Think Out Loud

Washington State University study finds anti-bias training can benefit police

By Sheraz Sadiq (OPB)
Oct. 30, 2023 4:14 p.m.

Broadcast: Tuesday, Oct. 31

A new study by researchers at Washington State University found a “small but significant” effect from training to counter implicit bias in police officers. Implicit bias occurs without conscious awareness and can shape prejudicial attitudes and behaviors. The study recruited 50 police officers in Sacramento, California to participate in video simulations of encounters between police officers and different community members. Two of the study’s authors developed the interactive training program which is being sold to police departments nationwide. The researchers found that officers who completed the training had fewer complaints of discrimination by community members than officers without the training. Police bodycam footage also revealed improvements in trained officers’ behavior, especially in their interactions with people experiencing homelessness. Lois James is the lead author of the study, and assistant dean of research and an associate professor in the College of Nursing at Washington State University. She joins us to talk about the results.


This transcript was created by a computer and edited by a volunteer.

Dave Miller: This is Think Out Loud on OPB. I’m Dave Miller. A new study by researchers at Washington State University found that implicit bias training for police officers had a small, but significant, positive effect. The study recruited 50 officers in Sacramento who participated in video simulations of encounters with various community members. Researchers found that officers who completed the training had fewer complaints of discrimination by community members than officers without that training. Lois James is the lead author of the study. She’s an assistant dean of research and an associate professor in the College of Nursing at Washington State University. She joins us now. Welcome to the show.

Lois James: Hi. Thank you so much for having me.

Miller: Thanks for joining us. What do you mean when you say implicit bias specifically in the context of police officers?

James: That’s such a great question. And honestly, it’s a concept that’s often misunderstood or not really understood very well. And what we mean when we say implicit bias is unconscious bias, the biases that lurk below the surface that we’re not even necessarily aware [of being] there. And the reason that implicit bias can be so problematic is, because of its subconscious nature, it can influence us in ways that we don’t intend or that we don’t anticipate. And it can shape some of our decisions and, unfortunately, can lead to some discriminatory behavior.

Miller: How much of what we talk about, in terms of policing in the US, is implicit bias as opposed to explicit bias?

James: Explicit bias, obviously, is [bias] that we’re aware of. Now that can manifest itself as bigotry, as prejudice, for example, racism, sexism, some kind of bias that a person would put their hand up and claim, “That’s who I am. That’s what I think.” And then you can also have covert explicit bias where somebody might be aware of it but they’re ashamed of it or they don’t, you know, they certainly wouldn’t put their hand up and say, “Yes, this is who I am.” So really with policing, it’s hard to say because if you can imagine giving somebody a questionnaire, unless they’re really overtly explicitly biased in which case, hopefully they’re not working for a department anymore. They’re not going to say, “You know, I believe this group of people are worse because…”

So really what we’re exploring is what are biases that are implicit and what are biases that are explicit but covert. So what we’re trying to do is we’re trying to shine a light on implicit bias for officers so they can actually bring them into conscious awareness so that they can recognize that, “Okay, maybe I do think that. Maybe I do have this association.” And if they’re aware of it, at least they have a chance of controlling it.

MillerThe implicit bias tests that I’ve done, and I have a feeling that many listeners at this point have done, often it’s how quickly can you see, say a person who may be Black or white and words that are negative or positive and you do these in various iterations. And you’re supposed to do it fast enough to be able to click on things and when you do that, you find out that without realizing it, your brain makes certain associations. That’s, I think, a common online version just for the general public. But what does the training that you’ve developed entail in its entirety?

James: The test you were talking about is the implicit association test out of Harvard. And it remains our best measure of implicit bias. What we do with the training is more behaviorally based. So with the training, we have officers go through these scenarios, these very realistic scenarios that are just based around fairly routine police community member interactions that often have the potential to be volatile. So things like interacting with people in crisis, responding to domestic disturbances, doing welfare checks, investigations of suspicious persons or circumstances, things like that. And the officers respond in this simulated setting. They respond to the scenario and then we have the ability to drive the scenario or dictate what the person on-screen does or says, based on how the officer is behaving.

So for example, if an officer approaches somebody in a scenario, introduces themself by name, explains why they’re there, offers some kind of concern or empathy, offers any kind of help to the person, any kind of resources, tries their best to de-escalate. So if it does go volatile, if an officer does all of those things, we have the ability then to end the scenario in a very positive manner. However, if that same officer were to, let’s say, put their hand on their holster, not be personable at all with the person, not introduce themselves, not explain why they’re there, just rapid fire questions, and behave in a way that we’re really trying to move away from, the really authoritative, not humanistic side of communication, if they do that, then we have the ability to send the scenario in a much more volatile negative way. So by going through these scenarios, officers start to A) practice the skills that we want them to practice and B) start to uncover themselves, “OK, do I respond differently based on the type of person that I’m interacting with?”

