May 22, 2011

How to Lie with Statistics for Church Money



I've been following more of the coverage on the John Jay Causes and Contexts study commissioned by the US Conference of Catholic Bishops. I am definitely not the only one having a huge problem with what reads more like apologia for the Church than a serious study of their sexual abuse problem.

An excellent column by Debra W. Haffner in the Washington Post points to some of the excruciating illogic.

It is nearly silent about the abusing priests’ flagrant disregard for the church’s teachings on sexuality and sexual behaviors, while condemning the culture of the 1960’s and 1970’s for these crimes against children. The extensive sections about other agencies serving youth and other denominations facing similar problems, and the concluding paragraphs that these others should take steps to prevent child sexual abuse seem to obviate at least in part what I wished would have been an urgent call for reform, lamentation, and restitution by the Catholic Church.

I can’t help but wonder if a single sexologist was asked to read and comment on the report before it was published. Although I was pleased to see that the writers directly address the fact that there is no evidence that gay priests are any more likely to abuse children than heterosexual priests, it was odd that they didn’t call then for the Catholic Church to end its proposed ban on gay seminarians. Further, the writers (inaccurately) define pedophilia as sex with children ten and younger--and then criticize the media for talking about “pedophile priests” when 22 percent of the victims--nearly one in four! --were these ages. Using the more accepted definition for pedophilia, their own data reveal that 73 percent of the victims were under the age of 14. These children and early adolescents were not capable of consent, regardless of the ages used.

Haffner's attempt to compile saner, more fairly representative statistical groupings poked at something that has been bothering me. As I said here, there is a murkiness to their age classifications and they're presented inconsistently within the study itself.

For one behavioral analysis they make age twelve the line of demarcation between pedophile and what they term ephebophile. From page 34:

For the purpose of this comparison, a pedophile is defined as a priest who had more than one victim, with all victims being age eleven or younger at the time of the offense. 164 An ephebophile is defined as a priest who abused more than one victim, with all victims being boys above the age of twelve.165

The endnotes do little to clarify these classifications.

164  Pedophilia is a diagnosable disorder in the Diagnostic and Statistical Manual of Mental Disorders (discussed at length in Chapter 3). However, these data are based on the behavior exhibited by priests rather than diagnoses. The behaviors are consistent with that which would be exhibited by an individual diagnosed with this disorder.

So endnote 164 cites the DSM but the current edition, the DSM-IV-TR, doesn't make the age cut-off for pedophilia eleven. It says thirteen. It is appallingly poor scholarship and downright disingenuous to make an assertion based on a source and misrepresent what that source actually says:

Over a period of at least 6 months, recurrent, intense sexually arousing fantasies, sexual urges, or behaviors involving sexual activity with a prepubescent child or children (generally age 13 years or younger).

As Haffner points out in her column, that age ballpark of thirteen is also the clinical standard. But it's not just that Karen J. Terry, et al., used a standard different than what the APA and others use. It's that they've simply made up a standard and cite nothing to justify it.

Again, assigning a specific age is always going to be a ballpark because the onset of puberty varies. But the authors seem to be deliberately skewing for the youngest possible cutoff. This is particularly troubling because most of the victims in this case are boys and boys typically hit puberty later than girls  do, so the likelihood that most of the victims were prepubescent is quite high.

The second endnote supporting their classification of ephebophilia is even more confusing; one might even say convoluted.

165  Ephebophilia, also sometimes called hebephilia, has been variously defined as follows: R. Blanchard, A.D. Lykins, D. Wherrett, M.E. Kuban, J.M. Cantor, T. Blak, “Pedophilia,  Hebephilia, and the DSM-V.” Archives of Sexual Behavior (2008): hebephilia is the erotic preference for pubescent   children (roughly, ages eleven or twelve to fourteen) and  ephebophilia is the erotic preference for adolescents (usually fifteen- to sixteen-year-olds); P. Cimbolic and P. Cantor, “Looking at Ephebophilia through the Lens of Cleric Sexual Abuse.” Sexual Addiction and Compulsivity, 13, 347-359 (2006) (in reference to clergy abuse): Ephebophilia refers to the persistent sexual attraction to pubescent or post-pubescent boys; J. Nunez, “Outpatient Treatment of the Sexually Compulsive Ephebophile.” Sexual Addiction & Compulsivity, 10:23-51 (2003): Ephebophile refers to adult men attracted to teenage boys; Goldberg (1992): Hebephile is a person who is sexually attracted to adolescents; K.V. Lanning, Child Molesters: A Behavioral Analysis. Alexandria, VA: Center for Missing and Exploited Children (1992): Hebephile and ephebophile describe an adult that has sexual attraction toward pubertal children; T. Rahman, “Ephebophilia: The Case for the Use of a New Word.” Forum for Modern Language Studies, 24(2), 126-141 (1988): refers to male sexual interest in boys and youths; K. Freund, H. Scher, S. Chan, and M. Ben-Aron, “Experimental Analysis of Pedophilia.” Behavior Research and Therapy, 20(2):105-112 (1982): Hebephilia is the sexual attraction to pubescent girls; ephebophilia is the sexual attraction to pubescent boys. The prevailing definition of ephebophilia today is of adolescent male victims, which is what we use as the definition here.

