The Supreme Court wrestled on Dec. 10 over how judges should weigh IQ scores when determining whether a convict was wrongly sentenced to death.

The case, Hamm v. Smith, centered on an Alabama man convicted of murder in the 1990s, but the legal questions involved stem from a 2002 decision by the high court known as Atkins v. Virginia.

In that case, a majority held that the Eighth Amendment’s bar on cruel and unusual punishment prevented death sentences for intellectually disabled individuals.

Questions remained, however, as to how individual states would determine disability and courts would in turn apply the Eighth Amendment to state determinations.

During oral argument on Dec. 10, the justices questioned how much flexibility states should have in guiding those determinations, as well as how courts should consider factors such as the margin of error for IQ tests.

Joseph Clifton Smith was sentenced to death in a state court, but two federal courts, after examining his IQ scores and other evidence, later said that the sentence was inconsistent with the Eighth Amendment.

According to court filings, Smith received five IQ scores—75, 74, 72, 78, and 74—near the 70-point benchmark Alabama has set for deciding whether someone is intellectually disabled.

When the U.S. Court of Appeals for the 11th Circuit reviewed his case, it said that it needed more certainty that his actual IQ score wasn’t below that level.

Justice Samuel Alito seemed skeptical of this approach and wondered how far courts could go in questioning IQ scores above 70.

“So if there are multiple IQ scores, there are five 100s and one 71, you still—the court can’t just say the defendant has not—cannot prove that he’s intellectually disabled?” Alito asked Seth Waxman, who was representing Smith.

Alabama had asked the Supreme Court to consider the cumulative effect of IQ scores rather than moving on to other factors, such as adaptive deficits, if a lower IQ score raised uncertainty about the defendant’s disability.

Justice Sonia Sotomayor’s questions indicated she was more supportive of the 11th Circuit and critical of Alabama’s attempt to reverse the appeals court’s decision.

She told Alabama Principal Deputy Solicitor General Robert Overing that he was asking the court to “undo” two precedents that followed Atkins.

“So for us to say what Petitioner wants would disavow that language in Hall,” she said. “It would disavow nearly the identical language in Moore.”

She was referring to the Supreme Court’s 2014 decision in Hall v. Florida and 2017 decision in Moore v. Texas.

In the first case, the court said that when an IQ test score falls within the test’s margin of error, the defendant must be able to present additional evidence of intellectual disability.

Moore v. Texas focused on a defendant who had six scores averaging 70.66, which the Supreme Court said indicated “mild intellectual disability.”

It faulted a lower court for using two higher IQ scores (74 and 78) to say that the defendant hadn’t shown significantly subaverage intellectual functioning.

In each of those cases, Alito, Justice Clarence Thomas, and Chief Justice John Roberts joined dissents.

Penning a dissent in Moore, Roberts said the court had erred “in concluding that the Eighth Amendment turns on the slightest numerical difference in IQ score.”

Thomas, who was the only member of the current court who was serving in 2002, also dissented from Atkins.

Roberts and Justice Brett Kavanaugh, however, did express skepticism of Alabama’s position on Dec. 10, suggesting its approach to weighing IQ scores was potentially unfair to defendants.

“What if the scores were 69, 68, 69, 69, 75?” Roberts asked. “I mean, would you concede in that case that it was—the actual IQ is below 70, or would you still argue, well, 75 is above, so we’ve got to look at all sorts of other things?”

After Overing said the defendant would have to “prove a likelihood that his IQ is 70 or below,” Roberts indicated that the approach was “results-oriented.”

“When you have scores above 70, you want to average them and discount the one below, but when they’re all below, you don’t do that,” Roberts said. “You instead say you’ve got to look at all these other factors?”