A Widely Used Criminal Justice Algorithm For Assessing Child Pornography Recidivism Is Flawed

Source: scientificamerican.com 5/20/24

The CPORT algorithm, commonly used to estimate the risk that a child pornography offender will offend again, hasn’t been validated for use in the U.S.

In today’s criminal justice system, there are more than 400 algorithms on the market that inform important legal decisions like sentencing and parole. Much like insurance companies use algorithms to set premiums, judges use risk assessment algorithms to estimate the likelihood someone will become a repeat offender when they render prison sentences. Generally speaking, lower-risk offenders can and do receive shorter prison sentences than higher-risk offenders.

Scientists and legal advocates have criticized the use of these algorithms as racially biased, opaque in how they operate and too generic for a criminal justice system that is supposed to treat everyone individually. Yet few people are paying attention to how these algorithms get this way—how they are being developed and validated before use. In the case of child pornography offenders, one algorithm is widely used by psychological experts in the criminal justice system with little thought to its development and, more importantly, its accuracy. The use of an unvalidated algorithm with unknown accuracy is dangerous, given the serious consequences associated with child pornography offenses.

The algorithm is called the Child Pornography Offender Risk Tool (CPORT). The State of Georgia uses the CPORT to determine which convicted sexual offenders should be placed on the public sexual offender registry, and experts commonly testify at sentencing hearings across the country about the results of the CPORT risk assessment. One might assume there is robust scientific evidence validating the CPORT on offenders in the United States. That assumption is incorrect.

Last year, we published a detailed methodological critique of the CPORT. Among other things, we noted that the sample used to develop the instrument was extremely small. The CPORT was developed by studying 266 child pornography offenders from Ontario, Canada, who were released from custody between 1993 and 2006. Within five years of release, 29 of the offenders were charged or convicted of a new sexual offense.

Developing an algorithm based on 29 recidivists is troubling because small sample sizes make statistical models unstable and not generalizable to the broader population of child pornography offenders. Other well-known risk factors, such as access to children or preoccupation with child pornography, were not predictive risk factors in this sample and thus were not included in the CPORT.

What’s more, the development data for the CPORT are potentially outdated given the enormous differences in technology that are used to access, store, and transmit child pornography since 2006—when the CPORT developmental sample was collected. Cell phones and other Internet technology did not come into widespread use until after 2006, significantly changing and expanding the way online child pornography offenses occur. Access to the internet is a common characteristic of child pornography offenders, but it is not included in the CPORT.

Read the full article

 

Related posts

Subscribe
Notify of

We welcome a lively discussion with all view points - keeping in mind...

 

  1. Submissions must be in English
  2. Your submission will be reviewed by one of our volunteer moderators. Moderating decisions may be subjective.
  3. Please keep the tone of your comment civil and courteous. This is a public forum.
  4. Swear words should be starred out such as f*k and s*t and a**
  5. Please avoid the use of derogatory labels.  Always use person-first language.
  6. Please stay on topic - both in terms of the organization in general and this post in particular.
  7. Please refrain from general political statements in (dis)favor of one of the major parties or their representatives.
  8. Please take personal conversations off this forum.
  9. We will not publish any comments advocating for violent or any illegal action.
  10. We cannot connect participants privately - feel free to leave your contact info here. You may want to create a new / free, readily available email address that are not personally identifiable.
  11. Please refrain from copying and pasting repetitive and lengthy amounts of text.
  12. Please do not post in all Caps.
  13. If you wish to link to a serious and relevant media article, legitimate advocacy group or other pertinent web site / document, please provide the full link. No abbreviated / obfuscated links. Posts that include a URL may take considerably longer to be approved.
  14. We suggest to compose lengthy comments in a desktop text editor and copy and paste them into the comment form
  15. We will not publish any posts containing any names not mentioned in the original article.
  16. Please choose a short user name that does not contain links to other web sites or identify real people.  Do not use your real name.
  17. Please do not solicit funds
  18. No discussions about weapons
  19. If you use any abbreviation such as Failure To Register (FTR), Person Forced to Register (PFR) or any others, the first time you use it in a thread, please expand it for new people to better understand.
  20. All commenters are required to provide a real email address where we can contact them.  It will not be displayed on the site.
  21. Please send any input regarding moderation or other website issues via email to moderator [at] all4consolaws [dot] org
  22. We no longer post articles about arrests or accusations, only selected convictions. If your comment contains a link to an arrest or accusation article we will not approve your comment.
  23. If addressing another commenter, please address them by exactly their full display name, do not modify their name. 
ACSOL, including but not limited to its board members and agents, does not provide legal advice on this website.  In addition, ACSOL warns that those who provide comments on this website may or may not be legal professionals on whose advice one can reasonably rely.  
 

