The MMRMA Deadly Force Project 
A Critical Analysis of Police Shootings  
Under Ambiguous Circumstances© 
 

 
Thomas J. Aveni, MSFP 
The Police Policy Studies Council 
 
 
February 9, 2008 

Structured to pursue quantification of the Graham v. Connor, “reasonable officer” standard, this study 
examined micro‐behavioral components of the deadly force decision‐making process. In doing so it also 
encompassed consideration of specific officer and suspect traits in an effort to determine their degree 
of correlation in the decision‐making process. The study concludes that the age, assignment, experience 
and gender of the officer had little significance in the apparent propensity to use deadly force under the 
circumstances  observed.  The  age,  ethnicity,  gender  and  attire  of  suspects  shot  by  officers  did  suggest 
significant  correlation,  but  in  ways  that  might  radically  alter  the  manner  in  which  the  police  use  of 
deadly force is examined in the future.  

2008 Copyright The Police Policy Studies Council

Table of Contents
Acknowledgements .......................................................................................................................... 1 
Abstract ............................................................................................................................................ 3 
Executive Summary ......................................................................................................................... 4 
Project Genesis and Overview ......................................................................................................... 5 
Prior PPSC Research ................................................................................................................. 6 
Faulkner’s “Response to Resistance” Project............................................................................. 6 
Data Analysis .............................................................................................................................. 8 
Right of Non-Participation ........................................................................................................... 9 
Participant’s Right to Review Video ............................................................................................ 9 
Artificiality of Video Simulation .................................................................................................... 9 
Scenario Acting Considerations ................................................................................................ 11 
Acting Quotient (AQ) ................................................................................................................. 12 
Empty-Hand vs. Object In-Hand Suspect Variations ................................................................ 14 
Participant Isolation ................................................................................................................... 14 
The Return-Fire Stimulus .......................................................................................................... 14 
First Shot Fired.......................................................................................................................... 15 
Agency and Jurisdictional Characteristics ................................................................................ 15 
Data Analysis & Research Findings ............................................................................................... 17 
Action vs. Reaction ................................................................................................................... 18 
Inter-Agency Shooting Variations ............................................................................................. 19 
The Suspect “Surrender” Phenomenon .................................................................................... 20 
Cover from Return-Fire ............................................................................................................. 21 
“Death by Defiance™” .............................................................................................................. 21 
Mugging Scenarios and the Bystander Influence ..................................................................... 22 
The Black Officer Phenomenon ................................................................................................ 23 
Training Implications ...................................................................................................................... 24 
OIS Investigative Implications ........................................................................................................ 26 
Policy Considerations..................................................................................................................... 27 
References ..................................................................................................................................... 29 
APPENDICES ................................................................................................................................ 31 
Biographical Information ........................................................................................................... 32 
Reaction Time Histograms........................................................................................................ 33 
First Shot Fired Scenario Summary .......................................................................................... 35 
Participant DeBrief Form Questions ......................................................................................... 36 
Participant Consent Form ......................................................................................................... 41 

2008 Copyright The Police Policy Studies Council

1

ACKNOWLEDGEMENTS
The nature and substance of this project was years in the making. However, it
may have never come to fruition without the gracious assistance of people who
recognized its potential benefit and then assisted in meaningful ways. The contributions
made by those who supported this project should be recognized from the outset.
The most important benefactor of this study was Bill Page, a senior risk
management specialist with the Michigan Municipal Risk Management Authority. Bill
was receptive to the nature of this study from the very outset and greatly assisted PPSC
in applying for an MMRMA grant. Bill patiently shepherded us through the grant
application process until it was approved. But Bill’s most stellar efforts were exhibited
after the grant was approved. There were countless times in which Bill Page nurtured
this project through many of its biggest logistical hurdles. Had it not been for Bill’s
persistent effort at seeking the substantial material and technical assistance from IES
Interactive Training, this study may not have been possible.
Dr. Robert E. Maurer was greatly instrumental in bringing this project to
completion. He was able to operationalize unstructured conceptualizations of what this
project intended to do. Given the complexity of what we wanted to accomplish, his
expertise in research design and statistical analysis was truly invaluable. Dr. Maurer’s
investment of hundreds of hours, obsessing over the smallest details, was instrumental
in bringing about a meaningful outcome for this study.
Robert McCue, President of IES Interactive Training, must be singled out for his
valuable support for this project. The utilization of a laser-based system was highly
beneficial in this endeavor, and the combination of IES hardware and software afforded
the integration of synchronized video capture and “return-fire” capabilities.
The bulk of the IES hardware employed in this project was generously provided
by Ed Gallagher, proprietor of Michigan Mobile Ranges. In addition, it was Ed Gallagher
who transported and set up the IES MILO system at each location, assured that it
operated properly, and then disassembled and transported it from each location at the
end of each week. The value of Ed’s assistance in this project cannot be overstated and
we are deeply indebted to him for his assistance.
Dave Young and Brent Radford of CAPS™ were crucial resources in the staging,
producing, filming and editing of the 80 research scenarios used in this project. This
was a major undertaking, involving sophisticated, high-resolution, digital camera
equipment, and a Tiffen “SteadiCam™” system that afforded smooth, stabilization of the
action scenes that involved camera movement. The equipment that they brought to bear
for this project was as valuable as was their expertise. We are extremely grateful to
Dave and Brent for their critical assistance at the developmental stages of this project.
Lieutenant Louis DiRenzo (Ret., Camden, NJ Police Department) flew into Grand
Rapids and greatly assisted with the last, crucial week of this study.

