According to the sexually abused-abuser hypothesis, childhood sexual victimization (CSV) among males increases the likelihood of later sexual offending against children. Why CSV is related to sexual offending against children, however, has yet to be determined. To explore mechanisms that may link CSV to sexual offending, we tested the relationship between CSV and the two main risk domains: pedophilic interest and antisocial orientation. In four studies, men convicted of sexual offences against children under 15 who reported experiencing CSV were generally more sexually interested in children—especially male children—and were more antisocial than those who did not report experiencing CSV. A meta-analysis of the results across our four studies showed that CSV was moderately associated with greater antisociality, but less so with pedophilic interests. Future research should test the extent to which pedophilic interest and antisocial orientation explain the relationship between CSV and sexual offending against children among convicted sexual offenders.
Many theories suggest that childhood sexual victimization (CSV) plays an important role in later sexual offending against children, but how? Perhaps CSV leads to sexual interest in children and antisocial tendencies which leads to sexual offending against children for some people.
The present paper further examines the relationship between CSV and indicators of pedophilic interest and antisocial tendencies in a series of studies.
We conducted statistical analyses on four datasets/samples that consisted of men convicted of sexual offences against children. Participants who reported experiencing CSV were generally more sexually interested in children—especially male children—and were more antisocial than those who did not report experiencing CSV. We averaged the results across the studies, and found that experiencing CSV was slightly related with more sexual interest in children, and was significantly related with more antisociality.
The findings from our paper and past research are consistent with (though certainly do not confirm) the hypothesis that CSV can lead to greater pedophilic interest and antisociality, which would increase the likelihood of sexual offending against children.
Childhood sexual victimization (CSV) was generally related to offending against more children. There were mixed findings on the relationship between CSV and viewing child pornography. CSV was related to patterns of antisociality (more violence, higher risk to reoffend). CSV was related to more impulsivity/antisocial lifestyle. CSV was related to fewer interpersonal/affective psychopathy traits.
Many theoretical models contend that childhood sexual victimization (CSV) plays an important role in later sexual offending against children (e.g.,
Researchers have hypothesized that CSV may lead to the development of pedophilic interest for some victims, which may in turn lead to sexual offending against children (e.g.,
Furthermore, theorists have speculated that CSV may cause desensitization to painful or anxiety-provoking experiences, callousness, empathy deficits, and disregard for the needs of others (
However, with respect to psychopathy, CSV appears to be more related to generic antisocial and unstable lifestyle traits, rather than to the interpersonal and affective traits. In a sample of people convicted of violent offences,
The present paper further examines the association between CSV and the two strongest predictors of sexual offending against children: pedophilic interest and antisocial orientation. We conducted secondary analyses on four datasets that were originally gathered for larger projects. Though our datasets did not permit tests of whether pedophilic interest and antisocial orientation mediate the relationship between CSV and sexual offending against children, we were able to test the first step of this hypothesized mediation; that is, the relationship of CSV with pedophilic interest and antisocial orientation. Based on the reviewed literature, we hypothesized that men who experienced CSV would have higher SSPI scores, more sexual offences against children and specifically male children, and view more child pornography. With regard to antisociality, we hypothesized that those who experienced CSV would score higher on actuarial tools assessing risk of general violent recidivism, score higher on unstable lifestyle/antisocial behavior traits (i.e., according to self-report psychopathy scales and proxy scales assessing Factor 2 psychopathy traits), have more total and violent convictions, violate more release conditions, and be younger at their first conviction.
In this study, we examined the association of CSV with indicators of pedophilia (SSPI, sexual offending against prepubescent children and male children; and viewing child pornography) and indicators of antisocial orientation (risk assessment instrument scores, Factor 2 psychopathy traits, number of total and violent convictions, conditional release violations, and age at first conviction).
The data examined in Study 1 were drawn from a dataset originally collected for a larger project. The larger dataset has been previously used for other manuscripts (e.g.,
Based on self-reported information documented in the file information (in the form of interview notes and written reports), individuals were classified as having experienced childhood sexual victimization if they were sexually abused and/or were exposed to sexual stimuli before the age of 13. Seventy-two (40.7%) men were classified as having experienced CSV prior to age 13 and 105 (59.3%) as not having experienced CSV.
