexual arousal in response to children.
In the current investigation there was not a great deal of evidence to suggest that offenders with younger victims were more likely to have been raised in families with serious problems. It is worth noting that roughly 40 percent of all participants in this study reported that they were sexually and physically abused themselves, a proportion that is slightly higher than the one-third of pedophiles who reported being sexually victimized during their childhoods in other research.39 Although intuitively appealing, the link between commission of sexual crimes and offenders’ own experiences of sexual abuse is a contentious one. Hanson and Slater39 found that the proportions of sex offenders against children who reported being sexually abused themselves were similar to those found in other sexual and nonsexual offender groups.
Psychological Measures
The results of the current study agree with the observations of Kalichman1 that child molesters who abuse prepubescent children demonstrate significantly more psychopathology and emotional disturbance than do offenders with older victims. The group of men with younger victims was clearly more apt to abuse alcohol than the group with older child victims. In addition, on the DSFI, a general measure of sexual functioning, sexual aggressors against prepubescent children rated themselves as having significantly more problems than the other group. While this observation lends support to the hypothesis that those who offend against younger children are likely to be more disturbed than those who do not, it should be noted that the mean score on the DSFI for both groups placed them in approximately the fifth to seventh percentile of the population at large.40 This suggests that there is difficulty with sexual function among all incest offenders. In addition, on the BPRS, which was used to determine the degree of psychiatric disturbance, the group with younger victims was found to be significantly more disturbed than the group with older victims. Psychopathy as measured by the PCL-R, revealed no difference between the two groups.
Phallometric Analyses
Laboratory measures of sexual arousal have generally indicated that child molesters have a relatively distinct preference for sexual activity with children34,35,41 compared with rapists and nonsexual offenders. As mentioned earlier, both groups exhibited clinically significant deviant sexual arousal to child stimuli. However, there were no significant differences between the groups on these indices of deviant arousal. It is possible that the stimuli simply did not adequately target the preference for such young victims. For example, the recordings describe a child but not a particular child and the descriptions are quite general, outside of gender. Perhaps other forms of stimuli that specifically address the physical, intellectual, and emotional qualities of a very young child may have elicited differential responses from the two groups.
Victim and Offense Characteristics
The results of the current study show sexual aggressors against infants and toddlers to be more likely to have multiple victims. Of course, more victims may simply be the result of more opportunities for abuse in this particular sample, but it may suggest that these offenders either possess less restraint or experience stronger inclinations to victimize. As discussed in the Results section, the results of analysis of this variable should be interpreted with caution, as the phrase "multiple victims" may connote serial offending or a degree of pathology not represented in the data. Most of the offenders against younger victims who had more than one victim had two victims. Another finding concerning victim characteristics was that the offenders with younger victims were more likely to abuse boys. This observation does not eliminate the possibility that these same individuals also offended against girls. In fact, in the entire incest database from which the two groups in this study were sampled, only 19 of 342 offenders did not have a female victim. It has been noted in the literature on incestuous sex offenders that the largest proportion of this population offend against females, specifically daughters and stepdaughters.42 It has also been observed clinically that sons are extremely unlikely to be molested by fathers, even though extrafamilial child molesters often target boys.43 If this reported preference for female victims among paternal offenders is also characteristic of the preferences of other incest offenders, the preponderance of sexual aggressors against infants/toddlers with male victims in this study may suggest that they are less selective in their choice of victims, perhaps due to a stronger urge to offend against or greater generalizability of arousal in response to young children.
Offenders of infants/toddlers were also significantly more likely to have offended against their nephew/niece or grandson/granddaughter than were the offenders against older children. Speculation about this finding leads to the conclusion that it may be easier for these more distant relatives to exploit younger children than older children because they have less power over the child than do closer relatives, whereas an uncle or grandfather in an elder role of esteem and trust would be provided ideal opportunities, such as visits and babysitting, to groom and victimize a vulnerable family member with frequency and privacy. This finding having been noted, one wonders why the role of father or stepfather might not have featured more prominently in this groups’ membership, given that they would presumably have even more direct access to their own children. Unfortunately, the data from the current study do not specifically address why a particular victim was chosen.
Conclusion
The current study represents a preliminary effort to determine whether differences exist between incestuous child molesters as a function of the victim’s age. Most of the differences found are in line with a general prediction that the offenders with younger victims represent a group with generally more disorder personalities. Previous attempts to find differences between groups of child molesters arranged according to victim age may have been hampered by small sample sizes, but an argument could be made that the incestuous component incorporated in the offenses of the current sample of child molesters may have had an effect on the results. In other words, if the sexual abuse of an unrelated infant or toddler represents an especially deviant act, it may follow that the same offense perpetrated by someone against a member of his own family—possibly a daughter or nephew—constitutes even further pathology and thus perhaps a more disordered offender.
A unique feature regarding the current study is the nature of the population considered. To date, a great number of studies regarding sex offenders have involved samples from maximum security psychiatric hospitals or prisons. In the present study, the subjects were convicted of sexual offenses against children and were assessed before sentencing in a sexual behaviors clinic located in the forensic unit of a large general psychiatric hospital. Therefore, this group may be much more representative of the wide array of men processed through the courts than those in many previous studies.
Limitations
A limitation of this study is the heterogeneity in the group of sexual offenders against younger children. Participants in this group were defined as anyone from the incest database who had offended against a child five years of age or younger. This criterion did not exclude offenders who had, for example, offended against a 14-year-old, so long as he had also offended against at least one child who was 5 years of age or younger. With this in mind, the 19 offenders who also had victims older than five years, but were classified as offenders with younger victims, may have been no more infantophilic than they were hebephilic, except where the ages of their victims are concerned. To the extent that victims represent an arousal preference, or an absence of one, it may be that this confound is one of the reasons that significant differences in phallometric results were not observed. Despite this limitation, the group that had offended against adolescents, defined as those whose victims were exclusively 12 years of age or older, was indeed homogeneous, and the fact remains that in the current study those who had offended against an infant or toddler could indeed be distinguished on several notable variables from those who had not.