2008 Copyright The Police Policy Studies Council

2
Fred Moriarty, proprietor of Creative Woods, a furniture store in Spofford, NH,
generously provided his store to us as a filming location. It served as an excellent venue
for filming the 80 scenarios that were ultimately used in this project. Two other business
locations were approved but not needed, due to the excellent site features afforded to
us at Creative Woods.
Sam Pilo, of the Actors Theatre Playhouse in Chesterfield, NH is credited not
only for his acting roles in this project, but also for his substantial effort in locating and
organizing the diverse range of actors needed for this project. Through Sam’s efforts,
we acquired the acting talents of Teresa Storti, Bill Forcion, Eric Morgan, and Solomon
Muhammad – to all of whom we are also very grateful.
Naturally, this project would never have been possible without the incredible
support of the participating agencies; Grand Rapids Police Department, Ingham County
Sheriff’s Department, Kent County Sheriff’s Department, Livingston County Sheriff’s
Department. Livonia Police Department and the Southfield Police Department. The
chiefs, sheriffs and their designated subordinates who assisted with the implementation
of this study deserve an inordinate amount of credit and gratitude for their work and
their foresight in recognizing the potential value of this study.

2008 Copyright The Police Policy Studies Council

3

ABSTRACT
Prior research on the police decision to use deadly force has tended to focus on
macro-situational variables, such as race, while largely forsaking critical microbehavioral issues that seem to have a significant cumulative effect on an officer’s
decision to employ deadly force. This study focused on multivariate situational and
behavioral variables and their relationship to the frequency with which participating
officers utilized deadly force. This paper employs data gathered from 307 officers from
six Michigan law enforcement agencies, and employs multivariate analyses to identify
situational and behavioral predictors of police shootings involving felony-in-progress
suspects, under circumstances that were purposely designed to appear ambiguous.
Findings from the analyses are then used in an effort to develop predictive
classifications as to when the use of deadly force might be deemed “objectively
reasonable.” In identifying situational and behavioral predictors of the threshold at which
officers might employ deadly force, we may begin to better understand the behavior of
officers forced, by the nature of their assignment, to make critical, split-second decisions
that may result in the taking of a life.

2008 Copyright The Police Policy Studies Council

4

EXECUTIVE SUMMARY
There were 307 officers/deputies from six agencies that participated in this study,
engaging a total of 117 “unarmed” suspects. In examining whether there was correlation
between officer’s attributes (race, sex, age, experience, type of duty location) and an
officer’s action, we found that no significant correlation exists. Nor was there any
significant correlation between officer action and order of scenario videos, subject race
or subject sex.
Significant correlation does exist between officer action and action of the subject
(shoot, surrender with object in-hand, and surrender without object in-hand), acting
quotient, and video setting (burglary, robbery, and mugging). Also, significant correlation
exists between an officer’s action and two attributes of the subject – subject’s age and
subject’s dress.
Officers/deputies are more likely to shoot when the subject is young (rather than
old), in punk dress (rather than business dress), and acting quotient is high (rather than
low). Officers are also more likely to shoot in robbery scenarios than in muggings and
more likely to shoot in mugging scenarios than in burglary scenarios. Lastly, officers are
more likely to shoot when a subject's action is “shoot” than when a subject’s action is
“surrender without,” and more likely to shoot when a subject’s action is “surrender
without” an object in-hand than when it is "surrender with an object in-hand." This
apparent anomaly is explained by the high correlation found in scenarios which had
been assigned high “acting quotients” (AQ) for the amplitude of critical acting variables.
The higher the acting quotient, the higher the correlation was for officers shooting
“unarmed” suspects.