So that's about as clear as mud and does nothing to explain how they arrived at age twelve, nor does it explain why they later revise the ages.

It is hard for me to take their age specifics seriously when it is so glaringly apparent that their albeit obliquely stated purpose is to refute the public perception that the church has a pedophilia problem. Why else would a "scholarly" research paper repeatedly refer to unspecified "media" reports as counterpoint to their statistics? One such reference is on page 53:

Media reports about Catholic priests who sexually abused minors often mistakenly have referred to priests as pedophiles. According to the DSM IV-TR, pedophilia is characterized by fantasies, urges, or behaviors about sexual activity with a prepubescent child that occurs for a significant period of time. Yet, the Nature and Scope data indicated that nearly four out of five minors abused were at least eleven years old at the time of the abuse. Though development happens at varying ages for children, the literature generally refers to eleven and older as an age of pubescence or postpubescence.

What literature it is that places the onset of puberty at eleven, the authors do not say. There is no endnote or other citation for that statement. But, as quoted above, the DSM-IV-TR they specifically do cite in that passage states a ballpark age of thirteen to define pedophiliac attraction.

Further, I would defy the authors to find a source that puts postpubescence at age eleven. Postpubescence is when the changes of puberty are complete, generally in the mid to late teens.

Wikipedia offers a fairly good overview:

Although there is a wide range of normal ages, girls typically begin the process of puberty at age 10 or 11; boys at age 12 or 13.[1][2] Girls usually complete puberty by ages 15–17,[2][3][4] while boys usually complete puberty by ages 16–18.[2][3][5] Any increase in height beyond the post-pubertal age is uncommon. Girls attain reproductive maturity about 4 years after the first physical changes of puberty appear.[4] In contrast, boys accelerate more slowly but continue to grow for about 6 years after the first visible pubertal changes.[6]

The onset of puberty, particularly in girls, has been rolling forward for decades.

For example, the average age of the onset of menstrual periods in girls was 15 in 1900. By the 1990s, this average had dropped to 12 and a half years of age.

It can be younger still. Some have tied this to growth hormones in meat and milk and to phytoestrogens in soy. Whatever the reasons, precocious puberty can start alarmingly young.

In girls, puberty usually starts around 11 years of age, but it may start as early as 6 or 7 years of age. In boys, puberty begins around 12 years as age, but may start as early as 9 years of age.

So is sex with a six year old girl pedophilia, you know, if she's started puberty? I'll look forward to later editions of Causes and Contexts for the answer to that.

So why did the authors move the goalpost mid paper? Was it just some of the general sloppiness that characterizes much of this work? That argument could definitely be made. Or was it a deliberate attempt to minimize the impact of disturbing statistics? I did a little more digging to see if I could find a different breakdown of age groups. They cite no source other than their previous paper for the statistical breakdown in the graph I posted previously.

Most sexual abuse victims of priests (51 percent) were between the ages of eleven and fourteen, while 27 percent were fifteen to seventeen, 16 percent were eight to ten, and nearly 6 percent were under age seven. Over 40 percent of all victims were males between the ages of eleven and fourteen. It is worth noting that while the media has consistently referred to priest-abusers as “pedophile priests,” pedophilia is defined as the sexual attraction to prepubescent children. Yet, the data on priests show that 22 percent of victims were age ten and under, while the majority of victims were pubescent or postpubescent. Figure 1.4 shows the overall gender and age distribution of the victims from the Nature and Scope data.

I had to do a bit of digging through the earlier research paper to find where they had represented that data because of the way the material is presented. I found two presentations of those age statistics both of which are slightly different. But I would invite anyone to download all of it, see if they can make heads or tales of it, and find the exact statistics referred to above. This is from the executive summary:

The largest group of alleged victims (50.9%) was between the ages of 11 and 14, 27.3% were 15-17, 16% were 8-10 and nearly 6% were under age 7. Overall, 81% of victims were male and 19% female. Male victims tended to be older than female victims. Over 40% of all victims were males between the ages of 11 and 14.

And this is the chart that breaks down individual ages from the section entitled Characteristics of children who alleged sexual abuse by Catholic priests:



Click to Enlarge



The numbers in all three sets of statistics are slightly different but very close. If I add them up based on the authors' breakdown in that chart, I come up with: 1-10 (22.6%), 11-14 (50.7%), and 15-17 (26.7%). But if I move eleven year olds back into the prepubescent category, as the authors suggest in their earlier classification, the picture and implication change dramatically: 1-11 (32.6%), 12-14 (40.7%), and 15-17 (26.7%). By that standard, the prepubuscent grouping is nearly one third of the overall number and is larger than the only age grouping that could arguably be considered postpubescent. But where it gets most interesting is when I tally the percentages based on that clinical standard of age thirteen: 1-13 (60.1%), 14 (13.2%), and 15-17 (26.7%).