The SSPI (
The VRAG (
The CATS (
The Static-2002 (
Cohen’s
The present research was approved by an institutional research ethics board and also a separate ethics board designated to review research involving the use and disclosure of health information. Three undergraduate- and one graduate-level research assistants received a full day of training on the variables and were examined on three cases to ensure that they coded the variables reliably. All variables examined in the current study were coded by research assistants retrospectively from official records in participants’ case files, which contained interview notes (self-reported information), formal assessment reports by the assessor at the clinic, official criminal records and police documentations provided with the referral, any prior collateral reports from correctional institutions and/or probation, demographic information, and some description of victim information. Most variables (e.g., CSV, pornography use) were coded based on self-report documented in the individuals’ files, unless collateral documentation provided details (e.g., police reports, correctional reports, previous assessments). Of note, due to some missing information, sample sizes may vary depending on the variables included in the analyses.
Variable | CSV < 13 years old |
Cohen’s |
|
---|---|---|---|
Yes |
No |
||
SSPI | 2.36 (1.51) |
2.10 (1.29) |
0.19 |
# of Victims Under 13a | 0.31 (0.17) |
0.28 (0.15) |
0.17 |
# of Male Victims Under 13 | 0.23 (0.52) |
0.12 (0.36) |
0.26 |
Child Pornography Use | 0.20 (0.41) |
0.19 (0.39) |
0.04 |
aVariable was log transformed.
With regard to the relationship between CSV and indicators of antisocial orientation (see
Variable | CSV < 13 years old |
Cohen’s |
|
---|---|---|---|
Yes |
No |
||
VRAG | 4.19 (10.39) |
-4.41 (9.48) |
0.87* |
CATS | 3.55 (2.73) |
1.84 (2.04) |
0.74* |
Static-2002: General Criminality Subscore | 1.81 (1.89) |
1.45 (1.61) |
0.21 |
Total Past Convictions | 6.03 (8.66) |
3.34 (6.72) |
0.35* |
Prior Violent Convictions | 1.21 (1.77) |
0.60 (1.29) |
0.40* |
Conditional Release Violations | 0.32 (0.47) |
0.20 (0.40) |
0.28 |
Age at First Conviction | 26.69 (14.70) |
31.71 (15.53) |
-0.33* |
*
In this study, we attempted to replicate results found in Study 1.
The data examined in this study were drawn from a dataset originally collected for a larger project. The larger dataset has been previously used in a thesis, conference presentations, and manuscript (e.g.,
The SSPI, VRAG, and CATS (see Study 1 Measures) were also used in Study 2.
As part of a comprehensive interview, participants were asked if they had experienced sexual abuse before the age of 12. Thirteen participants (out of 28) reported experiencing sexual abuse before the age of 12 and 15 participants reported no sexual abuse before the age of 12.
The SRP-III (
A self-report questionnaire was given to participants to complete under conditions of confidentiality. The questionnaire was developed for the original project and included questions on demographic and criminal history information. Some self-reported questions from the questionnaire included in this study were “Have you ever been charged with a sex offence against someone under the age of 12?” (0 =
As in Study 1, Cohen’s
Data for this study were collected as part of a larger project. The study was approved by the university research ethics board and the correctional agency. Similar to Study 1, files were coded retrospectively and no additional measures were administered for the purpose of this study. At the time of data collection, participants completed a questionnaire concerning demographic and offence history information, after consenting to participate. Therefore, available variables consisted of information collected from both official records and self-report.
Variable | CSV < 12 years old |
Cohen’s |
|
---|---|---|---|
Yes |
No |
||
Official Documentation | |||
SSPI | 2.85 (1.52) |
2.40 (1.30) |
0.32 |
Self-Report | |||
# of Charges Against Victims Under 12a | 0.30 (0.35) |
0.19 (0.31) |
0.34 |
# of Male Victims Under 12b, c | 0.13 (0.24) |
0.15 (0.33) |
-0.06 |
Child Pornography Use | 4.08 (4.33) |
3.93 (4.37) |
0.03 |
aOnly includes victims from reported offences. bIncludes all victims (i.e., from reported and unreported offences). cVariable was log transformed.