2008 Copyright The Police Policy Studies Council

5

PROJECT GENESIS AND OVERVIEW
While many professions embody a preponderance of decisional absolutes, there
are professions that require a significant amount of discretionary latitude. Law
enforcement officers are entrusted with a considerable degree of discretion, which is
essential in navigating a world of human behaviors and circumstances that are often
ambiguous. However, such latitude often creates greater hesitancy and indecision while
an officer attempts to determine the “appropriate” response to a given situation.
Most recent research one finds on the police use of deadly force has been
focused on what police do more than upon why they do it. Social scientists who have
wandered aimlessly into this research realm have done so largely ill-prepared to
discover anything of practical value.
Race has been explored extensively as a factor in the “disproportionate” police
use of deadly force, but as a research variable, race has almost always been
manipulated without meaningful context (Correll, Park, Judd, & Wittenbrink, 2002;
Greenwald, Oakes, & Hoffman, 2003; Payne, 2001). One is left to wonder why this
tendency persists. This study embedded officer and suspect race as research correlates
simply because their inclusion has become almost obligatory.
While many previous social science research endeavors have focused keenly
upon the issue of “stereotype congruence/incongruence” in officers’ decision-making
processes, this study delved deeply into the generally overlooked micro-issues that are
the very essence of the police decision-making process. Where race appears to play a
critical role in police decision-making it may well be attributed to behavioral nuance that
is perceived as being defiant and even threatening by police. Weitzer and Tuch (2004)
presented evidence that members of ethnic minorities often feel as if they’ve been
mistreated by the police (see also Sunshine & Tyler, 2003). The implication has been
that the police are racist and that officers use excessive force against minority suspects.
In response, it has been suggested that black people may engage in more
confrontational behavior toward police officers, perhaps adding to a cycle in which
hostility toward police might prompt more severe applications of force by police (Reisig,
McCluskey. Mastrofski, & Terrill, 2004).
Does defiant behavior toward authority contribute to the frequency of police
shootings – especially under ambiguous circumstances? That question appears to be a
legitimate one to ponder, since “death-by-defiance” (to police) seems to be of predictive
value when situational and behavioral cues are extreme in nature. One component of
this study attempted to quantify whether defiance (i.e., non-compliance to officer/deputy
commands) contributed to the officer/deputy perception of an imminent threat. After a
literature review of relevant research, we decided to establish a different research
pathway and not rely on the synthesis and/or reinterpretation of previous studies.