So if we use the clinical standard, as defined by the DSM-IV-TR, a substantial majority of sex abuse victims is prepubescent. It gets much harder to argue that the "media" is wrong and this isn't a pedophilia scandal based on that statistical breakdown. Using their original classification of prebuscent children as eleven and under doesn't look so hot either. I would have an easier time believing that this was anything other than an attempt to obfuscate if they had cited any source at all to justify their use of ages 1-10 as their prepubescent grouping... but they don't.

Their poor scholarship only gives fodder to exactly the kinds of myths they claim to debunk elsewhere. Specifically, they argue that homosexuality is not the issue but the odious Bill Donohue of the Catholic League looked at the numbers and analysis provided in the paper and found otherwise.

Finally, the report says that 81 percent of the victims were male and 78 percent were postpubescent. Since 100 percent of the abusers were male, that's called homosexuality, not pedophilia or heterosexuality.

Donohue is incorrect about the percentage of postpubescent victims. The largest grouping of 11-14 is, at most, pubescent, but it's the authors themselves who conflate pubescent and postpubescent as I quoted above.

Haffner also makes the point that the authors contend this wasn't a pedophilia scandal because many of the abusive priests weren't strictly pedophilic in their attractions.

I wonder if anyone will be comforted that the John Jay writers say that priests who sexually abuse children are more accurately labeled “indiscriminate offenders” than pedophiles...

I'd be willing to bet that the abused children would not be comforted. In fact, I suspect they'd consider that distinction meaningless.

They break abusers down into "specialists" and "generalists" and, once again, blame society for this compulsive behavior. From page 119:

Priest-abusers were not “pedophile priests.” The majority of priests who abused were not driven by particular pathologies, and most did not “specialize” in abuse of particular types of victims. The pathologically driven priests were not influenced by social factors as were the majority of abusers (for example, their behavior was consistent across the time period and did not peak from the mid-1960s to 1980s). “Generalists,” or indiscriminate offenders, constituted the majority of abusers and were influenced by social factors.

Wow. You know, I think it can be fairly stated that promiscuity became more common and accepted in the '60s and '70s, but I think you'd be hard pressed to find a lot of aging hippies who'd tell you that it's perfectly normal to fuck prepubsescent children as long as you also fuck older kids and adults. Think about what the authors are saying here; that it is not pathological to molest prepubescent children as long as you also have an attraction to more age appropriate partners. That is literally what they're saying.

There's a bit more in their breakdown of "specialists" and "generalists," but frankly, it just made my head spin around like Linda Blair.



Click to Enlarge


As you can see, pedophilia is solidly ten and below in their new calculus. Ephebophilia is now boys of thirteen to seventeen. And we must assume that no "specialist" has any interest in children aged eleven or twelve. I'm sure all the eleven and twelve year olds who've been molested will be very relieved.

As I mentioned here, there is an ongoing debate over adding hebephilia to the DSM-V. It seems obvious to me that there is a difference between attraction to children who are in the midst of the hormonal, physical, and emotional changes of puberty and those who are genuinely postpubescent. If we're going to genuinely look at the biological changes associated with puberty, perhaps we should acknowledge that those changes go on for years. (In contrast, the authors of the John Jay study contend that children become sexually viable as soon as they've conceivably begun the changes of puberty.) In that sense, making the cut-off for pedohebephilia fourteen doesn't even begin to cover changes that make kids extremely vulnerable into their upper teens. In practice, it would add one year to the age range of pathological attraction.

A column in Psychology Today explores some of the controversy over acknowledging attraction to the pubescent as a pathology.

Hebephiles are not sexually typical men who just happen to go for easy-pickin's because they can't get them a real man or woman. Don't get me wrong: there are some otherwise-sexually typical men, primarily attracted to adults, who will sexually abuse pubescent children (including their own children and step-children) because they can't get enough of what they really want. But hebephiles are different. They show offense histories and laboratory arousal patterns indicating that their peak sexual arousal is to pubescent body types. In this way, they are discernable, as a group, from men who show offense histories and laboratory arousal patterns indicating peak sexual arousal patterns to prepubescent children or to adults.

A good deal of this research has been conducted by my colleague (and friend) Ray Blanchard of Canada's Center for Addiction and Mental Health and the University of Toronto. Using an extraordinary compilation of data from thousands of men, Blanchard has convincingly shown that hebephiles consist of a sort of "missing link" between pedophiles (those attracted to pre-pubescent children) and teleiophiles (those of us attracted to sexually mature people). Blanchard's data clearly indicate that the sexual orientation of males (at least) isn't just composed of the sex of their partners, it's also composed of the age of their partners.


Comments on this entry are closed, on this blog. If you wish to comment, please find this and all newer blog entries crossposted on Celestial Reflections.