Learning to de-escalate: evaluating the behavioural impact of Verbal Judo training on police constables

ABSTRACT

This article presents findings from an assessment of the immediate behavioural impact of the ‘Verbal Judo’ de-escalation training program, delivered to officers in a Canadian police service in 2017, and represents the first evaluation of Verbal Judo. The evaluation utilized systematic observations of officers in videotaped simulations, comparing a baseline group of officers to officers trained in Verbal Judo. The study found that five of 15 coded behaviours were impacted at a statistically significant level in the hypothesized direction, and these tended to be simpler behaviours encouraged by the training (such as identifying oneself or providing a reason for contact), while more complex behaviours (such as empathizing with the subject) were not impacted. The evaluation also found that likelihood of utilizing de-escalation skills decreased as an officer’s years of experience increased. The paper closes by outlining the implications of these findings for de-escalation training, and elaborates the main limitations of the study.

KEYWORDS:

• Policing
• police training
• de-escalation
• evaluation
• Verbal Judo

Acknowledgments

The project team completed this research with the support of the members of the HRP training section, including Julie Moore, (then) S/Sgt. Penny Hart, and Sgt. Dennis Crowell, the course trainers, and the officers who participated in the baseline exercise, all of whom deserve thanks for making this possible. Externally, Darcy Pennock from Verbal Judo Canada, Dr Paul Quinton and Julia Wire from the College of Policing (UK), Dr Meg Ternes from Saint Mary’s University (Canada) and Dr Cody Telep from Arizona State University (US), all provided valuable input at different stages in the project. We also thank the anonymous peer reviewers and journal editorial staff for their comments. This project greatly benefited from the input from the wide range of stakeholders listed above, and any errors or omissions are those of the project team alone.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental data for this article can be accessed here.

Notes

1. Block training is an annual two-day exercise that all HRP officers are required to complete. Block training 2017 included Verbal Judo training, alongside ‘business as usual’ use-of-force re-certification training for the remainder of the two-day session.

2. Both Miller and Alexandrou, and Wheller and colleagues, also used administrative data (written grounds and stop and search rates in the former; victim feedback data in the latter), but in both cases these were treated as secondary outcome metrics, while attitudinal and (in Wheller et al) role play data were treated as ‘primary’ outcomes.

3. There are no appropriate data sources for longitudinal or pre-post measurement of officer stress and morale, and there have been far too few lawsuits brought against HRP officers relating to their behaviour in recent years for this to be considered a relevant outcome metric.

4. A full description of the categories and coder instructions can be made available on request to the authors.

5. Additionally, the coder’s work was supervised by the project lead, and after the first 100 videos were coded by the coder, 12 randomly-selected videos were ‘double-coded’ (independently, once by the project lead and once by the coder). The results were then tested for inter-rater agreement. Overall, there was 85% agreement between the two datasets, across 199 data points coded by each coder. Any points of disagreement were then discussed, coding instructions were modified for clarity, and any required modifications were retroactively made to the first 100 coded videos for consistency.

6. A random list of officers was generated using internal records, and email requests were sent for officers to attend a session at the training facility. In the email request, officers were advised that they were being asked to participate in an exercise to help assess our training, that their participation was voluntary, and that they were welcome to ask questions and given the opportunity to decline to take part. Some officers initially selected to take part were unable to attend due to operational needs or other responsibilities, so in total approximately 100 officers were invited to take place, and the first 64 to be available were included in the baseline group.

7. There were no female-only pairs in the baseline group, while female-only pairs made up approximately 8% of the treatment group. The female only pairs have been nonetheless included for statistical testing in this study; however even when accounting for female-only pairs the results of the analyses are effectively the same.

8. Using an independent samples t-test, the difference in average experience was significant at p < 0.01.

9. Due to the nature of the training, officers in the baseline exercise were by necessity included in the treatment phase, albeit in different pair combinations than observed in the baseline phase. While this means that some participants in the treatment phase had previously experienced the simulation scenarios, the results described later in the paper are not significantly altered if we isolate the analysis to only those treatment phase officers that had no involvement in the baseline exercise.

10. For more on integrated policing in Canada, see Giacomantonio 2015.

11. For a series of reasons – related to i) the need to blind ‘baseline’ participants as to the nature of the baseline exercise, ii) an inability to link anonymous post-training questionnaires to behavioural data and iii) limitations in personnel records related to prior training – it was not possible to control for prior de-escalation training in the study; however, this is a minor issue in that it is safe to assume that all officers had prior exposure of some sort to de-escalation concepts and methods.

12. While this evaluation seeks to understand whether Verbal Judo impacted officer behaviour, it is important to recognize that the evaluation does not involve any value judgements about the behaviour of officers who did not exhibit behaviours encouraged by VJ. This is primarily because we do not, at this stage, have evidence to show that VJ behaviours result in better outcomes in the field – while such a claim is supported in general terms by social psychology research on aggression in dispute contexts, there is no systematic evidence on this question specific to the policing environment.

13. The issue of shortening the training to fit available training time was discussed during the train-the-trainers session, suggesting that the Verbal Judo Institute has worked with police services providing VJ at a range of different program lengths.

14. As it turned out, this approach to measuring speed of use of force was primarily relevant to the ways in which officers used force in S1, since (a) there were very few uses of force in S2 and (b) there was no obvious ‘branch’ point in S2 after which force may be appropriate, while there was a clear singular branch point in S1 after which force was appropriate.

15. SBOR have been collected since before 2012; however, 2012 is the first year for which a consistent internal dataset has been produced internally.

Additional information

Notes on contributors

Chris Giacomantonio

Dr. Chris Giacomantonio is the Research Coordinator at the Halifax Regional Police in Canada, Adjunct Professor in Psychology at Saint Mary’s University and co-chair of the Canadian Association of Chiefs of Police Research Advisory Committee. He holds a PhD in Criminology from the University of Oxford and conducts mixed-method research on a wide range of policing and criminal justice topics.

Stephanie Goodwin

Ms. Stephanie Goodwin was a Research Assistant with Halifax Regional Police during the course of this study. She holds an Honours BSc in Forensic Psychology and currently works as a Psychology Technician at the NS Hospital Sex Offender Clinic.

Garland Carmichael

Sgt. Garland Carmichael is a 20-year veteran of Halifax Regional Police. He has worked in a wide range of roles including Training program developer with the Strategic Technology Section, Detective Constable with the Integrated Special Enforcement/Drug Section, and an Instructor with the HRP Training Division. Sgt Carmichael is an active peer referral agent with the Employee and Family Assistance Program and currently works as a Patrol Sergeant in HRP’s West Division.