The current study also relies somewhat on self-report data, which may be vulnerable to biased responses from those wanting to present themselves in a more favorable light—especially incarcerated sex offenders.44 However, the validity of many of the measures used in the current study are not unduly threatened by social desirability.
Future Research
It would be interesting to compare incestuous offenders against young children to a similarly defined group of extrafamilial offenders, with an adequate sample size to determine the differential psychopathology. Another important variable to investigate regarding incestuous offenders may be the extent of the caregiving role that the offender fulfilled and the amount of time the offender spent with the child prior to the offense. According to the Westermarck hypothesis, humans may have an innate incest-avoidance mechanism such that persons raised together from early childhood have an aversion to sexual intercourse with one another.45 One would speculate that a sufficient quality and duration of time spent with a young child in a caring capacity serves to inoculate adult family members against inclinations toward sexual interactions.
References
Kalichman SC: Psychopathology and personality characteristics of criminal sex offenders as a function of victim age. Arch Sex Behav 20:187–97, 1991
Williams LM, Finkelhor D: The characteristics of incestuous fathers: a review of recent studies, in Handbook of Sexual Assault: Issues, Theories, and Treatment of the Offender. Edited by Marshall WL, Laws DR, Barbaree HE. New York: Plenum Press, 1990, pp 231–55
Studer LH, Clelland SR, Aylwin AS, et al: Rethinking risk assessment for incest offenders. Int J Law Psychiatry 23:15–22, 2000
Frenzel RR, Lang RA: Identifying sexual preferences in intrafamilial and extrafamilial child sexual abusers. Ann Sex Res 2:255–75, 1989
Freund K, Watson R, Dickey R: Sex offenders against female children perpetrated by men who are not pedophiles. J Sex Res 28:409–23, 1991
Abel GG, Becker JV, Murphy WD, et al: Identifying dangerous child molesters, in Violent Behavior: Social Learning Approaches in Prediction, Management, and Treatment. Edited by Stuart RB. New York: Brunner/Mazel, 1981, pp 116–37
Langevin R, Watson R: A comparison of incestuous biological and stepfathers. Ann Sex Res 4:141–50, 1991
Cupoli JM, Sewell PM: 1049 children with a chief complaint of sexual abuse. Child Abuse Neglect 12:151–62, 1988
Dube R, Hebert M: Sexual abuse of children under twelve years of age: a review of 511 cases. Child Abuse Neglect 12:321–30, 1988
Greenberg DM, Firestone P, Bradford JM, et al: Infantophiles, in Serial Offenders: Current Thoughts, Recent Findings. Edited by Schlesinger LB. New York: CRC Press, 2000, pp 229–46
Finkelhor D: Crimes in the Nursery. Newbury Park, CA: Sage, 1988
Schetky DH: The sexual abuse of infants and toddlers, in American Psychiatric Press Review of Psychiatry (vol 10). Edited by Tasman A, Goldfinger SM. Washington, DC: American Psychological Association 1991, pp 308–19
Greenberg DM, Bradford J, Curry S: Infantophilia: a new subcategory of pedophilia?—a preliminary study. Bull Am Acad Psychiatry Law 23, 63–71, 1995
Hare RD: Manual for the Revised Psychopathy Checklist. Toronto, Ontario, Canada: Multi-Health Systems, 1991
Udry JR, Talbert L, Morris NM: Biosocial foundations for adolescent female sexual behavior. Demography 23:217–27, 1986
Schickedanz JA, Schickedanz DI, Forsyth PD, et al: Understanding Children and Adolescents. Boston: Allyn and Bacon, 1998
Derogatis LR: Derogatis Sexual Functioning Inventory. Baltimore: Clinical Psychometrics Research, 1978
Derogatis LR: Psychological assessment of psychosexual functioning. Psychiatr Clin North Am 3:113–31, 1980
Hanson RK, Cox, B, Woszcsyna C: Sexuality, Personality, and Attitude Questionnaires for Sexual Offenders: A Review. Ottawa, Ontario, Canada: Solicitor General Canada, Ministry Secretariat, Supply and Services Canada publication no. JS4-1/1991-13
Overall JE, Gorham DR: Brief Psychiatric Rating Scale. Psychol Rep 10:799–812, 1962
Dingeman PM: The Brief Psychiatric Rating Scale (BPRS) and the Nurses’ Observation Scale for Inpatient Evaluations (NOSIE) in the evaluation of positive and negative symptoms. J Clin Psychol 46:168–74, 1990
Gibbs LE: Validity and reliability of the Michigan Alcoholism Screening Test: a review. Drug Alcohol Depend 12:279–85, 1983
Magruder-Habid K, Durand AM, Frey KA: Alcohol abuse and alcoholism in primary health care settings. J Fam Pract 32:406–13, 1991
Marguder-Habid K, Stevens HA, Alling WC: Relative performance of the MAST, VAST, and CAGE versus DSM-III-R criteria for alcohol dependence. J Clin Epidemiol 46:435–41, 1993
Allnut SH, Bradford JMW, Greenberg DM, et al: Co-morbidity of alcoholism and the paraphilias. J Forensic Sci 41:234–9, 1996
Hucker S, Langevin R, Bain J: A double blind trial of sex drive reducing medication in pedophiles. Ann Sex Res 1:227–42, 1988
Rada RT: Alcoholism and forcible rape. Am J Psychiatry 132:444–6, 1975
Rada RT, Laws DR, Kellner R: Plasma testosterone levels in the rapists. Psychosom Med 38:257–68, 1976
Hare RD, Forth AE, Strachan KE: Psychopath and crime across the life span, in Aggression and Violence Throughout the Life Span. Edited by Peters RD, McMahon J, Quinsey VL. Newbury Park, CA: Sage, 1992, pp 285–300
Harris GT, Rice ME, Quinsey VL: Psychopathy as a taxon: evidence that psychopaths are a discrete class. J Consult Clin Psychol 62:387–97, 1994
Quinsey VL, Rice ME, Harris GT: Actuarial prediction of sexual recidivism. J Interpers Violence 10:85–105, 1995
Hare RD, Harpur TJ, Hakstian AR, et al: The Revised Psychopathy Checklist: reliability and factor structure. Psychol Assess 2:338–41, 1990
Abel GG, Blanchard EB, Barlow DH: Measurement of sexual arousal in several paraphilias: the effects of stimulus modality, instructional set, and stimulus content on the objective. Behav Res Ther 19:25–33, 1981
Firestone P, Bradford JM, Greenberg DM, et al: Differentiation of homicidal child molesters, nonhomicidal child molesters, and nonoffenders by phallometry. Am J Psychiatry 157:1847–50, 2000
Firestone P, Bradford JM, McCoy M, et al: Prediction of recidivism in extrafamilial child molesters based on court-related assessments. Sex Abuse 12:203–21, 2000
