Depending on whom you ask, “de-escalation training” is either a miracle cure or a four-letter word.

As high-profile, deadly confrontations between law enforcement officers and civilians continue to generate widespread public concern, de-escalation training has been hailed as the solution for this seemingly intractable problem. Public officials and policy makers from across the political spectrum have embraced de-escalation training as the key to safer interactions between police and the public.

But for some law enforcement officers, “de-escalation” is a loaded word. “What they hear is, ‘You’re teaching me to hesitate, and that’s gonna get me killed,'” said Robin Engel, a professor of criminal justice at the University of Cincinnati. “When we go in to implement this, we have to call it something else.”

Engel, a nationally respected criminologist, has conducted years of research on de-escalation training and its role in law enforcement. What she found early on surprised her: Despite the conflicting claims and strong emotions, very little was proven and consistent in de-escalation training. There weren’t many standards regarding what de-escalation training should teach or how it should be implemented, and absolutely no studies examining whether or not it worked.

Engel, whose work on de-escalation training is supported, in part, by grants from the Bureau of Justice Assistance (BJA), set out to find answers. And while de-escalation training is not a cure-all—it can’t be applied in every situation—a groundbreaking study by Engel found that de-escalation training can dramatically reduce injuries among civilians and law enforcement officers alike.

“I was surprised and thrilled that we found a reduction in injuries,” Engel said in a recent interview. “As a social scientist, most of the studies we do don’t find enough evidence to show anything. But in this case, I was really surprised by the size of the finding.”