2008 Copyright The Police Policy Studies Council

6

The genesis of this study actually stems from consideration of an ACLU
allegation made on their website in 1998 (http://www.aclu.org/police/abuse/index.html):
that 25% of all suspects shot by police are, “unarmed and not-assaultive.” One of the
three corroborative studies cited by the ACLU was conducted by Marshall Meyer
(1980). Meyer analyzed data provided by the Los Angeles Police Department from
1974-78 and reached a myriad of interesting conclusions about the LAPD’s use of
deadly force. While noting disparities in the number of black suspects shot by LAPD, he
noted;
“In almost all instances, the suspect’s act precipitating a shooting incident
is the final act that caused the officer to fire, that is, that act but for which
the shooting would not have taken place.” pg.101
It should be noted that the three studies (Meyer 1980, Kobler 1975, Milton et al
1977) cited by the ACLU as being illustrative of the frequency in which unarmed
suspects were shot by police, all predated the more restrictive legal constraints imposed
by Tennessee vs. Garner. It should also be noted that although the ACLU “police
abuse” webpage no longer asserts that 25% of unarmed suspects shot by police are
unarmed, it still promotes a variety of sources that are clearly dated.
Prior research (M.D. White, 2002) has suggested that certain types of police
assignments are highly correlated with the police use of deadly force. In this study
however, the presence of a lethal threat was largely implied by behavior, and to a lesser
extent, the nature of the scenario. This “threat ambiguity” was embedded as a critical
design feature. We were focused upon defining what a “reasonable officer” would do
when scenario characteristics were not clearly defined – much the way that they aren’t
in many of the more controversial police shootings.
Prior PPSC Research
In a multi-year study of shootings in Los Angeles County (Aveni, 2003), we found
that 18% of all officer involved shootings were of “unarmed” suspects. We excluded
incidents in which subjects had replica firearms, pellet guns or any other device that
convincingly looked like a weapon to the officer/deputy. We also excluded incidents in
which motor vehicles were used as lethal weapons against officers/deputies.
Investigative narratives regarding the 148 shootings examined (1998-2002) indicated
that over half of the officers reported that they used deadly force after non-compliance
with verbal commands and after suspects then exhibited “furtive movement.” These
findings influenced the structural design of this study.
Faulkner’s “Response to Resistance” Project
Lieutenant Sam Faulkner, Mechanicsburg (OH) PD, began conducting a use-offorce survey in 1990 that he now claims has been completed by more than 40,000
police officers. Lt. Faulkner’s stated mission has been to diminish the ambiguity of the
“reasonable officer” standard derived from Graham v. Connor. This effort represents a
significant effort of substantial value. It should be noted, however, that Lt. Faulkner’s
survey addresses a broad spectrum of use-of-force issues, without specific focus on
complex deadly force issues.

2008 Copyright The Police Policy Studies Council

7
The PPSC study took an alternative, experiential approach because we believe
that any self-reporting mechanism is inherently flawed for most serious research
applications. If one were to submit a survey to officers, asking them whether or not
they’d shoot an unarmed suspect under certain circumstances, chances are that few
would respond in the affirmative.
Where this study does employ a self-reporting mechanism (i.e., “Officer Debrief”
forms), it does so after the participating officer has already exhibited what his/her
judgment was, under realistically ambiguous and somewhat tense circumstances. In
essence, we asked officers not what they would do under specific circumstances; we
asked why they did what they had already done in a specific scenario exposure.

2008 Copyright The Police Policy Studies Council

8

Project Structure
Each participating officer/deputy was randomly subjected to a total of three
scenarios, with one of each of three types of scenarios:
•
•
•

Robbery-In-Progress
Burglary Alarm-In-Progress
Mugging-In-Progress

Of these three scenarios listed above, all officers were also randomly exposed to
three different outcomes of each scenario:
•
•
•

Suspect empty-hand surrender
Suspect surrender with (non-weapon) object in-hand
Suspect shoots at officer

To further determine whether outward appearance might influence a participant’s
tendency to use deadly force, we also embedded random variations of suspect attire in
the scenarios. In all scenario varieties, suspects were filmed in both “dress” and “punk”
attire. “Dress” attire was usually a shirt and tie, often with a dress jacket or coat as an
outer garment. “Punk” attire varied from individual to individual, based on what actors
brought to the film session. Clothing ranged from blue jeans, hooded sweatshirts and
leather jackets.
From a methodological standpoint, this study wasn’t structured to employ a
random sampling of officers/deputies from participating agencies. Given the manpower
constraints encountered during the summer months that this project entailed, this study
pursued the involvement of at least half of each agency’s patrol personnel. To
accomplish this goal, research scenarios were run from 10:00AM to 10:00PM to
straddle at least two patrol shifts on each day. On some occasions participation
stretched well into the midnight shift when it appeared that additional participants might
be made available.
During the initial orientation given to each police participant, all were told that
they were about to embark upon a data collection exercise, not a training exercise. They
were also told that because it was a data collection exercise, no value judgments would
be offered before, during or after the projection of each scenario.
Data Analysis
Employing stepwise regression and statistical correlation of variables, we
compiled facilitating data in the following categories:
•
•
•

Hesitation or proclivity to use deadly force
o Age/Assignment/Experience/Gender/Race of Officer
Hesitation or proclivity to shoot based on outward appearance of suspect
o Age/Attire/Dress/Ethnicity/Gender of suspect
Hesitation or proclivity to shoot by perceived nature of scenario
o Burglary/Mugging/Robbery

2008 Copyright The Police Policy Studies Council

9
•

Hesitation or proclivity to shoot by the perceived nature of subject’s
actions
o Non-compliance with officer’s commands
o Concealment of subject’s hands in conjunction with “furtive
movement”
o “Targeting glance” over subject’s shoulder at officer
o Rapid subject turn toward officer