Firestone P, Bradford JM, Greenberg DM, et al: Homicidal sex offenders: psychological, phallometric, and diagnostic features. J Am Acad Psychiatry Law 26:537–52, 1998
Tabachnick BG, Fidel LS: Using Multivariate Statistics (ed 2). New York: Harper & Row, 1989
Hanson KR: Age and sexual recidivism: a comparison of rapists and child molesters. Ottawa, Canada: Solicitor General Canada, Ministry Secretariat, Supply and Services Canada, publication no. JS42-96/2001
Hanson KR, Slater S: Reactions to motivational accounts of child molesters. J Child Sex Abuse 2:43–59, 1988
Derogatis LR: The DSFI: a multi-dimensional measure of sexual functioning. J Sex Marital Ther 5:244–81, 1979
Chaplin TC, Rice ME, Harris GT: Salient victim suffering and the sexual responses of child molesters. J Consult Clin Psychol 63:249–55, 1995
Gibbens TCN, Soothill K, Way CK: Sibling and parent-child incest offenders. Br J Criminol 18:40–52, 1978[Free Full Text]
Rice ME, Harris GT: Men who molest their sexually immature daughters: is a special explanation required? J Abnorm Psychol 111:329–39, 2001
Nugent PM, Kroner DG: Denial, response styles, and admittance of offences among child molesters and rapists. J Interpers Violence 11:475–86, 1996
Shepher J: Incest: A Biosocial View. New York: Academic Press, 1983(Philip Firestone, PhD, Kr)
Abstract
The purpose of the present study was to compare incest offenders (IOs) whose victims include infants or toddlers to IOs with adolescent victims on several variables commonly examined in the sexual offender literature. Participants were 48 men whose youngest victim was less than 6 years of age (younger-victim incest offenders; YVs); and 71 men whose youngest victim was 12 to 16 years of age (older-victim incest offenders (OVs). In general, YVs showed more emotional disturbance and pathology than OVs. Compared with OVs, YVs had a greater history of substance abuse and more current problems with alcohol. In addition, YVs reported significantly poorer sexual functioning and were significantly more psychiatrically disturbed. YVs were also more likely to have a male victim, to have victimized a nephew/niece or grandson/granddaughter, and to have denied their offense(s). It was evident that both the YVs and OVs demonstrated clinically significant difficulty with normal sexual functioning and exhibited deviant sexual arousal.
While sexual victimization of any age group is unacceptable, the current admonitions surrounding such offenses against adolescents could be viewed as largely attributable to societal/cultural norms and mores and to investigations regarding the physical and psychological damage endured by a young person who has been abused. That is, in an effort to protect young people, most modern societies have designated those under the age of roughly 16 as psychologically unprepared for sexual interactions (particularly with adults), irrespective of physical traits or reproductive competencies present that may suggest otherwise. For research and management purposes—if not in everyday parlance—it is primarily this arbitrary victim age demarcation that has formed the basis for labeling a sexual offender as a child molester. However, perhaps under the blanket designation of child molester there are important subgroups based on the victim’s age. Sexual abuse of an infant or toddler not only represents the same breach of society’s values and laws that a similar crime against an adolescent does, but it also contravenes any semblance of adaptive sexual behavior, biologically or otherwise. With this difference in mind, it is reasonable to predict that a person who offends against a very young child would differ, on one or many levels, from a person whose sexual assault victims appear to be limited to adolescents. Supporting this line of thought, Kalichman1 has reported that child molesters who abuse prepubescent children demonstrate significantly more psychopathology and emotional disturbance than those with older victims.
Although incest offenders are often treated in the literature as a discrete category, it has been noted that no universal characteristics or single profiles of such offenders have been found.2 Relatively common characteristics include passive personalities, dependent personalities, physical and emotional maltreatment during childhood, marital dissatisfaction, sexual dissatisfaction, and disturbances in empathy and attachment.2 Studer et al.3 observe that because their offending is limited to family members, it is commonly held that incest offenders pose a limited threat to society as a whole, and as a result of this perception, they are often given less intensive treatment. This characterization may be partially sourced from the results of comparisons between incestuous and nonincestuous child molesters that suggest that the arousal patterns of incest offenders are less sexually deviant.4,5 However, some studies have failed to uncover differences in this regard.6,7 We contend that the inconsistency in sexual preference findings in incest offenders and the difficulties experienced by researchers in developing a reliable profile of this population stem from a failure to attend to a possible heterogeneity within this category, based on the victim’s age.
Sexual abuse of infants and toddlers is not uncommon, with cases involving victims as young as two to three months having been reported.8,9 Accurate estimates of the number of very young children who are sexually abused are difficult to obtain, as the limited verbal and abstraction skills of the victims, along with the damaging effects of the abuse itself, present significant impediments to full or even reasonable disclosure.10 Finkelhor11 found that of 270 cases of sexual abuse of children in daycare, 6 percent occurred against children younger than two years and 60 percent involved children less than four years of age. Schetky12 surveyed Child Protection Services serving three small counties in Maine and found 53 substantiated cases of sexual abuse involving children younger than four years between the years of 1985 and 1988. According to Schetky, the perpetrators of sexual abuse against infants and toddlers tend to be family members, babysitters, or daycare workers who capitalize on the access to defenseless victims furnished by their caregiving role, an opportunity not available to a stranger.
