nextgenforensic

The link between pedophilia and height

Ian McPhail

If I were to ask the question, “what are some markers of early development experiences that might be linked to the development of pedophilic interests?”  I doubt most people would suggest “physical height”.  But to the contrary, pedophilic sex offenders are, on average, 1.7cm shorter than other groups of people (for example, other sex offender groups, non-offenders).  A more recent study by Fazio and colleagues found pedophilic offenders to be 3.09cm shorter than non-pedophilic offenders.

That pedophilic men are on average shorter seems to be a replicated and consistent finding.  I will note, all of the studies to date have used samples of pedophilic sexual offenders, so the link between pedophilia and height in non-offending pedophiles is unknown.  This point is important because, as a recent post on this blog re-iterated, not all pedophiles commit sexual offences.  Also, I keep repeating “on average” because this link is based on group averages and does not mean any individual pedophile will be shorter than an individual without this interest or that a shorter individual is pedophilic.

So, why do researchers do studies on the height of pedophilic men?  The main reason is that shorter height in adulthood is an indicator of whether a group of people may have suffered neurodevelopmental impairments in the womb or environmental hardship in childhood.  For instance, males born to mothers who smoke during pregnancy tend to be about 1.1cm shorter or to mothers who drink during pregnancy are about 0.9cm shorter.

Shorter height in adulthood is an indicator of whether a group of people may have suffered neurodevelopmental impairments in the womb or environmental hardship in childhood

As Fazio and colleagues suggest, “conditions affecting height pre- or peri-natally may heighten the likelihood of developing and/or expressing pedophilia.”  The finding that, on average, pedophilic men are shorter provides some indirect evidence that pedophilia may have a neurodevelopmental basis and/or there have been hardships in their early childhood.

Methodology matters

Another recent study by Levenson and Ackerman also examined the link between pedophilia and height.  They found, in a large sample of 22,228 sexual offenders, that men with victims under the age of 13 years are around 0.59cm shorter than men with victims over the age of 18.  This study adds another replication to the association between pedophilia and height.

As in all studies, there are some notable methodological limitations.  I made a closer inspection of the methods after reading that the study found pedophilic men to be 0.59cm shorter; a height difference about 1/3 to 1/5 of what other studies have found.  This suggested to me there may have been some features that served to ‘shrink’ the magnitude of the height difference.

First, care must be paid to how we identify pedophilic offenders in samples of sexual offenders.  Levenson and Ackerman used the age of sex offenders’ victims as a means of identifying who was pedophilic in their sample.  Pedophilic offenders were identified if they had victims under the age of 13 (that is, 12 years of age or younger).  The authors do acknowledge that victim age is an imperfect measure of pedophilic interests (indeed there is as of yet no one perfect measure).

What is a bit more of an issue is that when victim age is used to identify pedophilic interests, victims under the age of 11 years of age (that is, 10 years of age or younger) is an empirically supported cutoff for identifying pedophilic interests.  Using victim age of 12 years as the cutoff to identify pedophilic sexual offenders is problematic because this is over the age that the typical child starts puberty.  Sexual interest in early pubescent children is called hebephilia.  Similar to a large number of authors, Levenson and Ackerman used a higher age cutoff that likely resulted in categorizing non-pedophilic men as pedophilic (false positives).  The result is that these authors used an imperfect measure imperfectly.  The effect this had on their finding was likely to reduce the difference in height between pedophilic and other sex offenders.

Secondly, the method by which height was measured in the dataset used by Levenson was unknown.  This is because the data used by the authors was taken from public sex offender registries, so they had to make do with what they had.  No fault of their own, but this likely adds to the noise in their findings.  In contrast, in the Fazio study I mentioned above, the research team directly measured height.  When they compared measured height with self-reported height, men typically overestimated their height by 2 to 3½cm.

I wanted to comment on these limitations because I think they have some important ramifications for research that continues to examine the link between pedophilia and height and the factors that might contributed to this link.  The overall difference in height, across almost all studies published to date (the exception being the Fazio paper), is relatively small.  If care is not taken to measure the two variables under study, pedophilia and height, this leads to a reduction in an already small difference.  Levenson and Ackerman likely could detect differences because they were working with a sample of 7,500 ‘pedophilic’ offenders and 4,000 ‘non-pedophilic’ offenders.