As in Study 1, those who reported experiencing CSV had higher SSPI scores (
With regard to the relationship between CSV and indicators of antisocial orientation (see
As for self-reported psychopathy (i.e., SRP-III), participants who reported experiencing CSV had higher criminal tendencies (
Variable | CSV < 12 years old |
Cohen’s |
|
---|---|---|---|
Yes |
No |
||
Official Documentation | |||
VRAG | 3.77 (7.65) | 1.40 (10.78) | 0.25 [-0.50, 1.00] |
CATS | 0.77 (2.45) | -0.33 (2.58) | 0.44 [-0.31, 1.19] |
Total Past Convictions | 0.61 (0.55) | 0.31 (0.49) | 0.58 [-0.18, 1.33] |
Self-Report | |||
Prior Violent Convictions | 2.00 (3.24) | 0.33 (0.82) | 0.73 [-0.04, 1.50] |
Conditional Release Violations | 1.00 (2.20) | 0.33 (0.82) | 0.41 [-0.34, 1.16] |
SRP: Interpersonal Manipulation | 35.08 (7.16) | 36.20 (7.76) | -0.15 [-0.89, 0.59] |
SRP: Callous Affect | 35.08 (8.17) | 37.33 (7.74) | -0.28 [-1.03, 0.46] |
SRP: Erratic Lifestyle | 46.00 (7.21) | 45.33 (8.13) | 0.09 [-0.66, 0.83] |
SRP: Criminal Tendencies | 41.62 (11.35) | 37.93 (10.35) | 0.34 [-0.41, 1.09] |
SRP: Total | 157.77 (23.75) | 156.80 (27.86) | 0.04 [-0.71, 0.78] |
In this study, we attempted to replicate results, as well as clarify inconsistent findings, found in the previous studies.
The data examined in this study were drawn from a dataset originally collected for a larger project. The larger dataset has been previously used in manuscripts and a conference presentation (e.g.,
The SSPI and the self-report questionnaire from Study 2 were also used in Study 3.
As part of a larger questionnaire, participants were asked to self-report the number of times they had been sexually abused before the age of 12. Participants responded on a 10-point scale ranging from 0 (
The SIR-R1 (
Again, Cohen’s
Procedure for this study was identical to the procedure in Study 2. The study was approved by the university research ethics board and the correctional agencies.
Variable | CSV < 12 years old |
Cohen’s |
|
---|---|---|---|
Yes |
No |
||
Official Documentation | |||
SSPI | 2.54 (1.39) | 2.36 (1.08) | 0.15 [-0.61, 0.90] |
Age of Youngest Victim | 7.31 (2.36) | 6.21 (2.78) | 0.42 [-0.34, 1.19] |
Self-Report | |||
# of Charges Against Victims Under 12a | 0.62 (0.77) | 0.93 (1.00) | -0.35 [-1.11, 0.41] |
# of Male Victims Under 12b | 0.31 (0.85) | 0.00 (0.00) | 0.52 [-0.25, 1.29] |
Child Pornography Use | 2.92 (4.05) | 2.14 (3.76) | 0.20 [-0.56, 0.96] |
aOnly includes victims from reported offences. bIncludes all victims (i.e., from reported and unreported offences).
With regard to the relationship between CSV and indicators of antisocial orientation (see
Variable | CSV < 12 years old |
Cohen’s |
|
---|---|---|---|
Yes |
No |
||
Official Documentation | |||
SIR-R1 | 9.83 (7.18) |
15.69 (7.64) |
-0.79 |
Self-Report | |||
Prior Violent Convictionsa | 0.14 (0.26) |
0.13 (0.33) |
0.04 |
Conditional Release Violationsa | 0.02 (0.08) |
0.05 (0.19) |
-0.18 |
aVariable was log transformed.