Although the authors are unaware of any published research in which incest offenders are compared as a function of victim age, some comparisons of extrafamilial child molesters have been conducted. Greenberg et al.,10 in a comparison of extrafamilial child molesters who had offended against victims younger than five years with those who had offended against victims aged 8 to 12 years, found few differences on measures of victim characteristics, offense characteristics, or psychological characteristics of the offender. However, offenders who victimized younger children were significantly younger than those who victimized older children. It is possible that the failure to find further statistically significant differences was due to the small sample size. A follow-up study was conducted by Greenberg et al.,13 and, as in the previous study, the offenders against younger children were observed to have features similar to those whose offenses were against older children, with the exception, once again, of age. In addition, offenders against young victims were significantly more likely to deny their offenses and to have a higher score on Factor 2 (antisocial lifestyle) of the Psychopathy Checklist-Revised (PCL-R14). Although this study involved 192 offenders against older children, there were only 18 whose victims were younger children, which may have again failed to provide adequate statistical power to detect real differences that may have existed between the two groups. In light of Schetky’s12 observation of the preponderance of family members involved in cases of infant and toddler sexual abuse and given the importance of frequent and isolated access to the victim in this form of sexual offending, it could be expected that this phenomenon is more likely to be witnessed in incest offenders. The studies by Greenberg and his colleagues13 involved samples of extrafamilial child molesters only. The exclusion of incest offenders may have contributed to the small sample sizes of the groups with victims younger than five years.
In the current study, we investigated whether incest offenders who have offended against very young children are distinguishable from those whose youngest victims were pubertal (as far as criminal and psychiatric files indicate). There is some similarity in the current study to the methodology and focus of the earlier studies of sexual abusers of infants and toddlers by Greenberg et al.10,13 Greenberg and his colleagues are also affiliated with the Sexual Behavior Clinic at the Royal Ottawa Hospital and used data from a database similar to the one we have used. However, Greenberg et al. studied extrafamilial child molesters, while our investigation focused on a sample of incest offenders, and so there is no overlap in participants. To our knowledge, there have been no other studies of child molesters who have offended against children under the age of five in such a population of offenders. A potential limitation of the studies by Greenberg et al. concerns the fact that the group with victims 5 years of age or younger was compared with a group with victims as old as 16 but as young as 6 years. This may have masked differences that existed between offenders in the two groups whose victims’ ages were more dissimilar. With this in mind, another key difference in the protocol of this study is the use of a comparison group consisting of men who have exclusively offended against post-pubertal victims in their index offenses. This decision was made to maximize the contrast between the ages, and thus the physical and emotional characteristics of the victims of the two groups of offenders. Twelve years was chosen as the lower limit for the group with older victims based on the fact that the average age for the onset of menses in girls in the United States is 12.5 years,15 and the average age at which pubic hair develops in boys is 13.16
It is our contention that a man who sexually abuses an adolescent who presumably displays at least some semblance of sexual maturity, while no less criminal in his actions, presents a less disturbed person than one who offends similarly against society’s most vulnerable and least sexually provocative demographic: infants and toddlers. The current study is preliminary and descriptive in nature, and the existing literature does not lend itself to many specific hypotheses, due to the current study’s differing population (i.e., incest offenders) and methodology. However, several differences are expected to emerge and to be congruent with the central prediction that the group with victims younger than six years of age are more generally disordered and pathological. The groups were compared on demographic and historical variables, psychological test scores, phallometric indices, index offense and victim characteristics, and criminal histories. Preliminary comparisons of this nature can contribute to the elucidation of the traits of those in the incest offender category. Furthermore, evidence supporting heterogeneity within this population may hold significance in treatment and management of these offenders.
Method
Participants
The present sample consisted of two groups of adult men convicted of hands-on sexual offenses against family members (biological child, stepchild, nephew, niece, grandchild, or sibling) who were 16 years old or younger. All participants were assessed at the Royal Ottawa Hospital, Sexual Behaviors Clinic. In one group were 48 men whose youngest victims were less than six years of age (younger-victim incest offenders; YVs). In the other group were 71 men whose youngest victims were at least 12 years of age (older-victim incest offenders; OVs). Ideally, the YV group would have consisted of men whose oldest victims were also less than six years of age. However, this degree of purity could not be furnished by the sample. Signed informed consent was obtained from all participants. The Research Ethics Committee of the Royal Ottawa Hospital approved the research protocol. Although the offenders’ names appeared on both criminal and psychiatric records, confidentiality was maintained by numerically labeling each case in the computer database.