Solid operationalizations of these basic constructs are going to be needed to detect effects reliably

If research follows the advice of Fazio and colleagues and begins to examine the neurodevelopmental and early environmental experiences linked with pedophilic interests, solid operationalizations of these basic constructs are going to be needed to detect effects reliably.  Using weak operationalizations may obscure important findings.

Accumulation

As is my empirical bent, I criticize with an eye to accumulating the available evidence.  To this end, I have taken the time to update the meta-analysis on pedophilia and height James Cantor and I did a few years back.  In the analysis, I have added in the Fazio and colleagues study and the Levenson and Ackerman study.  To highlight the impact of the Levenson study on reducing the overall effect size, I present the results of the meta-analysis both with and without this study included.  These results are presented to show how studies that have attenuated effect sizes, most likely due to (easily remedied) methodological limitations, can impact the accumulation of knowledge through meta-analysis.

Here are the results of the meta-analysis without the Levenson and Ackerman study:

wout-la

We can see that the overall effect size is about 0.21.  This means that, on average, pedophilic offenders are 0.21 standard deviations shorter than other groups.  To help understand this difference, the average height of American males aged 30–39 is 176.4cm.  The standard deviation around this mean is 11.08cm.  Applying our meta-analytic finding here, we would expect the average pedophilic offender to be 2.33cm shorter than the average American male.  Though this is the finding when all studies are combined, I would add that the Fazio study used the strongest methodology.  Future studies that look at this association are encouraged to reproduce the method of the Fazio study.

We would expect the average pedophilic offender to be 2.33cm shorter than the average American male

Here are the results of the meta-analysis with the Levenson and Ackerman study:

w-la

We can see, with the addition of Levenson and Ackerman’s findings, the overall effect size shrinks from 0.21 to 0.15.  Indeed, this study qualifies as an outlier.  Applying this to the example used above, if we relied on this estimate of the difference in height, we would expect the average pedophilic offender to be 1.66cm shorter than the average American male, a reduction of 0.67cm (from the 2.33cm difference calculated from the original meta-analysis).  Keeping in mind that although these differences in height seem small, some well-known experiences that impact adult height are not much greater than 0.67cm (e.g., males born to mothers who smoked rather heavily during pregnancy can be around 0.9cm shorter).

Where to go from here?

As it stands, there are six published studies that suggest pedophilic (offending) men are shorter than other groups, on average.  Applying this to less abstract and indirect topics, what might research look at that would account for a height difference of this magnitude?  What developmental or early childhood experiences might occur in pedophilic men’s lives that are powerful enough to affect their physical development to this degree?

Both Levenson and Ackerman and Fazio and colleagues make several interesting suggestions about next steps that I think bear repeating and consideration.

Levenson and Ackerman suggest that intergenerational transmission of trauma in utero might be one cluster of experiences that affects development for some in this population of men.  Childhood adversity, such as emotional neglect, childhood sexual abuse, or mental illness in the home, might be experiences that are elevated in pedophilic populations.  Fazio and colleagues cite research that suggests having a serious illness in childhood, parental separation, and maternal exposure to stressors are associated with height.

Taken together, the association between pedophilia and height and the ideas about sources of this link both suggest that developmental experiences, starting in utero, are worth exploring with samples of pedophilic men.

Suggested citation:
McPhail, I. V. (2017, January 8). The link between pedophilia and height [Weblog post]. Retrieved from https://nextgenforensic.wordpress.com/2017/01/08/the-link-between-pedophilia-and-height/


NextGenForensic welcomes analytical, thoughtful, cogent, relevant, respectful, and proportional comments, either supportive or opposing, on posts. Comments that are vague, off-topic, factually inaccurate, or include arguments replete with logical fallacies, will not be approved.

Want to submit your own post? Click here to find out how!

Advertisements
%d bloggers like this: