In the wake of recent verdicts, the Johnson & Johnson defense team might be seeing its hopes of convincing jurors to follow its scientific advice going up in a cloud of smoke…or perhaps a cloud of potentially carcinogenic talcum powder. Johnson & Johnson has lost a string of suits, most recently in April ($117 million and $26 million) and July ($46 billion), with the common factor being a tendency for jurors to find the plaintiffs’ anecdotes to be more compelling than the defendant’s data. The science supporting the plaintiffs’ cases, now numbering more than 9,000, is still quite controversial. While at least some scientists might be shaking their heads, the Johnson & Johnson jurors aren’t necessarily off-base.

As explained in a recent article in Popular Science entitled, “The Problem with Taking Scientific Questions to Court,” the lack of agreement between what the scientific community says and what the verdicts say can be traced to the different standards of law and science. “What the legal system considers enough evidence to establish that exposure causes illness is different from the standards of science—and trying to fit the two together can be hazardous.” Of course, understanding that difference isn’t likely to satisfy parties like Johnson & Johnson who have no choice but to bring science into the courtroom. At times, the defense will need jurors in the courtroom to think and act a bit more likely they’re scientists in the lab. In this post, I will take a closer look at the Popular Science article and share a couple of thoughts on how to encourage jurors to adopt that mindset.

Law v. Science: 51 Percent or 95 Percent?

The legal standard is generally the familiar “preponderance of the evidence” standard of “more likely than not,” which plaintiffs often operationalize as just slightly above the balanced midpoint: 51 percent, or a feather placed on one side of the scale.

For scientists, of course, the standard is different. To reject the “null hypothesis” of “no relationship,” scientists need much more than 51 percent. The most commonly accepted error rate is usually expressed as equal or less than .05, meaning that there is a 95 percent or greater probability that the effect is caused by the phenomena under study and a 5 percent chance or less that it is due to sampling error. When the observed relationship doesn’t reach that 95 percent threshold, the relationship isn’t “statistically significant,” meaning it could be an accident of sampling and not an actual relationship.

Of course, when that science makes it into court, jurors aren’t told they need to be 95 percent sure, they’re told it only needs to be more likely than not. The result is that a lot of science that is still quite uncertain can still end up being used, creating some uncertainty for juries. “One of the questions is, what are juries really supposed to do?” asks Ed Cheng, who studies scientific and expert evidence at Vanderbilt Law School. “You have this sort of immature science out there, which isn’t clear. If you’re a juror, you really want to get it right, but what are you supposed to do when the numbers aren’t clear?”

The Response for Defendants

The response for defendants is to give jurors that guidance. Plaintiffs might be happy to apply the more relaxed civil standard to science, but being true to science while still staying within the law can be frustrating for defendants. Here are a couple of ideas for handling this.

Emphasize the Frame

How are you framing the trial and placing jurors in a role? For plaintiffs, the jury is likely framed as a voice for the injured and the powerless, with the verdict as a way to wrest popular justice from large and uncaring organizations. For defendants, the right frame might be to emphasize their role, not as dispensers of “justice but as investigators of truth. And when it comes to questions of science, approaching that truth requires a scientist’s commitment. It requires high standards, skepticism, and even doubt. Scientists begin assuming that the null hypothesis is true, and checking to see whether there is compelling evidence that allows one to be nearly certain in rejecting it. In that role, the jurors are the gatekeepers, and their role is to prevent flawed or hasty assumptions from getting past.

Emphasize the Threshold

There’s a reasonable point to be made that, when the strength of an observed relationship is more than the law’s 51 percent, but less than the scientific 95 percent, it is not a relationship with modest support, rather it is still a relationship with no support. The reason for that is that scientists don’t typically like to think of their work as “tipping the scale slightly.” Instead, within their discipline, they are either rejecting the null hypothesis or they’re not. That means that a relationship with 70 percent reliability isn’t 70 percent supported, it is zero percent supported. The science did not reach the threshold, so science is not able to offer an answer to this question.

This, of course, can be asking a lot from jurors. It will still be easier to look at the anecdote, or look at the not-yet-reliable association and conclude that there must be a relationship. But jurors can be motivated to reject the easy way. Wanting to believe that they’re doing a good and thorough job, they can be encouraged to, in a small way, think like a scientist.