Measures
Sexual Functioning
The Derogatis Sexual Functioning Inventory (DSFI) is designed to assess general and specific dimensions of sexual functioning.17,18 It collects information by using numerous items at once to grasp "the fundamental components judged essential to effective sexual behavior" (Ref. 18, p 117). The 10 subscales are Information, Experience, Sexual Drive, Sexual Attitude, Psychological Symptoms (also known apart from the DSFI as the Brief Symptom Inventory), Affect, Gender Role Definition, Sexual Fantasy, Body Image, and Satisfaction. The Sexual Functioning Index (SFI) is a global measure derived by adding the 10 subtest scores. It provides an overall measure of an individual’s level of sexual functioning. The DSFI has been used with large nonforensic samples, but its use with sexual offenders is limited. There is some suggestion that sexual offenders show high levels of sexual dissatisfaction.19
Psychiatric Disturbance
The Brief Psychiatric Rating Scale (BPRS)20 is an inventory designed to provide clinician ratings on psychiatric symptomatology in 16 domains, such as anxiety, emotional withdrawal, somatic concerns, and other symptomatology. For each domain, the clinician is required to rate the degree to which symptomatology is present in the subject. Ratings may range from a low score of zero (not present) to a high of six (extremely severe). In addition to providing information about specific symptomatology, a total score provides an overall indication of psychiatric disturbance. An acceptable internal consistency coefficient of .63 has been reported.21
Alcohol Abuse
The Michigan Alcoholism Screening Test (MAST), a 24-item self report inventory, is used to identify behavior indicative of alcohol abuse.22 The internal consistency is adequate, with a reported overall coefficient of .87, and the test is relatively unaffected by age or by denial of socially unacceptable characteristics.23,24 Scores of five or six are considered suggestive of alcohol problems, and a score of seven or more is considered strongly indicative of alcohol abuse.25 The MAST has been found to correlate with DSM-III-R criteria for alcohol dependence.24 The MAST has been extensively used as a screening tool for alcoholism, and many studies have used samples of sex offenders.25–28
Psychopathy
The Psychopathy Checklist-Revised consists of 20 clinical rating scales designed to assess behavior and personality characteristics considered fundamental to psychopathy.14 Rigorous testing has indicated that the PCL-R is a psychometrically sound instrument.29 Valid PCL-R ratings can be made on the basis of high-quality archival information.30,31 The existence of two factors has been replicated using various samples: (1) the degree of personality, interpersonal, and affective traits deemed significant to the construct of psychopathy and (2) the degree of antisocial behavior and an unstable, corrupted lifestyle.32 Previous studies have shown the interrater reliability and internal consistency of both factors to be high, despite the small number of items per factor.14,32 In the present investigation, the PCL-R was completed by two research assistants who examined descriptive material contained in institutional files. These files included criminal records, court reports, and psychiatric records. A random sample of 100 clinic files was independently rated by each researcher, resulting in satisfactory interrater reliability (r = 0.88; p < .0001).
Measurement of Sexual Arousal
Changes in penile circumference in response to audio stimuli were measured by means of an indium-gallium strain gauge.
Stimulus Presentation.
The order of stimulus presentation, held constant for all participants, was computer controlled. Participants were presented with one or more of three series of audiotapes, according to the nature of the participants’ sexual offense. The audiotapes consisted of 120-second vignettes that described activities varying in age and gender of the victim or partner, and degree of violence portrayed.33 After being instructed to allow normal arousal to occur, each participant was presented with at least one full set containing one vignette from each category. The categories used in the assessment of arousal in response to children that are relevant to the current study were: (1) a child initiating sexual contact, (2) the child participating mutually in sexual activity, (3) nonphysical coercion of a child, (4) physical coercion of a child, (5) sadistic sex with a child, (6) nonsexual assault of a child, and (7) consenting sex with an adult. The audiotape series used to identify sexual attraction to rape included two scenarios of two minute’s duration in each of three categories: (1) consenting sex with an adult female, (2) the rape of a female adult, and (3) the nonsexual assault of a female adult.
Scoring.
The Pedophile Index (PI) was computed by dividing the highest response to the child initiates or child participates mutually stimulus by the highest responses to a consenting adult stimulus. Adequate criterion validity for the PI (utilizing the same stimuli as the present study) has been demonstrated by its ability to differentiate child molesters and non-offenders significantly at levels of accuracy above chance.34
The Pedophile Assault Index (PAI) was computed by dividing the highest response to an assault stimulus involving a child victim (nonphysical coercion, physical coercion, sadistic sex, or nonsexual assault) by the highest response to the child initiates or child participates mutually stimulus. Adequate criterion validity for the PAI (utilizing the same stimuli as the present study) has been demonstrated in past studies that found significant differences between homicidal sex offenders and non-offenders.35,36
Procedure
Most of the data examined in the current study, such as demographic, historical, psychological test, phallometric, and index sexual offense information, were derived from the medical records or the participants’ assessments, which routinely involved numerous interviews and extensive testing. After the assessments, data on prior criminal charges were gathered from the Canadian Police Information Center (CPIC), which has compiled a national database of criminal arrests and convictions, including INTERPOL reports from the Royal Canadian Mounted Police.
Results
Before statistical tests were performed, the data were screened to ensure that the assumptions underlying the tests were not violated. In the very limited number of cases in which outlying values occurred, they were adjusted upward or downward according to the direction of the problem. This method is appropriate when case retention is desirable and when it does not unduly influence the group means.37 Student’s t tests were performed on continuous variables with = .05; two-tailed and dichotomous variables were analyzed by chi-square, cross-tabular, independent group tests with = .05.
Demographic and historical data for YVs and OVs are presented in Table 1. The groups did not differ in age or number of years of education. However, a statistical trend (p < .10) suggested that the YVs had lower full-scale IQ scores than the OVs. The YV group was also significantly more likely to have a history of drug abuse and alcohol dependency. In terms of family background, the group with younger victims was significantly more likely to have been raised in a family with a history of criminality, but there were no other significant differences.
The two groups of incest offenders were compared on a variety of psychological and phallometric measures (Table 2). The YVs had poorer overall sexual functioning than the OVs, as evidenced by significantly lower group mean scores on the DSFI. It is interesting to note that neither group scored above the fifth percentile on this sexual functioning inventory. In addition, scores on the MAST indicated that YVs are significantly more likely than OVs to be alcohol dependent; was adjusted for this comparison, because the assumption of homogeneity of variance was not met. Offenders against younger children also exhibited significantly more psychiatric disturbance than the other group on the BPRS. The PCL-R did not distinguish between the YV and OV groups. As indicated at the bottom of Table 2, phallometric analysis revealed that both groups showed clinically meaningful levels of deviant sexual preferences. Nevertheless, there was no significant difference between the YV and OV participants.
Nonparametric analysis of data regarding the offender’s sexual offenses and the victim’s characteristics revealed that YVs were more likely to have multiple victims (Table 3). This variable was calculated with dichotomous data (i.e., one victim versus more than one victim). Examination of the frequencies, however, revealed that the majority of YVs (87.5%) had two or fewer victims, while the remaining 12.5 percent had between three and seven victims. In other words, it may be more appropriate to consider YVs more likely to have two victims than OVs.