Right of Non-Participation
All participants were required to read and sign a “Consent to Participate In
Research” form (See Appendix) before they were exposed to any scenarios. Several
participants were turned away when they voiced reservations about taking part in this
study. The importance of voluntary participation was stressed against a frequent
perception (by participating officers) that they were being compelled by their
administration to participate. Officers who expressed reservations about participating
were told that they had been “rejected” for the study as a means to spare them any
discomfort with their administration for non-participation. In all, eight officers/deputies
were excluded from taking part in this study due to reservations they expressed about
participation.
Participant’s Right to Review Video
In recognition of a raging controversy about whether officers should be permitted
to view dash-cam video of an officer-involved shooting before being compelled to
provide an oral or written statement to agency investigators, this project was structured
to permit participants the right to review their actions in each video before being tasked
with completing a “Debrief Form” for each completed scenario. ALL participants opted to
review the scenarios in which they used deadly force but typically showed much less
interest in reviewing the scenarios in which they did not employ deadly force. It is worth
noting that when participants opted not to review a video replay of their performance
they typically had difficulty remembering many of the situational and behavioral
elements embedded in those scenarios. Subsequently, there was an unsatisfactorily
high number of “N/A” Debrief Form responses when participants opted not to review a
completed scenario. The “N/A” response was an option given to participants when they
thought a debrief question was either “not relevant” to what they saw or when they
couldn’t recall that element being embedded in the scenario.
Artificiality of Video Simulation
Having used video firearms training simulators for many years, we had some
expectations regarding the likelihood of getting some measure of disingenuous
decision-making from artificial situational and behavioral stimuli. For instance, a
subject’s age, attire and gender may influence a cognitive bias on the street more so
than in a simulated training environment. The reasons for this are simple. For many
years, video simulation exercises have stressed the importance of not making
“assumptions” based on a subject’s outward appearance. Consequently, a substantial
number of training videos have been introduced involving female assailants, or older
male assailants who commit violent crimes and assault police with deadly weapons. In
video simulations, the prevalence of such “stereotype incongruence” has become

2008 Copyright The Police Policy Studies Council

10
somewhat routine – perhaps much more than encountered on the street, if Uniform
Crime Report data are used as a standard of comparison. Consequently, officers are
more acclimated to the notion that “anything goes” in a video-based training
environment than they might be accustomed to expect on the street.
Another common criticism encountered pertinent to using video simulation for
research purposes is the notion that officers/deputies will “shoot” in a simulation
environment much more readily than they might on the street. The consensus about
why this might occur seems to be that there are few negative consequences when
deadly force is used in simulation. The degree to which this concern might influence the
data derived from the study is probably impossible to quantify. However, to some
degree we believe that this concern was mitigated by the apprehension that
officers/deputies expressed about their participation being video-taped. Many
participants voiced concern about any “mistakes” they might make in the study would be
scrutinized and potentially used against them in the future. Although this concern was
addressed through reiteration of a contractual assurance of participant confidentiality, it
likely remained as a constraining influence on each participant’s propensity to use
deadly force in this study.

