Today the Supreme Court of the United States ruled that, in death penalty cases, the state of Florida must take into account the inherent imprecision of IQ tests.
Why are IQ tests used in death penalty cases? It is unconstitutional to execute a person deemed to be intellectually disabled (Intellectual disability is the current term for what was previously known as mental retardation.). Diagnosing intellectual disabilities is a complex matter but the diagnosis hinges to a large degree on the person’s performance on a well-constructed IQ test. Although high-quality IQ tests are more reliable than most psychological measures, even the best IQ tests are imperfectly precise. There is a potentially large risk that a person with an observed score slightly above the threshold set by Florida law may have a “true score” that is below the threshold.
It was an unexpected honor to have my work cited in both the court’s decision (written by Justice Kennedy) and the dissenting opinion (written by Justice Alito). My contribution to the argument (relevant portion reproduced here) is a technical one and played an admittedly small role in the proceedings . My main point was that when multiple IQ tests have been administered to the same individual, we should not average the scores but make them into a composite score in the same way that we combine psychological scores in any other context. Doing so gives a more accurate estimate of the IQ and a smaller confidence interval around the score. I hope that the application of this procedure results in fewer incorrect decisions and a fairer administration of justice.
I am grateful to Cecil Reynolds for giving me the opportunity to write the paper and to Kevin McGrew for encouraging me to re-write and publish the argument on the web, demonstrating its application to death penalty cases. Although it was the published chapter that was cited and used by the defense, it was the free web version that initially caught the attention of the law firm representing the defendant.