posted 10-27-2008 10:20 PM            

5 Principles for RQs
http://www.raymondnelson.us/mm/PCSOT_RQ_principles_October2008_Utah.html

7 Rules for RQs
http://www.raymondnelson.us/mm/PCSOT_Relevant_Questions_7_rules.html

and

3 Rules for what Mr. Holden calls a "valid testing model"
http://www.raymondnelson.us/mm/PCSOT_Valid_Testing_Model-3_Big_Rules_slides_136-138.ht ml

r

------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)

posted 10-28-2008 09:39 AM            

I just read an article in APA magazine that discussed "Special" requirements and certifications for examiners running Post Conviction Tests for Domestic Violence Offenders. I can certainly make a convincing argument that almost every type of crime has significant differences from every other type of crime. (That’s why statute books are so thick). But even with those differences field examiners manage to muddle through the process of constructing exams based on the issues to be resolved, which are mostly provided by the person requesting the results.
Is specialization necessary?
Where will this “specialization” end?

I may be taking a posture here that is over simplistic but here goes.

If I am aware of a specific allegation supported by information or suspicion that something has actually occurred, I run a specific issue format. If I don't have a specific issue then I run some type of screening format which is often a mixed issue or multiple issue test. In the context of PCSOT the mixed issues tests are restricted to either maintenance issues with similar consequences involving treatment or monitoring issues with similar consequences involving probation or parole violations. I understand the thought process that decided that the relevant questions should have similar consequences although I am not sure that "validity" is the appropriate explanation as to why we should do it that way.

When I run a pre-employment test, it is always a multiple issue test and while the "face" consequences of failing appear the same i.e. "not getting the job"; the real underlying consequences run the gamut from not getting hired to going to prison. When I stop to consider the underlying consequences of either maintenance or monitoring exams they are very similar to pre-employment tests as they also run the gamut from reprimand to prison.

In all of these tests isn't it the examinee that actually assigns the significance of each stimulus?

Other than the utility of being able to tell a treatment provider or P.O., with some conviction that significant reactions on the exam pertain to either maintenance or a monitoring issue why can’t we mix maintenance and monitoring issues on a single test?

Either way, absent an admission, shouldn’t we be running breakdown exams to identify the cause of the reaction anyway?

Does anyone know of anybody getting busted out of treatment or sent back to prison based ONLY on the examiner’s opinion that the examinees showed significant reaction to either non-specific “maintenance” OR “monitoring” issues?

posted 10-28-2008 11:22 AM            

Thanks for responding. Most folks are either too confused, too apathetic, or overly confident in their fearless-leader to carefully examine the content of the PCSOT proposals. Or perhaps they are scared of losing their good-housekeeping-seal-of-approval.

How's the new 'puter working?

Your question...

quote:

---

Other than the utility of being able to tell a treatment provider or P.O., with some conviction that significant reactions on the exam pertain to either maintenance or a monitoring issue why can’t we mix maintenance and monitoring issues on a single test?

---

Good question.

However, point of clarification:

In the present proposed model policy, "monitoring" refers to test questions regarding new offenses while under community supervision and in treatment. "Maintenance" refers to test questions pertaining to non-compliance with rules.

There is arguable face-validity to the separation of re-offense concerns from non-compliance concerns.

Reoffending is presumably a lower base-rate concern than non-compliance with rules. Also, consequences for reoffending are potentially very different from the consequences for non-compliance with rules. Reoffending is a go-directly-to-jail-without-passing-go kind of issue, while non-compliance issues may be more correctable. Because of their presumed lower, or different, base-rates, error estimation will differ for re-offense questions - with a potential increase in the proportion of false-positive errors among persons would produce "significant reactions."

Obviously, re-offense rates differ for different typologies of offenders. Psychopaths and other high-risk offenders will re-offend at higher base-rates. Sexual compulsives may also reoffend more frequently, and we need to treat compulsivity behaviors carefully (e.g., peeping, exhibitionism, underwear theft, public masturbation).

I assume you are referring to the artificial vector of separation between Maintenance-treatment and Maintenance-probation questions.

Elizabeth, at APA, conceded to the large room full of examiners that their is overlapping content in probation and treatment contracts. Therefore, any argument that these represent "distinct" concerns - as the present training materials indicate - will require proof.

I predict there will be no proof of validity for the notion of some vector of distinct separation between superivision and treatment rules or an offender's non-compliance therewith.

It makes no sense for the APA to etch this in stone - except that it fits someone's fanciful mental model. The separation of non-compliance concerns to probation and treatment rules is not based in science, and lacks face validity. The committee members have already admitted the content overlap.

.012

r

------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)

posted 10-28-2008 11:40 AM         

The more "rules" you place on a testing format, the more training required to understand, the more money made giving training. Why not stick to the proven methods of testing? Mixed issues exams have limited accuracy, I believe we all agree on this. The percentages I have heard are 86% if the examiner is well trained and conducting the exam properly. ( and that's the high %)

Break out exams should be mandatory, you increase accuracy and limit the scope of the exam. You already have the required training.

