Re: re unrelated males in household

["John M Price PhD" <jmprice@xxxxxxxxxx>]

On Fri, 21 Jun 2002, Scott Linder wrote:

> I am not clear on what you mean by false positives. 

I read that as part of the problem of using these data for the development
of a predictive measure.  You will have as a result a high rate of
people/families where the predictive measure would indicate the kid(s) was
going to end up dead, but in real life they would not.  This would waste
resources at the least (not even to mention issues regarding the family's
privacy, etc.).

I am concerned also about a previous mention of the CPS worker's 'hunch',
and I do seem to recall that that very term was used.  Have we all
forgotten Meehl's research on the variability, indeed pretty much
uselessness, of professional opinion vis a vis actuarial analysis?  In
that regard, I see this effort to nail down predictors laudable, as long
as they are really tested psychometrically and end up being more than
simply surface valid.

 Remember, the subject 
> is in regards to children that have been violently killed.  You wrote:
> 
> >>In practical terms, the biggest problem, as has been pointed out by 
> >>several contributors, is the inevitable high rate of false positives when 
> >>the outcome variable has low incidence. <<

-- 
John M. Price, PhD                                     jmprice@xxxxxxxxxx
Life: Chemistry, but with feeling!      |      PGP Key on request or FTP!
Comoderator: sci.psychology.psychotherapy.moderated          Atheist# 683