This study was conducted as another attempt to replicate or clarify results found in the previous studies.
The data examined in this study were drawn from a dataset originally collected for a larger project. The larger dataset has been previously used in a thesis, conference presentations, and manuscript (e.g.,
The SSPI, SIR-R1, and Self-Report Questionnaire (see Study 1 and 3 Measures section) were also used in Study 4.
The measurement and operationalization for CSV in this study were the same as those in Study 3. Eight out of 17 participants reported experiencing CSV and nine reported never experiencing CSV.
Again, Cohen’s
Procedure for this study was identical to that in Study 2. The study was approved by our university research ethics board and the correctional/forensic agencies.
Variable | CSV < 12 years old |
Cohen’s |
|
---|---|---|---|
Yes |
No |
||
Official Documentation | |||
SSPI | 4.00 (1.51) |
3.60 (1.67) |
0.25 |
Age of Youngest Victim | 6.25 (2.82) |
10.78 (2.64) |
-1.66* |
Self-Report | |||
# of Charges Against Victims Under 12a | 0.48 (0.31) |
0.04 (0.11) |
1.89* |
# of Male Victims Under 12 | 3.43 (3.87) |
0.00 (0.00) |
1.14 |
Child Pornography Use | 3.00 (4.07) |
1.44 (2.92) |
0.44 |
aVariable was log transformed.
*
With regard to the relationship between CSV and indicators of antisocial orientation (see
Variable | CSV < 12 years old |
Cohen’s |
|
---|---|---|---|
Yes |
No |
||
Official Documentation | |||
SIR-R1 | 2.50 (7.67) |
8.17 (8.59) |
-0.70 |
Self-Report | |||
Prior Violent Convictions | 2.88 (3.76) |
1.67 (2.29) |
0.39 |
Conditional Release Violationsa | 0.10 (0.19) |
0.12 (0.27) |
-0.10 |
aVariable was log transformed.
Although results were generally consistent across Studies 1 to 4 for some variables, results for other variables seemed to vary and we often lacked adequate statistical power due to small sample sizes. Therefore, to synthesize findings, we conducted a fixed-effects meta-analysis of the observed group differences of CSV with indicators of pedophilic interest and antisocial orientation. We examined the combined and synthesized effects of the two broad domains (i.e., indicators of pedophilic interest and indicators of antisocial orientation) across all four studies.
Within each study, we combined Cohen’s
Overall, the combined effect sizes from all four studies demonstrated that experiencing CSV was associated with slightly more pedophilic interests than not experiencing CSV (fixed effects weighted average
The first goal of this paper was to examine the relationship between CSV and indicators of pedophilic interest. Although the relationship between CSV and SSPI scores did not reach significance in any of the studies, our findings remain in the direction that is consistent with the literature across all four studies. Specifically, we found that those who had experienced CSV before the age of 12 or 13 were more pedophilic (as indicated by their SSPI scores), which is consistent with the literature (e.g.,
On the other hand, our findings were mixed with respect to the age of the youngest victim. For example, results from Study 4 indicated that men who experienced CSV offended against significantly younger child victims, whereas results from Study 3 indicated that men who experienced CSV offended against older victims, but this finding was not significant. Furthermore, even though Studies 3 and 4 demonstrated that men who experienced CSV were generally more likely to view child pornography, Studies 1 and 2 showed no difference between the two groups, which is inconsistent with past findings. However, it is important to note that these relationships were not significant. Not only is viewing child pornography a criminal offence, it is also a social taboo; thus, some participants may have been fearful to disclose this information.
