Thursday, January 7, 2010

Dufour v Florida (2009) update: State still has its head in the sand re basic psychometrics (esp. SEM)

Update on Dufour v Florida (2009). 

ICDP's original post re: this case, including links to the court decisions, is available here.  Also, as previously posted, AAIDD submitted an Amicus Brief on behalf of Dufour.  And, I continue to articulate disbelief in the state of Florida's continued reliance on the Cherry court-based draw-a-line-in-the-sand at IQ=70 and don't cross it....period..and damn the standard error of measurement (SEM) approach to Atkins MR/ID death penalty cases also reflected in Brown v FL.

The defense subsequently submitted a response brief (click here) that focused on a number of arguments, with the bright-line cut-off score of 70 and failure to recognize the standard error of measurement (SEM) being what has concerned me the most.  And yes...there are the usual cast of issues being argued by the state and defense in this case (evidence of low or high adaptive functioning; malingering or not; dueling expert witness credibility, etc.).  But as I stated in my original post:
"Although I've not completed an exhaustive and critical analysis of all documents, it continues to concern me that in this case, as in other Florida cases, the state appears to continue to ignore the concept of the standard error of measurement (SEM). Cheery v Florida( 2007) continues to rear its head in many of these decisions. It appears the state of Florida continues to rely on the Cherry court to establish an absolute IQ score of 70 or below...period. The state of Florida apparently does not understand, or does not want to accept, the concept of SEM."
I was flabbergasted to read  the prosecutions answer (click here) for ignoring the SEM.  The brief stated "the standard error of measurement is not a criteria for diagnosis, but, a 'statistical term.' " A statistical term!!!  Even after admitting that experts from all sides accept and recognize the professional and scientific concept of SEM in psychological measurement, the primary argument is that is is just a statistical term....what kind of logic or argument is that?  It is not an argument. It is based on no logic.  To me this statement reflects a deliberate attempt to stick ones head in the sand when one knows they have lost an argument based on accepted science, professional standards, and simple logic.  It is analogous to arguing against the concept of "gravity" and saying it should not be recognized because it is "just a physics term."

Beyond my laser-like focus on the bright-line cut-score and SEM issues, a reading of these two new documents are a good education for the manner in which psychological information and data are torn apart and argued, especially in the acknowledged messy area of determining adaptive behavior functioning.

In summary, first there was the inability of the state of Florida to accurately count votes in a presidential election (Gore v Bush).  Now we have the deliberate attempt to not recognize a very basic numerical/statistical concept (SEM).  What's up with the quantitative literacy at the highest levels of power in FL?  SEM is easy to understand and I have provided a very readable document that should make SEM understandable for even those -1 SD below the mean on quantitative literacy. 

Come on Florida..this is not that hard. 

Technorati Tags: , , , , , , , , , , , , , , , , , , , , ,