The procedure by which the YV and OV groups were constructed in the current study created a potential threat to internal validity. Specifically, the inclusion criteria for the YV and OV groups could have been responsible for the higher victim counts of the YVs and, consequently, for the other differences that were found between the groups. The YVs could have any number of victims older than five years (but <16 years), so long as they had at least one victim aged five years or younger. Meanwhile, for the sake of maintaining purity in the OV group, an individual who offended against a 13-year-old but also had a 7-year-old victim would not be included. To determine whether the exclusion of offenders from the OV group, because they also had victims between 6 and 11 years old, artificially lowered the number of victims in the OV group, the medians for each group were examined. Both those included in and excluded from the OV group had a median of one victim. This suggests that the stringent criteria for the OV group did not unduly affect the results on this variable.
The YVs were more likely to have offended against a male victim than the OVs. The victim’s relationship to the offender was derived from a categorical variable that listed the following options: stepson/stepdaughter, biological son/daughter, sibling, and relative (grandson/granddaughter, niece/nephew). The only significant difference between YVs and OVs concerning the victims’ relationships to the offenders was that of relative. All other victim relationships to the offender were equally likely in both groups. Another way of describing this finding is that YVs were more likely than OVs to have been the grandfathers or uncles of their victims. It was also evident that the YVs were more likely than OVs to deny having committed their offenses. Finally, there were no significant differences in criminal history between the two groups (Table 3).
Discussion
In general, the results of the current study provide support for the notion that incest offenders with prepubescent victims are a group with more disordered personalities than men who have sexually assaulted only pubescent or postpubescent children. With one exception, whether statistically significant or not, those men convicted of sexual offenses against younger children scored more poorly on nearly all measures studied, with the only exception being previous sexual offenses.
Demographic and Historical Information
Unlike the previous findings of Greenberg et al.13,10 in extrafamilial child molesters, the group with the younger victims in this study did not differ in age from the group with older victims. However, the average age of all participants in this study (mean (M) = 40.1 years) is considerably older than the ages of participants in the report by Greenberg and his colleagues (M = 27.9 years). The difference in age may be due to the participants in this study being incest offenders. Such offenders are usually older than extrafamilial child molesters.38 The two groups of men did not differ in the number of years of education completed. There was a statistical trend that emerged in the data suggesting that sexual aggressors against prepubescents have lower intelligence than sexual aggressors against adolescents, as measured by IQ scores. The rates of marriage in the histories of the two offender groups did not differ.
Williams and Finkelhor2 have noted that many incest offenders rationalize their crimes by blaming their offending behavior on alcohol. In the present study, there were no differences between the two groups regarding whether they reported being influenced by drugs or alcohol during the commission of the offense. Nevertheless, sexual offenders with young victims were more likely to have reported drug and alcohol abuse, than were offenders against older victims. The offenders with younger victims scored well beyond the cutoff of seven on the MAST, which is strongly indicative of alcohol abuse.25 The effects of recreational drugs or alcohol may augment an offender’s libidinal drive, distorting his perceptions or lowering his inhibitions, any number of which may lead to impaired judgment regarding sexual interactions with a very young member of his family. It is also possible that the sexual offenses against prepubertal children are part of the paraphilic attraction that these men have. It was certainly evident that both groups of offenders in the present investigation experienced clinically meaningful levels of deviant sexual arousal in response to children.
In the current investigation there was not a great deal of evidence to suggest that offenders with younger victims were more likely to have been raised in families with serious problems. It is worth noting that roughly 40 percent of all participants in this study reported that they were sexually and physically abused themselves, a proportion that is slightly higher than the one-third of pedophiles who reported being sexually victimized during their childhoods in other research.39 Although intuitively appealing, the link between commission of sexual crimes and offenders’ own experiences of sexual abuse is a contentious one. Hanson and Slater39 found that the proportions of sex offenders against children who reported being sexually abused themselves were similar to those found in other sexual and nonsexual offender groups.
Psychological Measures
The results of the current study agree with the observations of Kalichman1 that child molesters who abuse prepubescent children demonstrate significantly more psychopathology and emotional disturbance than do offenders with older victims. The group of men with younger victims was clearly more apt to abuse alcohol than the group with older child victims. In addition, on the DSFI, a general measure of sexual functioning, sexual aggressors against prepubescent children rated themselves as having significantly more problems than the other group. While this observation lends support to the hypothesis that those who offend against younger children are likely to be more disturbed than those who do not, it should be noted that the mean score on the DSFI for both groups placed them in approximately the fifth to seventh percentile of the population at large.40 This suggests that there is difficulty with sexual function among all incest offenders. In addition, on the BPRS, which was used to determine the degree of psychiatric disturbance, the group with younger victims was found to be significantly more disturbed than the group with older victims. Psychopathy as measured by the PCL-R, revealed no difference between the two groups.
Phallometric Analyses
Laboratory measures of sexual arousal have generally indicated that child molesters have a relatively distinct preference for sexual activity with children34,35,41 compared with rapists and nonsexual offenders. As mentioned earlier, both groups exhibited clinically significant deviant sexual arousal to child stimuli. However, there were no significant differences between the groups on these indices of deviant arousal. It is possible that the stimuli simply did not adequately target the preference for such young victims. For example, the recordings describe a child but not a particular child and the descriptions are quite general, outside of gender. Perhaps other forms of stimuli that specifically address the physical, intellectual, and emotional qualities of a very young child may have elicited differential responses from the two groups.
