Washington State’s Lois James on training police to recognize their biases

interview by Sheila G. Miller

Lois James is a professor in the College of Nursing at Washington State University in Spokane. With her husband, Steve James, she researches use of deadly force in law enforcement, and what underlying biases may be at play when police draw their weapons. She runs a lab with high-end use-of-force simulators, conducting research about how and why officers respond to shoot-or-don’t-shoot scenarios. Now she’s developed a training program that takes a portable simulator on the road.

How did this project get underway?

We wanted to be able to study things that are difficult to study in the field. This way we can study it in a controlled environment that still exposes officers to situations that are as real as they can get without being real.

What did your research show?

It’s really important obviously to put this stuff in context. There is a well-founded and controversial debate about whether officers are influenced by a suspect’s race when they make a decision to shoot. We found two things. One, the officers tend to have very strong implicit associations between African-Americans and the threat of weapons. But we also found a counter-bias effect, which is that because of this implicit bias when they were tested in the simulator they tended to display hesitancy to shoot African-American suspects versus white suspects.

We are a long way from figuring out what this means. We just tested officers in one department so there’s a lot more to do with research, but neither is desirable for officers on the street.

How did you come up with the training scenarios?

I took thirty years of data on officer-involved shootings from the FBI’s Law Enforcement Officers Killed and Assaulted data, and looked at all different types of encounters. I created scenarios based on those.

I look at things like, what’s the average distance between them, what is the time of day, what kind of a call is it—a domestic disturbance or some kind of traffic violation? We created sixty scenarios and ultimately really they are “shoot-or-don’t-shoot” scenarios.

As for demographics, the scenarios match the people most likely to assault officers. So it’s primarily white, then black, and then Hispanic. Those are the only three racial and ethnic groups in the scenarios because the other ethnicities are much fewer in number in terms of the demographics of those who have killed or injured cops.

How does the training work?

Law enforcement tends to favor scenario-based training because, if you practice a skill as opposed to hearing it, that skill is going to become a muscle memory, an ingrained cognitive muscle memory.

We use counterconditioning—the argument there is we very carefully select which scenarios to use, and they differ based on department.

Officers do this training in teams of five and their fellow officers are in the room with them and are observing. So officers go through the scenario and then are asked to do a self-reflective debrief with careful and specific questions designed to dig out and uncover any bias.

Fellow officers are asked to comment on an officer’s performance, what they saw, what they would have done the same or different. What we’re trying to do is get away from the stigma of bias training, because officers can feel that it’s quite judgmental and be defensive about it.

What’s next?

I’m submitting a grant proposal with the National Institute of Justice to do an evaluation of implicit bias training, because one of the big elephants in the room is that nobody really knows whether it really works because no one has ever tested it. In the realm of law enforcement there is very little known about it—some voices say implicit bias training might actually do more harm than good.