Randomized controlled trial of social interaction police training

• Author: Scott E. Wolfe,Michael R. Smith,Kyle McLean,Geoffrey P. Alpert,Jeff Rojek
• Published date: 01 August 2020
• DOI: http://doi.org/10.1111/1745-9133.12506
• Date: 01 August 2020

DOI: ./- .

SPECIAL ISSUE ARTICLE

CUTTING-EDGE RESEARCH IN POLICE POLICY AND PRACTICE

Randomized controlled trial of social

interaction police training

Kyle McLean1Scott E. Wolfe2Jeff Rojek2

Geoffrey P. Alpert3Michael R. Smith4

Clemson University

Michigan State University

University of South Carolina

University of Texas at San Antonio

Correspondence

KyleMcLean, Department of Sociology,

Anthropology,and Criminal Justice, Clem-

sonUniversity,  BrackettHall, Clemson,

SC.

Email:kdmclea@clemson.edu

Thisresearch was funded by the National

Instituteof Justice under Grant -IJ-

CX-.Points of view and opinions

providedare those of the authors and do

notrepresent the official position of the

U.S.Department of Justice. An earlier ver-

sionof this article was presented at the 

AmericanSociety of Criminology meeting

inAtlanta, GA. The authors would like to

thanktheir partners in the Fayetteville and

Tucsonpolice departments for making this

evaluationpossible. They also thank John

MacDonaldfor his helpful comments. All

errorsbelong to the authors. The authors

holdno conflicts of interest in evaluating

thetraining program in this study. The

authorsacted as independent evaluators

whosedata collection efforts were sepa-

ratefrom the program delivery by Polis

Solutions.

Research Summary: We conducted a randomized-

controlled trial (RCT) of a social interactiontraining pro-

gram to determine its effectiveness in improving atti-

tudes and behaviors among police officers. Survey data

and a series of difference-in-difference tests found that

participating in the training program improved attitudes

with treatment group officers placing higher priorities

on procedurally fair communication during a hypotheti-

cal officer–citizen encounter. An interruptedtime- series

analysis of official use-of-force reports provided no evi-

dence that the training program altered officer behavior.

Policy Implications: Policing scholars and reform-

ers have increasingly called for improvements to

police training that emphasize communication and de-

escalation skills. Although many programs addressing

these issues exist, evidence of their effectiveness has

been scarce. Our findings provide evidence that such

training may improve police officer attitudes but per-

haps not behaviors.

KEYWORDS

de-escalation, evidence-based policing, police training, procedural

justice, randomized controlled trials, social interaction

In the wake of numerous controversial police use-of-force encounters with citizens in recent

years, the need for improved social interaction and de-escalation skills has been featured as the

Criminology & Public Policy. ;:–. ©  American Society of Criminology 805wileyonlinelibrary.com/journal/capp

806 MCLEAN  .

centerpiece of police reform efforts (President’s Task Force on st Century Policing, ). To

improve these skills, many have suggested improvements in the quality and availabilityof police

training on interactions with citizens. As such, academic researchers often argue that success-

ful police training interventions should focus on improving officers’ ability to apply procedural

justice or effective communication principles consistently throughout their interactions with cit-

izens (Mazerrole, Antrobus, Bennett, & Tyler, ; Sargeant, Antrobus, & Platz, ). Weknow,

for example, that citizens are more likely to comply when they believe officers have treated them

with respect, provided them with a voice, and been unbiased in their decision-making—that is,

when they have acted with proceduraljustice (Tyler, ; Walters& Bolg er,; Wolfe, McLean,

Rojek, Alpert, & Smith, )—so it is only logical that improving officers’ behavior along these

dimensions should improve police–citizen interactions. Yet, as Nagin and Telep () recently

noted, rarely is theory or research evidence integrated into police training, and evenwhen it is, it

is also rarely subjected to rigorous scientific evaluation.

A critical development in the professionalization of policing over the past  years has been

the implementation of data and science into the practices of police agencies (Sherman, ).

The evidence-based policing movement has demanded a model of policing that “links external

demands on police .. . to research evidence on how to meet those demands”(Sherman, ,p.)

or as Alpert (, p. ) called it, “linking data to decisions.” Yetwith respect to police training,

Skogan and colleagues (, p. ) accurately described the state of scientific knowledge by

declaring, “We know virtually nothing about the short- or long-term effects associated with police

training of any type.” Similarly, in attempting a systematic review of studies of police training,

Huey () concluded that there were too few studies on any single topic to conduct a review.

Thus, modern policing is currently in a precarious situation, with external demands for police

agencies to implement training programs for which little research evidence exists. Nowhere is

the need for an evidence base more acute than in the area of officer social interaction skills and

de-escalation.

Tothis end, the present study used a randomized controlled trial to evaluate a social interaction

training program in two midsize police departments in the United States. This program involved

the repetitive practice of social interaction skills on a biweekly basis over several months. We

collected data on both treatment and control officers’ priorities in a hypothetical officer–citizen

scenario before and after the training program, as well as official reports of use-of-force incidents

before, during, and after the training program was implemented. The results indicate that the

training program was effective at increasing the priority officers place on procedurally fair com-

munication but was not effective at reducing the number of reported use-of-force incidents for

officers undergoing the training program. Prior to discussing our experimental results, we first

explore the literature on social interaction and related police training and detail the elements of

the training program we evaluated.

1 POLICING, SOCIAL INTERACTION, AND THE USE OF FORCE

A fundamental challenge of police organizations in a democratic society is the exercise of the

social control function in a way that reflects just action and restraint and is consistent with social

and legal expectations. This challenge is most evident in police–citizen interactions that involve

officers using some form of physical force. Although extensive research has been conducted

on police use of force (e.g., Alpert & Dunham, ;Eith&Durose,; Garner & Maxwell,

; Kaminski, Engel, Rojek, Smith, & Alpert, ; Paoline, Terrill, & Ingram, ; Terrill &

Mastrofski, ), most of these efforts have treated these events as static. An interactionist