Victim and Offense Characteristics
The results of the current study show sexual aggressors against infants and toddlers to be more likely to have multiple victims. Of course, more victims may simply be the result of more opportunities for abuse in this particular sample, but it may suggest that these offenders either possess less restraint or experience stronger inclinations to victimize. As discussed in the Results section, the results of analysis of this variable should be interpreted with caution, as the phrase "multiple victims" may connote serial offending or a degree of pathology not represented in the data. Most of the offenders against younger victims who had more than one victim had two victims. Another finding concerning victim characteristics was that the offenders with younger victims were more likely to abuse boys. This observation does not eliminate the possibility that these same individuals also offended against girls. In fact, in the entire incest database from which the two groups in this study were sampled, only 19 of 342 offenders did not have a female victim. It has been noted in the literature on incestuous sex offenders that the largest proportion of this population offend against females, specifically daughters and stepdaughters.42 It has also been observed clinically that sons are extremely unlikely to be molested by fathers, even though extrafamilial child molesters often target boys.43 If this reported preference for female victims among paternal offenders is also characteristic of the preferences of other incest offenders, the preponderance of sexual aggressors against infants/toddlers with male victims in this study may suggest that they are less selective in their choice of victims, perhaps due to a stronger urge to offend against or greater generalizability of arousal in response to young children.
Offenders of infants/toddlers were also significantly more likely to have offended against their nephew/niece or grandson/granddaughter than were the offenders against older children. Speculation about this finding leads to the conclusion that it may be easier for these more distant relatives to exploit younger children than older children because they have less power over the child than do closer relatives, whereas an uncle or grandfather in an elder role of esteem and trust would be provided ideal opportunities, such as visits and babysitting, to groom and victimize a vulnerable family member with frequency and privacy. This finding having been noted, one wonders why the role of father or stepfather might not have featured more prominently in this groups’ membership, given that they would presumably have even more direct access to their own children. Unfortunately, the data from the current study do not specifically address why a particular victim was chosen.
Conclusion
The current study represents a preliminary effort to determine whether differences exist between incestuous child molesters as a function of the victim’s age. Most of the differences found are in line with a general prediction that the offenders with younger victims represent a group with generally more disorder personalities. Previous attempts to find differences between groups of child molesters arranged according to victim age may have been hampered by small sample sizes, but an argument could be made that the incestuous component incorporated in the offenses of the current sample of child molesters may have had an effect on the results. In other words, if the sexual abuse of an unrelated infant or toddler represents an especially deviant act, it may follow that the same offense perpetrated by someone against a member of his own family—possibly a daughter or nephew—constitutes even further pathology and thus perhaps a more disordered offender.
A unique feature regarding the current study is the nature of the population considered. To date, a great number of studies regarding sex offenders have involved samples from maximum security psychiatric hospitals or prisons. In the present study, the subjects were convicted of sexual offenses against children and were assessed before sentencing in a sexual behaviors clinic located in the forensic unit of a large general psychiatric hospital. Therefore, this group may be much more representative of the wide array of men processed through the courts than those in many previous studies.
Limitations
A limitation of this study is the heterogeneity in the group of sexual offenders against younger children. Participants in this group were defined as anyone from the incest database who had offended against a child five years of age or younger. This criterion did not exclude offenders who had, for example, offended against a 14-year-old, so long as he had also offended against at least one child who was 5 years of age or younger. With this in mind, the 19 offenders who also had victims older than five years, but were classified as offenders with younger victims, may have been no more infantophilic than they were hebephilic, except where the ages of their victims are concerned. To the extent that victims represent an arousal preference, or an absence of one, it may be that this confound is one of the reasons that significant differences in phallometric results were not observed. Despite this limitation, the group that had offended against adolescents, defined as those whose victims were exclusively 12 years of age or older, was indeed homogeneous, and the fact remains that in the current study those who had offended against an infant or toddler could indeed be distinguished on several notable variables from those who had not.
The current study also relies somewhat on self-report data, which may be vulnerable to biased responses from those wanting to present themselves in a more favorable light—especially incarcerated sex offenders.44 However, the validity of many of the measures used in the current study are not unduly threatened by social desirability.
Future Research
It would be interesting to compare incestuous offenders against young children to a similarly defined group of extrafamilial offenders, with an adequate sample size to determine the differential psychopathology. Another important variable to investigate regarding incestuous offenders may be the extent of the caregiving role that the offender fulfilled and the amount of time the offender spent with the child prior to the offense. According to the Westermarck hypothesis, humans may have an innate incest-avoidance mechanism such that persons raised together from early childhood have an aversion to sexual intercourse with one another.45 One would speculate that a sufficient quality and duration of time spent with a young child in a caring capacity serves to inoculate adult family members against inclinations toward sexual interactions.
