If you have never heard of this diagnostic tool, it is a means of predicting recidivism by sex offenders. Unfortunately, it is a device with so little accuracy (particularly when applied to juveniles - the questions don't recognize age differences as to prior sexual experience) that it is almost useless. At the Legal Theory Blog, Solum links to an article on SSRN about just how bad the Static 99 is. Here is the abstract:
Sexually violent predator (SVP) laws are inherently suspicious because they incarcerate people not because of what they have done, but because of what they might do. The implicit assumption that sex offenders cannot control themselves is contradicted by recidivism data published by the U.S. Department of Justice in 2003. In practice the SVP laws violate due process by relying on an instrument - the Static 99 - that is so inaccurate that it condemns seven individuals for every one that would re-offend. Furthermore, the Static 99 fails to meet the constitutional criteria laid out by the U.S. Supreme Court in Kansas v Hendricks because it does not link an individual's mental illness to his dangerousness.
The problem with such instruments is that they are inherently invalid. What they are trying to to is determine whether or not a person is devient or a law breaker using a device that is based on actuarial science. I did research and wrote a paper on this type of instrument. It is neither valid nor relieable in dealing with specific individuals. In fact, it will be correct for specific persons less than 10% of the time. That would meen that over 90% of individuals would be improperly classified. If you desire further information, please contact me.
Posted by: Kenneth J. Bond | July 14, 2007 at 12:57 AM