The Problem With Slow Motion

This article was originally published in the New York Times on August 5, 2015.

Watching slow-motion footage of an event can certainly improve our judgment of what happened. But can it also impair judgment?

This question arose in the 2009 murder trial of a man named John Lewis, who killed a police officer during an armed robbery of a Dunkin’ Donuts in Philadelphia. Mr. Lewis pleaded guilty; the only question for the jury was whether the murder resulted from a “willful, deliberate and premeditated” intent to kill or — as Mr. Lewis argued — from a spontaneous, panicked reaction to seeing the officer enter the store unexpectedly.

The key piece of evidence was a surveillance video of the shooting, which the jury saw both in real time and in slow motion. The jury found that Mr. Lewis had acted with premeditation, and he was sentenced to death.

Mr. Lewis appealed the decision, arguing that the slow-motion video was prejudicial. Specifically, he claimed that watching the video in slow motion artificially stretched the relevant time period and created a “false impression of premeditation.” Did it?

We recently conducted a series of experiments whose results are strikingly consistent with that claim. Our studies, published this week in the Proceedings of the National Academy of Sciences, show that seeing replays of an action in slow motion leads viewers to believe that the actor had more time to think before acting than he actually did. The result is that slow motion makes actions seem more intentional, more premeditated.

In one of our studies, participants watched surveillance video of a fatal shooting that occurred outside a convenience store during an armed robbery. We gave them a set of instructions similar to those given to the jurors in Mr. Lewis’s case, asking them to decide whether the crime was premeditated or not. We assigned half our participants to watch the video in slow motion and the other half to watch it at regular speed.

Those who saw the shooting in slow motion felt that the actor had more time to act than those who saw it at regular speed — and the more time they felt he had, the more likely they were to see intention in his action. (We found similar results in a separate study involving video footage of a prohibited “helmet to helmet” tackle in the National Football League, where the question was whether the player intended to strike the opposing player in the proscribed manner.)

We also ran a statistical simulation to estimate how likely it would be to get a unanimous decision from a 12-person jury selected from our participants. Juries who saw the slow-motion video were nearly four times more likely to return a unanimous first-degree murder verdict than juries who saw the regular-speed version.

In Mr. Lewis’s case, the Supreme Court of Pennsylvania upheld his conviction, affirming the trial court’s decision that the slow-motion video was more probative than prejudicial. The court’s reasoning implied that, even if jurors were biased when they saw the slow motion, they had enough knowledge to correct for that bias: They were aware that Mr. Lewis shot the officer within two seconds of noticing him; and they saw the real-time version of the video in addition to the slow-motion version.

But our research suggests that such knowledge can fail to offset slow-motion bias. In another study using the convenience store video, we repeatedly informed participants of how many seconds had elapsed during the original event, and we verified that they could accurately report this information back to us. Nonetheless, they still judged the shooting as more intentional when they saw it in slow motion.

In other studies, we showed some participants the regular-speed version before they saw the slow-motion version. This reduced the bias but did not eliminate it. When we again used our results to simulate jury-selection outcomes, juries of participants who saw both versions were still about 50 percent more likely to be unanimous in a first-degree murder verdict than juries of participants who saw only the regular speed version.

We live in a YouTube age. These days, if something disputed happens in public, there is a fair chance it will be captured on video and analyzed in detail. It is important to keep in mind that video evidence can distort human judgment as well as sharpen it. Our studies suggest that the courts, at least, may not want to be so fast to slow things down.

Eugene M. Caruso is an associate professor of behavioral science at the University of Chicago, Zachary C. Burns is an assistant professor of organization, leadership and communication at the University of San Francisco, and Benjamin A. Converse is an assistant professor of public policy and psychology at the University of Virginia.