References
Kalichman SC: Psychopathology and personality characteristics of criminal sex offenders as a function of victim age. Arch Sex Behav 20:187–97, 1991
Williams LM, Finkelhor D: The characteristics of incestuous fathers: a review of recent studies, in Handbook of Sexual Assault: Issues, Theories, and Treatment of the Offender. Edited by Marshall WL, Laws DR, Barbaree HE. New York: Plenum Press, 1990, pp 231–55
Studer LH, Clelland SR, Aylwin AS, et al: Rethinking risk assessment for incest offenders. Int J Law Psychiatry 23:15–22, 2000
Frenzel RR, Lang RA: Identifying sexual preferences in intrafamilial and extrafamilial child sexual abusers. Ann Sex Res 2:255–75, 1989
Freund K, Watson R, Dickey R: Sex offenders against female children perpetrated by men who are not pedophiles. J Sex Res 28:409–23, 1991
Abel GG, Becker JV, Murphy WD, et al: Identifying dangerous child molesters, in Violent Behavior: Social Learning Approaches in Prediction, Management, and Treatment. Edited by Stuart RB. New York: Brunner/Mazel, 1981, pp 116–37
Langevin R, Watson R: A comparison of incestuous biological and stepfathers. Ann Sex Res 4:141–50, 1991
Cupoli JM, Sewell PM: 1049 children with a chief complaint of sexual abuse. Child Abuse Neglect 12:151–62, 1988
Dube R, Hebert M: Sexual abuse of children under twelve years of age: a review of 511 cases. Child Abuse Neglect 12:321–30, 1988
Greenberg DM, Firestone P, Bradford JM, et al: Infantophiles, in Serial Offenders: Current Thoughts, Recent Findings. Edited by Schlesinger LB. New York: CRC Press, 2000, pp 229–46
Finkelhor D: Crimes in the Nursery. Newbury Park, CA: Sage, 1988
Schetky DH: The sexual abuse of infants and toddlers, in American Psychiatric Press Review of Psychiatry (vol 10). Edited by Tasman A, Goldfinger SM. Washington, DC: American Psychological Association 1991, pp 308–19
Greenberg DM, Bradford J, Curry S: Infantophilia: a new subcategory of pedophilia?—a preliminary study. Bull Am Acad Psychiatry Law 23, 63–71, 1995
Hare RD: Manual for the Revised Psychopathy Checklist. Toronto, Ontario, Canada: Multi-Health Systems, 1991
Udry JR, Talbert L, Morris NM: Biosocial foundations for adolescent female sexual behavior. Demography 23:217–27, 1986
Schickedanz JA, Schickedanz DI, Forsyth PD, et al: Understanding Children and Adolescents. Boston: Allyn and Bacon, 1998
Derogatis LR: Derogatis Sexual Functioning Inventory. Baltimore: Clinical Psychometrics Research, 1978
Derogatis LR: Psychological assessment of psychosexual functioning. Psychiatr Clin North Am 3:113–31, 1980
Hanson RK, Cox, B, Woszcsyna C: Sexuality, Personality, and Attitude Questionnaires for Sexual Offenders: A Review. Ottawa, Ontario, Canada: Solicitor General Canada, Ministry Secretariat, Supply and Services Canada publication no. JS4-1/1991-13
Overall JE, Gorham DR: Brief Psychiatric Rating Scale. Psychol Rep 10:799–812, 1962
Dingeman PM: The Brief Psychiatric Rating Scale (BPRS) and the Nurses’ Observation Scale for Inpatient Evaluations (NOSIE) in the evaluation of positive and negative symptoms. J Clin Psychol 46:168–74, 1990
Gibbs LE: Validity and reliability of the Michigan Alcoholism Screening Test: a review. Drug Alcohol Depend 12:279–85, 1983
Magruder-Habid K, Durand AM, Frey KA: Alcohol abuse and alcoholism in primary health care settings. J Fam Pract 32:406–13, 1991
Marguder-Habid K, Stevens HA, Alling WC: Relative performance of the MAST, VAST, and CAGE versus DSM-III-R criteria for alcohol dependence. J Clin Epidemiol 46:435–41, 1993
Allnut SH, Bradford JMW, Greenberg DM, et al: Co-morbidity of alcoholism and the paraphilias. J Forensic Sci 41:234–9, 1996
Hucker S, Langevin R, Bain J: A double blind trial of sex drive reducing medication in pedophiles. Ann Sex Res 1:227–42, 1988
Rada RT: Alcoholism and forcible rape. Am J Psychiatry 132:444–6, 1975
Rada RT, Laws DR, Kellner R: Plasma testosterone levels in the rapists. Psychosom Med 38:257–68, 1976
Hare RD, Forth AE, Strachan KE: Psychopath and crime across the life span, in Aggression and Violence Throughout the Life Span. Edited by Peters RD, McMahon J, Quinsey VL. Newbury Park, CA: Sage, 1992, pp 285–300
Harris GT, Rice ME, Quinsey VL: Psychopathy as a taxon: evidence that psychopaths are a discrete class. J Consult Clin Psychol 62:387–97, 1994
Quinsey VL, Rice ME, Harris GT: Actuarial prediction of sexual recidivism. J Interpers Violence 10:85–105, 1995
Hare RD, Harpur TJ, Hakstian AR, et al: The Revised Psychopathy Checklist: reliability and factor structure. Psychol Assess 2:338–41, 1990
Abel GG, Blanchard EB, Barlow DH: Measurement of sexual arousal in several paraphilias: the effects of stimulus modality, instructional set, and stimulus content on the objective. Behav Res Ther 19:25–33, 1981
Firestone P, Bradford JM, Greenberg DM, et al: Differentiation of homicidal child molesters, nonhomicidal child molesters, and nonoffenders by phallometry. Am J Psychiatry 157:1847–50, 2000
Firestone P, Bradford JM, McCoy M, et al: Prediction of recidivism in extrafamilial child molesters based on court-related assessments. Sex Abuse 12:203–21, 2000
Firestone P, Bradford JM, Greenberg DM, et al: Homicidal sex offenders: psychological, phallometric, and diagnostic features. J Am Acad Psychiatry Law 26:537–52, 1998
Tabachnick BG, Fidel LS: Using Multivariate Statistics (ed 2). New York: Harper & Row, 1989
Hanson KR: Age and sexual recidivism: a comparison of rapists and child molesters. Ottawa, Canada: Solicitor General Canada, Ministry Secretariat, Supply and Services Canada, publication no. JS42-96/2001
Hanson KR, Slater S: Reactions to motivational accounts of child molesters. J Child Sex Abuse 2:43–59, 1988
Derogatis LR: The DSFI: a multi-dimensional measure of sexual functioning. J Sex Marital Ther 5:244–81, 1979
Chaplin TC, Rice ME, Harris GT: Salient victim suffering and the sexual responses of child molesters. J Consult Clin Psychol 63:249–55, 1995
Gibbens TCN, Soothill K, Way CK: Sibling and parent-child incest offenders. Br J Criminol 18:40–52, 1978[Free Full Text]
Rice ME, Harris GT: Men who molest their sexually immature daughters: is a special explanation required? J Abnorm Psychol 111:329–39, 2001
Nugent PM, Kroner DG: Denial, response styles, and admittance of offences among child molesters and rapists. J Interpers Violence 11:475–86, 1996
Shepher J: Incest: A Biosocial View. New York: Academic Press, 1983(Philip Firestone, PhD, Kr)