And what we’re really trying to do with the training is to generate that self-awareness. As I said, shining a light on implicit bias and making an officer aware of what their mental filter looks like, that they see the world through. That’s where the true learning occurs. Not just in the responses to the scenarios, but in all of the great conversation afterwards, both with their peers who are observing and then with the facilitators and the trainers as well.

Miller: You’ve said that you’ve seen small but significant decreases in disparities in terms of how officers treated different types of people. What do you see as the most important results from this study?

James: I do think that reduction and disparity piece is really important. It was encouraging and it was great to find that the officers in the intervention group had generally better performance scores after the training. And what I mean by that is regardless of who they were interacting with, they tended to display more of the indicators of really good policing, based in procedural justice, based in de-escalation, based in all of the great things that we want our officers to be doing as they protect and serve community members.


But the thing that was really encouraging is we looked at all of our baseline data because we looked at a year essentially - not every single video of course, but a random selection of body camera footage videos from the year before the training intervention - to really see what the disparities in this city with this department look like. And the biggest one, by far, was how officers interacted when they were interacting with community members suffering homelessness versus when they were interacting with community members who were not suffering homelessness.

Miller: That was higher than the disparity between, say Black residents and white residents?

James: Correct. So the homelessness piece was the biggest disparity that we saw and that’s homelessness kind of controlling for the race of the person or the gender of the person or the size of the person. So you’ve got body camera footage, there are a lot of variables that you can control for, because you’re seeing this person. Now, a small caveat there that is important is that oftentimes, of course, we did have to make judgments. So it’s like, “appears” to be suffering, homelessness, “appears” to be Black, “appears” to be white, “appears” to be male or female. Oftentimes these are assumptions that we’re making. But we did a great deal of training with our research assistants to really make sure that they were as well trained as possible, as they coded this footage.

Miller: And I imagine we’re also talking about assumptions anyway, on the part of officers. So hopefully there might be some convergence there. My understanding is that it’s not uncommon these days for police departments around the country to do some kind of implicit bias training. But a press release about your study said this is the first known research to provide evidence that this kind of training can actually produce positive effects. If that’s the case, why has this training become so common?

James: I should say, ours is the first to produce positive behavioral effects. So, in the past studies have certainly looked at [whether] an officer’s knowledge around bias increases or improves. And do their attitudes or their intent to put into play the lessons that they learned during training, does that improve? But this is the first study to kind of look at what officer behavior looks like before the intervention? And now what does it look like after the intervention?

Miller: That seems like the most important thing, from the public’s perspective. What are officers actually doing as opposed to what do they intend to do? What have they learned from a book?

James: Exactly. That’s what the public cares about. And also honestly, that’s what departments care about too because training is intended to produce behavior change.

Miller: So how do you explain that? Why is it that these have become so common? If yours is the first study to show positive behavioral changes?

James: I think part of the answer to that is that it’s often mandated. If a department is going through any kind of consent to care or collaborative reform then, typically, some kind of anti-bias training is going to be front and center, especially if that’s one of the reasons that they’re being audited or observed closely anyway. And then I think another reason is that departments sometimes can be very proactive. They can have communication with their community leaders for example, that can recommend it. So I think that oftentimes departments think that it’s a good idea.

We just haven’t, to this point, had any really solid evidence that it could be a positive thing. And that’s the concern that some people have had with implicit bias training. Is there any potential that it could be producing negative effects or damaging effects which, thankfully, was not observed in our study. But regardless, I do think that additional research needs to occur because this was a fairly small study and it’s one police department. Whether that’s generalizable to all police departments, is just something that we don’t know.

Miller: I should also mention that tool that you studied is one that you helped develop. If it’s seen as effective, it could lead to more business for you. More police departments could buy it. How did you deal with that potential conflict of interest?

James: So we dealt with it in a number of ways. This study was actually funded by the National Institute of Justice, which is a Department of Justice research branch. And because there was that potential conflict of interest, we had to put several steps in place. And the big one that we put in place was that, as our research assistants were coding data, they were completely blinded to the group that they were coding. So in other words, when they were reviewing a body camera footage video, they did not know whether that officer was trained or not.

So there was no potential for even any kind of subconscious… “I think that this officer might have been in the intervention group so I’m going to lean towards scoring them more favorably.” That just was completely removed from the equation. And that’s the big one. And honestly, that’s very typical with randomized control trials with experimental designs, that notion of blinding because it’s possible that even subconsciously, if people knew that, they could score things slightly differently. So that was the big one. And then the other big one of course is that we were not actually doing the coding ourselves.

Miller: Lois James, thanks very much for joining us. I appreciate having you on.

James: Yeah, thank you for having me. I really appreciate the chance to talk about this.

Miller: Lois James is an assistant dean of research and an associate professor in the College of Nursing at Washington State University, the lead author of a recent study that found positive behavioral benefits after Sacramento police officers took part in training to counter implicit bias.

Contact “Think Out Loud®”

If you’d like to comment on any of the topics in this show or suggest a topic of your own, please get in touch with us on Facebook, send an email to, or you can leave a voicemail for us at 503-293-1983. The call-in phone number during the noon hour is 888-665-5865.