Tuesday, December 1, 2009

Larry v Branker (NC, 2009). More voodoo psychometrics via SEM misuse/abuse and use of SSSQ

Another addition to the Atkins Court Decisions section of the ICDP blog. And, unfortunately, another case the reflects the courts failure to recognize professionally accepted psychometric standards (viz., failure to consider standard error of measurement--SEM; and the use of adaptive behavior instruments that are inappropriate for norm-referenced "relative standing" in a population comparisons).  Another case of voodoo psychometrics.

Thanks to Kevin Foley for sending the Larry v Branker (NC, 2009) decision.

Upon review, this appears to be yet another case where a bright line IQ score of 70 is used to render a Dx of MR.  That is....70 or below...DX=MR....71 to 75 = not MR.  The defendant obtained a Full Scale IQ score on the WAIS-III of 74, which is within the professionally accepted range for a Dx of MR (when taking into account "measurement error".)  Not taking into account the standard error of measurement (SEM) is at odds with the two major professional associations (AAMR/AAIDD; American Psychiatric Association DSM) that are always mentioned as providing the operational definition of MR in Atkins decisions.  APA and AAIDD (as well as all major measurement associations) make it extremely clear that SEM must be part of "best practices" when diagnosing MR.  The use of SEM is interpreted by professional associations, psychometric experts, and most psychological testing experts, as recognizing scores as high as 75 (70 plus or minus 5 IQ pts) in a possible Dx of MR. 

The continued failure of certain courts, judges, states, and some psychologist "experts" to recognize the well-established measurement concept of "measurement error" (as reflected by the SEM statistic) is troubling....very disturbing.  I will eventually be making a post re: the concept of SEM...in hopes of educating folks on the logic, statistics, and support for recognizing measurement error in MR Dx....esp. in Atkins cases.  The recent law review article by Blume is an excellent read regarding the failure of many courts to recognize the accepted practice of bracketing IQ scores with the SEM---and often in a manner that appears to be a deliberate attempt to circumvent the intention of the U.S. Supreme Courts original landmark Atkins v Virginia decision.  Some courts and judges continue to play foot-loose and fancy-free with psychometric concepts that they clearly fail to understand or, simply deliberately interpret inappropriately to justify a predetermined biased decision.  People with MR are continuing to be executed due to SEM misuse and abuse.   It appears the Larry Court hid behind statutory language: "The statute defines '[significantly subaverage general intellectual functioning' as '[a]n intelligence quotient of 70 or below.' Id" rather than employing reflective critical thinking in accordance with accepted professional standards. 

Also, as has been seen in other decisions,the SSSQ was inappropriately accepted as a measure of adaptive behavior.  Another example of voodoo psychometrics (click here for my prior comments on the use of the inappropriate use of the SSSQ in the formal Dx of MR. ) The SSSQ may be an appropriate instrument for which it was developed and/or normed, but is not an appropriate nationally normed measure of AB...which is required as per the Joint Test Standards when making inferences regarding a persons relative standing within the population.

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