Subject: Supreme Court to review statistics for a death penalty case involving IQ tests and the SEM.
Briefly as possible, the decision to execute a prisoner depends on the SCOTUS assessing if the lower bound of the smallest of five "IQ scores adjusted by the standard error of the mean" is sufficient to declare the defendant 'intellectually disabled' and therefore execution is 'cruel and unusual punishment' and the defendant is "Ineligible for execution." Further complicating matters is that Justice Sotomayor seems to suggest that the defendant survived a previous attempt at execution (I haven't confirmed that)
The use and interpretation of statistics and the lower bound of an "SEM adjusted score" will be a deciding factor in this case. ---------------------- >>>>>>>First, I am not an attorney!!!!! <<<<<<< 😊
Below, I have liberally inserted cut-and-pasted comments from the news, along with some of my interpretation. I inserted a question mark ?. where I was unsure or unable to find the records for the case in Alabama, and the actual IQ score "data" for the case.
Possibly, an ASA member here can locate the case records in Alabama and
a) Track down
a. the name of the IQ test name (such as "Stanford Binet Intelligence Scale") ,
b. IQ test scores AND
c. the IQ score standard errors of the means, or "margin of error, " and the definitions /formulae used for the calculations of those quantities for the five IQ Tests.
b) d. locate the transcripts of the expert witness testimonies because one of the online sources suggests there were "dueling expert testimonies".
Statistical summary.
There is no dispute that Smith brutally murdered Hamm, and details are available online. The Reddit link below gives some of the gory details. There is no dispute that Smith was administered five "full-scale" IQ tests. In Alabama, the IQ score must be less than 70 for Smith to be considered intellectually disabled. According to the "Atkins" If Smith is determined to be intellectually disabled, then execution is considered cruel and unusual punishment (under the Eighth Amendment to the Constitution) , and he will not be executed. Not readily publicly available (or I can't find them) are the "standard error of the mean" or the "margin of error" for each IQ test. Of the five tests, using the "adjusted IQ score, adjusting by the "SEM", the lower bound of the SEM adjusted interval is less than 70. Thus, based on one test of five, Smith could be declared intellectually disabled and not be executed.
Effectively, the five IQ tests are correlated. The court appears to have ruled out a multiple comparisons procedure (MCP).
SCOTUS
The case occurred in Alabama in 1997. Smith murdered Hamm. Further details below
https://www.supremecourt.gov/opinions/24pdf/23-167_heim.pdf
The Supreme Court will review a case from Alabama. Briefly, Joseph Clinton Smith murdered Durk Van Hamm. The deciding factor is whether Smith is intellectually disabled, and if so, the death penalty will not be applied because it is cruel and unusual punishment. This requires the SCOTUS to assess whether the evidence that five (5) IQ tests confirm that Mr. Smith is intellectually disabled,
The statistical issues before the court are several-fold.
1) Smith was administered five different IQ tests- the scores are below.
2) The margin of error for the smallest IQ test value, 72, must be "3" because the score with "adjustment " (court/attorney's words) is 69, consistent with Smith being intellectually disabled.
a. Caveat Emptor: Not clear to me if that means the same test five different times or five entirely different tests.
b. Also, not clear what the phrase "full scale IQ test" means
c. An issue is the precise definition of the "margin of error" vs "standard error of measurement" vs "range of standard error", that the court is applying
3) The Supreme Court (?) ruled that the assessment jointly of the five tests "is a complicated endeavor," excerpting
"Analyzing Smith's intellectual functioning requires evaluating his various IQ scores. In Hall v. Florida, 572 U. S. 701, 714 (2014), this Court stated that "when a person has taken multiple tests, each separate score must be assessed" considering the standard error of measurement. The Court further noted that "the analysis of multiple IQ scores jointly is a complicated endeavor." Ibid. This Court has not specified how courts should evaluate multiple IQ scores. See ibid.; Moore v. Texas, 581 U. S. 1 (2017); Brumfield v. Cain, 576 U. S. 305 (2015)
Further complicating matters, Alabama failed in its attempt to execute Smith by use of nitrogen hypoxia.
Excerpting from Sotomayer
https://www.supremecourt.gov/opinions/23pdf/23a688_ap6c.pdf
Alabama tried and failed to execute him before. In November 2022, Alabama botched Smith's execution by lethal injection. It was Alabama's third failed execution in a row in five months. See Barber v. Ivey, 600 U. S. ___ (2023) (SOTOMAYOR, J
The allegations against Joseph Clifton Smith
In 1997, Joseph Clifton Smith brutally beat Durk Van Dam to death with a hammer and saw-inflicting thirty-five blunt-force injuries including brain bleeding, rib fractures, and a collapsed lung-in order to steal $140, the man's boots, and some tools. Smith was convicted of capital murder during a robbery.
The argument over intellectual disability
At sentencing, Smith's defense argued that he was intellectually disabled and thus ineligible for execution under Atkins v. Virginia (2002), which prohibits executing individuals with intellectual disabilities. But under Alabama law at the time, an individual was presumed not intellectually disabled if they scored above 70 on an IQ test. Smith's IQ was measured at 72.
In total, Smith has received five full-scale IQ scores as an adult: 72, 74, 74, 75, and 78. He also had two scores measured when he was under 18, scoring 74 and 75. At the federal evidentiary hearing, both sides presented expert testimony. The district court found that while Smith's intellectual functioning was a "close case", it fell within the range (70-75) where further evidence of adaptive functioning must be considered per Hall v. Florida (2014) and Moore v. Texas (2017).
The district court ultimately found Smith intellectually disabled under Atkins, citing not just his IQ scores, but also extensive evidence of deficits in adaptive functioning-across social, conceptual, and practical domains-going back to childhood. These included special education placements, poor academic achievement, social naivety, and limited independent living skills. The Eleventh Circuit affirmed, deferring to the district court's factfinding and concluding there was no clear error.
https://www.washingtonpost.com/politics/2025/06/07/scotus-iq-death-penalty/
Appellate Court Statistical Reasoning
https://cases.justia.com/federal/appellate-courts/ca11/14-10721/14-10721-2024-11-14.pdf?ts=1731630699
excerpting
We unambiguously reject any suggestion that a court may
ever conclude that a capital defendant suffers from significantly
subaverage intellectual functioning based solely on the fact that the
lower end of the standard-error range for his lowest of multiple IQ
scores is 69. And we didn't so conclude the last time we opined in
this case
Hamm v. Smith, the state of Alabama is arguing that Joseph Smith
https://deathpenaltyinfo.org/united-states-supreme-court-sends-case-of-alabama-death-sentenced-prisoner-back-to-11th-circuit-court-of-appeals
reddit
includes gory case details.
https://www.reddit.com/r/supremecourt/s/lg6VF0vG4i
probability (National Association of Attorney General)
https://www.naag.org/attorney-general-journal/supreme-court-report-volume-32-issue-2/
dueling experts
excerpting
Relying on these background principles, the Eleventh Circuit noted the district court's finding that Smith's IQ could be "as low as 69" given the standard error; and the district court had also determined, after reviewing dueling experts' testimony, that Smith's lowest score could not be "thrown out as an outlier." The Eleventh Circuit held that the district court had properly applied Hall and Moore and had not clearly erred in determining that Smith had shown "significantly subaverage intellectual functioning."
------------------------------
Chris Barker, Ph.D.
Adjunct Associate Professor of Biostatistics
www.barkerstats.com---
"In composition you have all the time you want to decide what to say in 15 seconds, in improvisation you have 15 seconds."
-Steve Lacy
------------------------------