32 Comments
Inline Feedbacks
View all comments

I wonder if any of our fine legislators are even familiar with a pub like Scientific American. Or maybe they are and just don’t care about data, science, and facts.

29 people actually committed a new sex offense, or a new “sex offense” because they brock some supervision rule and any violation is counted as a “sex offense”?

First, and foremost, another study outta the Great White North that misled and misleads people, much like the -99R test we all dislike. While they do have some refreshing takes on the registry the US could learn from, they are not doing their best in areas like this.

The ones who should be challenging the use of this test is the legal defense community while on retainer in their capacity. While in court, they need to challenge it’s use as well as challenge it on its own outside of a defense capacity to rid the system of using it. The blatant disregard of no outside endorsement of it should outlaw its use, but the same is said for the polygraph as well (where both are believed in by their own people and only their own people, much like the -99R).

With that being said, COTS (commercial off the shelf) is a preferred way to spend gov’t money on something to avoid all sorts of development time and effort, but is NOT always the way to a better solution, which is usually only found out well after the money has been spent on it and it is too late.

However, until something better comes along, which should be proven independently, those who wants this will use it until told otherwise, sadly.

Everytime I read about “predictive algorithms” I immediately get a “scifi” sort of vibe, and start thinking about “The Minority Report” with Tom Cruise.

The precise methods these “predictive” systems utilize seem intentionally vague and obscured in a sort of “arcane mystique” only available to a “chosen few”, resembling something more akin to pure fantasy, (or, perhaps, a cabal of “technomancers” are divining chicken bones while consulting a triad of genetically enhanced psychics who are all connected to Hal 9000 and Deep Thought)…maybe “sci-fantasy” is a more accurate description.

Whatever they’re trying to sell, it’s ultimately just junk-science (or just plain junk, since there’s really no science to speak of) and propaganda to generate more profit and impose increasing control over a select segment of the population…just like the polygraph, “eye detect”, “static-99r”, and every other ridiculous “predictive” device or system they come up with.

How do we know they didn’t just wait until they had 29 people that had red offenses before they finally released report. Like saying you caught 5 fish when in fact you dropped your hook 200 times in the water. You can make any study say anything you want .

Whatever validity this tool may have had is outdated. The article makes that clear, in the final paragraphs that discuss the relationship between CP and the internet.

In 1993, physical media, (printed photos, VHS/DVD/Film, paper Magazines) where about the only kind of CP that existed. Thus the people in the study where the kind of people that actually sought this stuff out. People that were willing to go out looking for this stuff, risk getting caught trying to acquire it, and then keep it in their possession long enough to get caught with it. This might explain why the study group was so small, how many people like that are going to be available to study at any given moment? I’m actually surprised the study group was as large as it was.

There are several of us in this forum that we convicted for Non-Prod CP, me being one of them. How many in this group were caught with physical media CP? How many even attempted to acquire Physical media? I know I never did. Every single image, both still and video, was quickly, easily, and freely acquired from P2P networks right in the comfort of my own home.

So comparing me to some Canuck that searched the frozen tundras of Canadia for magazines/DVDs back in ’93 is not valid. Not once did I ever even consider leaving the house, risk getting caught trying to pick up the stuff from my connection, nor spending a single penny on anything. These folks would have had to put serious time, money and effort into building up their collections, effort I was not willing to make! Unlike them, getting this stuff wasn’t important enough to me to make any real effort. With me, If I did a search that didn’t immediately produce the results I was looking for, I was likely to just give up and walk away.

Thus, it’s not all that surprising that slightly better than 10% of these folks continued to pursue interest in this. These folks were dedicated to this pursuit, where for me this was more of an occasional hobby that I sometimes engaged in before I returned to my couch to drink until I passed out. As such…

An AlsoME/Canuck comparison isn’t Apples to Apples, nor in fairness is it apples to oranges, more like Apples to Pears. While similar in may ways, there are some critically distinct differences that render the comparison meaningless.

Looks like the Japanese Canadian sexologist also has his paws on CPORT. But I really can’t blame Michael Seto for American criminologists & law enforcement putting too much weight in his studies. I’ve even seen his research cited on residency ordinance statues that claim higher risk levels amongst those that commit sex crimes. Maybe Seto, and his buddy Karl Hanson, should come down from Ottawa to Georgia and straighten these people out. But then again, those Georgia state bureaucrats & LEOs may hurt them…