The one portion of PCSOT testing I see requiring more training is on sexual offenders thought processes and how this relates to violation of treatment rules and re offense.

I no longer do PCSOT testing at all. I can see in the distance a court case going to the supreme court regarding a false positive and some one loosing freedom as a result.

We need more studies using current cases that are verifiable. I'm not a research person so I will leave that to the experts in that area. More "Rules" will not solve the basic problems, more "Research" will help.

posted 10-28-2008 09:06 PM            

If the PO says "you are in violation if you are no complying with treatment rules" Then
treatment says "if you are not obeying your rules of probation, then you are in violation and I am kicking you out of treatment." Doesn't that make probation and treatment pretty darn similar?

Quite a few offenders go to jail or prison for getting kicked out of treatment, since that's a violation of probation.

Since no one would want a poor sex offender to have to pay for a treatment exam and then a probation exam these issues are getting mixed all the time.

The Washington DOC has a new rule that started this year; a Probation Officer has to be present when any offender on probation is polygraphed! I pretty much do not need an office anymore, but the DOC schedules are full more often and it cut down on the examiner shopping. But, Treatment providers are not asking for polys or sending questions to the PO to be asked,

sorry if I started rambling and lost my train of thought,

posted 10-28-2008 10:06 PM            

Even if there is face-validity to the untested hypothesis that monitoring/reoffense behaviors represent a different level of potential consequences, compared to maintenance non-compliance behaviors, a review of conviction numbers will reveal that most offenders who removed from the community and sent to prison are resentenced for technical violations involving violation of their probation rules (not reoffending) or their treatment contracts.

Don't think for a minute that offenders get warm milk and cookies for admitting treatment rule violations.

Get canned from treatment and you go to a more restrictive program.

That's the way it should be. They don't follow the rules... and we send them to their rooms for as many years as we can.

Probation and treatment go hand-in-hand. That's what containment is.

Keep in mind that many professionals have a tendency to be lazy, and to drift away from the ugliest details. Offenders love that. Offenders also love to compartmentalize people's roles. They prefer to deal with one professional in one way, and other professionals in other ways. Its how they manage their deviant worlds. The goal of containment is to disrupt that pattern, through cross-training and shared agendas.

Years ago, and offender would have refused to talk with the PO about fantasies and deviant impulses - just too embarasing, and too personal to discuss with anyone outside of the black-box of confidential therapy. Compliance with rules was not the concern of therapists - whose bleeding-heart value system made them an easy mark.

In the containment model, therapists are not permitted to pretend they are engaging in productive therapy when offenders won't even follow the rules. Try treating someone's drug problem while they are still using. Following probation rules is a central concern of treatment. BTW we don't really pretend we can fix them, but it's helpful to have clinical expertise involved so that we can understand them and get close to them. That's the rub, getting close to offenders is ugly bid'ness and therapists don't really want to do it. Offenders would like for one professional to be closer to them than the others, because that improves their ability to split he containment team and act out of bounds. So, a goal of containmen, in addition to cross-training, is that the team members all need to be more closely connected to each other than any of the team members are with the offender.

Containment is cross-training. Containment is multi-systemic.

It will be a substantial step backward to impose a false separation of treatment and probation rule compliance.

.012

r

------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)

posted 10-29-2008 11:13 AM            

quote:

---

There is arguable face-validity to the separation of re-offense concerns from non-compliance concerns

---

Maybe so but once "overlap" is acknowledged it becomes a pretty short argument.

I would go one step further in the discussion of overlap. It seems logical to me that re-offense will also overlap both probation compliance and treatment compliance issues, assuming, of course, the rules have been written sufficiently constraining to provide containment. If the rules are properly written, the offender should not be capable of re-offending without violating either his probation rules or treatment rules. Once you acknowledge that overlap occurs, you can no longer argue that a test format has the ability to discriminate between reactions caused by re-offense issues and reactions caused by compliance issues. Arguing the need to separate those types of questions into two different exams becomes moot.Life was alot easier when we thought outside issue questions were effective.

If the most significant reactions occur at questions involving re-offense, then maintenance violations have probably occurred. If the most significant reactions occur at questions involving compliance with probation or treatment rules, then re-offense has possibly occurred. We’re casting a pretty wide net; we need to figure out how to make the best use out of whatever we catch.

I’m thinking that a format more similar to the TES where re-offense is addressed on the first series and compliance is addressed on the second series makes more sense than running compliance and re-offense tests several months apart because it allows us the broader range of specific issue tests for follow up.

I'm just riffing on PCSOT and this is where my thoughts went this am.

------------------
Ex scientia veritas

ps the new 'puter looks like it's working fine. We held a vote and democratically decided that my wife would use it for college and I wouldn't complain.
ebv