Nextgenforensic at ATSA 2014: Ian McPhail
Below are links to a pair of presentation slides for talks that I gave at the 2014 Association for the Treatment of Sexual Abusers (ATSA) conference in San Diego. In this post, I wanted to briefly outline a few of the main points I think are most important to take away from these projects on pedophilia.
Pedophilia and Height
In this research, Dr. James Cantor and I set out to provide a meta-analytic estimate for the difference in height between pedophilic and non-pedophilic adult males. This idea arose after reading a recent article on the same topic that reached a conclusion that we did not expect (that pedophilic males are not shorter, on average). So we naturally wanted to summarize the available literature as a way of response to the recently published article by Dr. Sandy Jung and colleagues and to understand the pedophilia-height association better. The talk I gave at ATSA provides a brief summary of what we found in our meta-analysis. On a side note, the research is currently in press in the journal Deviant Behavior.
For a bit of a preamble, the association between height and pedophilia may seem to be an odd thing to be looking at and requires a bit of explanation. Height in adulthood is thought to be an indicator of certain experiences in utero or in early childhood; experiences that have the capacity to impact physical development (that is, these experiences are nontrivial to development). Specific events that are found to be associated with shorter stature in adulthood are things like poor prenatal nutrition, chronic childhood illness, and psychosocial stressors (for example, family tension, parental abandonment prior to age 7).
What Dr. Cantor and I found in our meta-analytic research is that the research on this topic does indeed show that adult pedophilic males are, on average, shorter than adult non-pedophilic males. This finding is consistent across studies, including the recent article that suggested the association between pedophilia and height was not present in their sample. On average, pedophilic males were shown to be 1.71cm shorter than non-pedophilic males in the published literature. This difference in height may not seem like much in absolute terms, however, given that height is almost completely determined by genetic factors, this difference in height seems to be a robust indicator of adverse experiences in utero or in childhood experienced by pedophilic males.
Slides can be downloaded here
A version of the in press article by Dr. Cantor and I can be downloaded here
Meta-analytic Research on Discriminative Validity of Phallometry
The second project that I presented is a different piece of meta-analytic research Stephanie Fernane, Chantal Hermann, Drs. Yolanda Fernandez, James Cantor, Kevin Nunes, and myself have been working together on. We examined the published literature (and had some datasets available to us as well, thanks to Sébastien Brouillette-Alarie, Jean Proulx, and Jan Looman) looking at the ability of different phallometric measures of pedophilic sexual interests to discriminate sex offenders with child victims from other groups.
I won’t go into great detail about phallometry and what it is, but the unfamiliar reader is encouraged to check out the Wikipedia page for phallometry (aka, penile plethsymography), since there is good introductory information about the procedure there. Further, I would also encourage any interested reader to look at critical reviews of the phallometric method, as I think the research I presented at ATSA 2014 will best viewed from a critical context about what we can say about what type of information phallometry is providing.
What we found in this research is that across a decently large amount of published literature, phallometry was consistently able to produce large differences between sexual offender against children samples and other groups (sex offenders with adult victims, nonsexual offenders, and nonoffenders). These findings indicate that sex offenders with child victims, on average, do indeed show greater sexual arousal to pedophilic stimuli and show less sexual arousal to non-deviant stimuli (that is, these offenders are less sexually aroused by adult females compared to other groups). This is what we would expect, since pedophilic sexual interests is assumed to be one of the main risk factors that would contribute to someone sexually offending against a child.
In an additional set of analyses, we looked at what the literature could tell us about the validity of phallometry in differentiating offenders who have offended against unrelated child victims from other groups. Indeed, we found that there was a fairly large and robust difference in pedophilic sexual interests on phallometric measures, indicating that sex offenders with unrelated child victims do indeed show greater sexual interest in children than comparison groups. One last group-based research question we were interested in was to see how pedophilic incest offenders are compared to other groups, since this is a sort of a half-answered question in the literature. What we found is that the amount of literature published that includes incest offender samples is limited and provides partial evidence that phallometric measures do distinguish incest offenders as having greater pedophilic interests compared to non-offenders.
In a final set of analyses, we wanted to see if we could identify phallometric procedures that were associated with stronger validity or weaker validity. The variety of phallometric procedures, or nonstandardization of phallometric procedures used in the field, has been one of the main criticism of the phallometric method. So we were interested if a meta-analytic approach could shed some initial light on which, if any, phallometric methods are associated with greater validity. These procedural differences are rather technical, so I won’t go into great detail here on the blog, but I will summarize the findings in the following way. The type of sexual activity being depicted in the stimuli used during phallometric testing does seem to be associated with the discriminative validity of the test. As well, how the data is treated also seems to be associated with discriminative validity of the test.
We do have more data to examine to provide a fuller picture of this literature, but this presentation is a good starting point for summarizing the phallometric literature.
Slides can be downloaded here
My contact information can be found on my academia.edu page if anyone has questions about these presentations.