Are Your Officers Well Trained?

Variability of Response

Publication

R. Horn, R., J. Lewinski, W., Sandri Heidner, G., Lawton, J., Allen, C., Albin, M. W., & P. Murray, N. (2023). Assessing between-officer variability in responses to a live-acted deadly force encounter as a window to the effectiveness of training and experience. Ergonomics, 1–16. https://doi.org/10.1080/00140139.2023.2278416

What Was the Issue?

Police officers must be able to effectively respond to deadly force situations. Yet, actual deadly force encounters are (fortunately) rare. This means that officers will not become proficient or expert at responding to these encounters simply by being on the job. Officers must be trained if we want to see them exhibit high levels of skill during deadly force encounters.

Officers receive training in the academy, during their post academy field training, and during other training sessions over the course of their careers; yet little research has explored the effects of this training on tactical performance - especially during deadly force situations.

This study seeks to address this knowledge gap. The most direct way to do this is by conducting a pretest, providing some training, and then having participants complete a post test. This study does not present data on that type of tightly controlled experiment. Instead, they present data on the variability of participants, who are all members of the same department, in responding to a deadly force scenario. As such, it gives insight into whether members of the same department who have received similar training respond in a similar way.

How Did They Look at It?

The authors tested 42 police officers who were all from the same police department. This agency uses an academy that provides 968 hours of training (which is more than 300 hours beyond the state requirement) and more than 100 hours above the national average for academy training. They also provide about 100 hours of post academy training to all new officers.

The scenario is the same one described in this post. The scenario was quite complex and included receiving an initial low priority call that was then upgraded to a higher priority call, driving through a course, and then interacting with multiple people at the scene. Four role-players were involved in the scenario. The assailant was the driver of one car involved in the accident. The other driver was an off-duty police officer. The third was a passenger in the assailant’s car, and the fourth was a motorist that stopped to help. Two officers responded to the scenario. During the interaction on scene, the assailant became increasingly agitated and eventually shot at the police officers.

During the scenario, the researchers tracked where the officers parked their cars, where the officers moved on scene, when they drew their guns, and when they fired at the assailant (if they fired).

What Did They Find?

The major finding was that there was a tremendous amount of variability in officers’ responses. During each scenario run, each team of officers parked in different places and moved along different paths. The figure below illustrates these differences. Each pair of officers also drew their weapons at different times. There was substantial variation in how long it took officers to fire at the suspect after he started shooting (with some officers failing to fire at all). When controlling for the age of the officer, years in law enforcement, specialized training (like SWAT or military experience), hours awake before testing, number of hours slept in the last sleep cycle, and a self-report of how rested the officer felt, only the hours awake before testing was significantly predictive of the time it took officers to return fire. Those who were awake longer were slower to react.

So What?

The authors acknowledge that some variability is unavoidable, but argue that the level of variability seen in this study indicates that the officers failed to plan how they would approach the situation and that they did not possess a shared tactical understanding. To the authors, this indicated that the officers were not adequately trained. This is despite the fact that this department had more training than was required by the state and more than the national average. The authors also indicated that much training is not scenario based and as such does not help officers to develop the skills needed to deal with the complexities of actual encounters.

The authors additionally emphasized that being awake longer resulted in a slower reaction to the assailant’s shots and that this area should receive more attention in the future.

My .02.

Much of what we are trying to accomplish in training is to create perception-action loops that transfer to the performance environment (to the street in policing). We want trainees to be able to ignore the noise in a situation and focus on what is important so that they have information that allows them to act effectively. In the literature, this is called specifying information. It specifies one course of action from another.

Consider the scenario presented in this study. There is a lot going on that the officers must sort through. Ultimately, they must understand what is happening and, based on that understanding, decide what to do. The behavior of the person who eventually becomes the assailant should have indicated to the officers that they needed to be prepared to deal with him. To accomplish this intention requires moving to a position where the officers can see the suspect and prevent him from assaulting the other people on scene. There is more than one position that will afford this, and different officers will have different characteristics which affect what positions will work for them. As such, I would expect to see variation, but I would expect it to be in clusters. People who have similar capabilities will choose similar positions. What matters in evaluating the response of the officers is not the specific position that the officer chose, but rather whether that position allowed the officer to see what he needed to see and act accordingly.

The same goes for the second officer. Since the first officer is primarily engaged with the soon to be assailant, the second officer’s intention should probably be to protect his partner. If this is his intention, his focus should be primarily on the other people, but he must also have a view of and be able to assist with the assailant. The second officer must move to a position that allows him to pickup specifying information and allows action based upon that. Again, I would expect there to be clusters based upon the capabilities of the second officer and locations/actions of the other people on scene. As long as the position allowed the officer to see and act, it is was a good position.

So I agree with the authors of this study that the variation they are seeing indicates that the officers are not adequately trained. Not simply because there is variation, but because that variation does not seem to be functional. Some of the officers appear to be in positions that do not allow them to see the important information and/or act effectively.

Clusters of similar behaviors will only emerge if the officers are “well trained”. The only way for the officers to become “well trained” is to put them in scenarios over and over again to allow them to learn what they should be trying to accomplish, where to focus their attention, and what their capabilities are.  They can’t learn it from reading it in a book or doing decomposed training (like hitting a bag, driving around cones, punching holes in paper targets, or performing room entries on empty rooms). Initially, these scenarios probably need to be much more constrained than the one in this study. This will help the trainees to understand what they are trying to do, identify the important information, and appreciate their capabilities. As the trainees improve, the scenarios can (and should) become more complex. Producing skilled officers will take a lot of time and effort. Which is why I continue to argue that policing must become a training culture.