These mixed findings could also reflect the differences in our samples and settings. For example, Studies 1 and 2 participants were recruited both from prisons and outpatient services in Western Canada, whereas participants in other samples were all incarcerated at the time of these studies. Therefore, it is possible that the former samples may have been less comfortable disclosing their offences against children and other victim information (as these are more socially unacceptable crimes) and may have withheld their offence information due to fear of potential legal repercussions (e.g., jail time), particularly those who were free in the community. Additionally, it is also possible that some discrepant findings across our studies resulted from the differing risk levels among our samples. For example, the average SIR-R1 scores in Studies 3 and 4 were 14.30 and 5.42, respectively. This substantial variance in level of general risk between the two studies could account for the fact that Study 3 results indicated a negative relationship between CSV and the number of charges for sexual offences against victims under 12, whereas Study 4 results indicated a significant and positive relationship between these two variables. In fact,
The second goal of this paper was to examine the relationship between CSV and indicators of antisocial orientation. Consistent with past research (e.g.,
Interestingly, we found that CSV was significantly and positively associated with criminality/antisocial lifestyle, as assessed by the CATS in Studies 1 and 2. When psychopathy was assessed using the SRP in Study 2, findings indicated that CSV was related to criminality/antisocial lifestyle, but negatively associated with interpersonal and affective psychopathy traits, though these results did not approach significance. These results are consistent with some past research (e.g.,
From a neurobiological perspective, CSV has been found to be associated with long-term dysfunction in emotional and behavioral outcomes (e.g.,
We conducted a meta-analysis of the observed group differences to quantitatively synthesize the findings across our four studies. As expected, we found that men who experienced CSV had more antisocial features than those who had not experienced CSV. Further, though the average difference in pedophilic interests did not reach statistical significance, it was in the expected direction; that is, men convicted of sexual offences against children who experienced CSV had more pedophilic interests than those who had not experienced such abuse. Overall, our meta-analytic results were consistent with the broader literature.
Some limitations in our studies should be noted. First, as mentioned above, most of our sample sizes were quite small, which, not only affects our ability to generalize our findings, but also limits our statistical power to conduct more complex and advanced statistical analyses. Certainly, relationships between psychologically meaningful constructs are often influenced by additional/confounding variables; thus, the relationship between experiencing CSV and endorsing indicators of pedophilic interest and antisocial orientation could be further probed for the influence of additional variables, such as risk level, to more fully understand the path through which CSV may lead to pedophilic interest and antisocial orientation. However, we did not have enough power to conduct more sophisticated analyses, such as moderated regression; therefore, future research should further investigate how CSV may be related to pedophilic interest and antisocial orientation, and whether other variables may have an impact on this relationship.
Second, most variables included in Studies 2-4 were self-report (e.g., number of charges against victims and males under the age of 12, child pornography use, prior violent convictions, conditional release violations, etc.) and some would argue that the information may not be as reliable as information from official documentation. Relatedly, many of these self-report variables were based on single items with untested validity. Indeed, some researchers have raised concerns about the accuracy of self-report data (e.g.,
A third limitation is that the definition of CSV in Study 1 includes exposure to sexual stimuli. However, research shows that children under the age of 13 may be watching or seeking out pornography of their own accord (
Finally, all of our studies were cross-sectional; therefore, we were unable to determine a causal relationship, or even the potential direction of influence, between variables. Using non-experimental longitudinal designs and statistically controlling for potential confounds would provide more convincing circumstantial evidence regarding the potential causal effects of CSV. Additionally, though the purpose of our meta-analysis was to synthesize findings across our four studies, the number of studies was rather small (
Notwithstanding these limitations, the results from the current paper and past research generally suggest that CSV may be linked to greater sexual interest in children and greater antisociality among convicted sexual offenders, which are both important predictors of sexual offending against children. Future research should use more rigorous methodology to further examine the extent to which CSV impacts pedophilic interest, antisociality, and sexual offending, as well as whether pedophilic interest and antisociality mediate the relationship between CSV and sexual offending.
We are grateful to Kelly Babchishin, Renee Malcom, Ian McPhail, Cathrine Pettersen, Rikki Sewell, and Nicole Kostiuk for collecting the data used in the current studies. We gratefully acknowledge the support provided by the Northern Alberta Forensic Psychiatry Services, and the Faculty of Arts & Science and the Research Office of MacEwan University.
The data for this research study will not be available to other researchers.
The authors have no funding to report.
The authors have declared that no competing interests exist.