2008 Copyright The Police Policy Studies Council

11

Scenario Acting Considerations
The material used to conduct this research was hybrid in nature. The scenarios
were scripted in such a way that they might elicit a variety of participant responses
within a controlled sequence of delivery.
Actors were hired from a small local theater and briefed on the nature of the
project. They were told that they’d have virtually no oral script, and that all acting would
be confined to what they could convey through their scripted physical activities. Prior to
filming the scenarios, all actors were instructed on how to convincingly convey their
assigned roles of criminal activity. They were given a choreographed session of “furtive
movement” instruction which entailed the sequence of veiled hand movements,
“targeting glances,” “digging” for secreted weapons, and nuanced hand movement as
they pivoted toward the camera. Hand movement was practiced with and without
objects in their hands.
Actors were instructed to initiate all hand movement (1) from waist level when
making their final turn toward the camera and (2) that their hand movement should
thrust forward and then up from the waist position as they completed a full turn toward
the camera. This procedure was to be adhered to in all scenario variations that didn’t
involve the actors turning toward the camera with a firearm. In the scenario variations
that did involve the actors turning toward the camera with a gun in-hand, they were
instructed that hand movement was to begin at waist level when they were initiating the
turn toward camera and end with one or both hands outstretched in a two-handed firing
position when they were completing their turn toward the camera. For acting evaluation
purposes, it was emphasized that at least one hand (the “gun hand”) maintain a waistlevel disposition through the first half-turn toward the camera.
Choreograph things as well as you might attempt to, actors seldom move and
turn in similar ways. Some exhibited fluid movement, others not. Some initiated their
turns with movement in the hips, buttocks, and/or shoulders. Others seem to initiate
their turns by noticeably shifting their weight to the foot from which they’d pivot. It is
important to note that many of the discernable, yet subtle, body movements (isolated in
retrospective video analysis, noted preceding suspect turns) toward the camera could
have been missed as behavioral cues indicating that a turn toward the camera/officer
was being initiated.
Likewise, many of the suspect’s subtle movements toward the camera, which
weren’t directly connected to his/her turn toward the camera, could have been
misinterpreted by officers as initial movement into a turn toward them. In each scenario,
actors were instructed to look over their shoulder at the camera at least once and then
pause for a few moments before committing to a full-body turn toward the camera.
Officers often learn or are trained to perceive such over-the-shoulder glances by
suspects as “targeting glances.” The degree to which such rearward glances involve
more substantive shoulder and/or hip rotation toward the officer might influence an
officer’s perception that a turn is being initiated when, in fact, it might not be. Since
much of this process of perception is context-driven, one might reasonably surmise that

2008 Copyright The Police Policy Studies Council

12
movement toward the officer would be interpreted differently within the context of a
robbery-in-progress call than it would within the context of a larceny-in-progress call.
The one common thread embodied in all turns in all scenarios was that they were
initiated toward the camera as left-turns.
Given the fact that some turns were “more compelling” than others, one might
reasonably suspect that certain actor’s turns elicited higher shooting responses than did
other actor’s turns (see illustration below).
Acting Quotient (AQ)
In an effort to pursue more meaningful data interpretation – primarily to explain
(or predict) certain anomalies in the way that participating officers reacted to specific
scenarios, we scored acting qualities for “values” that appear to add varying levels of
“compelling quality” to each scenario. For instance, if an actor turned more quickly
toward the camera in one scenario than another, a participating officer might react to
that turn with more belief that it represented the initiation of a threat than if that turn had
been less rapid. Likewise, if the actor turns with his/her hands at waist level – where a
secreted weapon is believed to have been accessed, the officer might be more inclined
to shoot than if the actor turned with his/her hands held high, in more of a “surrender”
posture. AQ values are scored cumulatively and it is assumed that a higher AQ
cumulative value (3-4) will result in a higher frequency of “unarmed” suspects being shot
than in scenarios with lower AQ cumulative scores (0 - 2.5). For example, a turn
characterized by E,LH,PC, CH would have an AQ score of 3.5.
Actor’s Quotient Values
Actor’s Action
Tepid Turn
Energetic Turn
High Hand(s)
Low Hand(s)
Upright Stance
Partial Crouch
Full Crouch
Open Hand(s)
Half-Closed Hand(s)
Closed Hand(s)

Symbol
T
E
HH
LH
U
PC
FC
OH
HC
CH

Value
0
1
0
1
0
.5
1
0
.5
1

An anomaly that wasn’t fully considered until the research project began was the
realization that the actors behaved differently in those scenarios in which they were
given a handgun and instructed to turn and fire at the movie camera. None of the five
actors utilized identified themselves as recreational shooters – or even gun owners. All
were given an orientation with the .38 Smith & Wesson M640 revolver used in the
“armed” scenarios. The actors were not instructed in the intricacies of “combat
shooting,” they were merely familiarized with function and safety of that specific

2008 Copyright The Police Policy Studies Council

13
handgun. It came as a surprise, when later reviewing each filmed scenario, that actors
behaved noticeably differently with the handgun in-hand. They tended to turn with more
of a body crouch, with their shoulders forward and knees somewhat bent. In addition, a
grimace was somewhat noticeable on actor’s faces when turning with a handgun. There
may have been a practical reason for that, since the full-flash “Hollywood Blanks”
utilized for effect were loud and bright during night-time filming. Whatever influenced the
actors to grimace, it added more visual emphasis to their turn toward the camera. The
combination of crouching and grimacing contributed to a more “intense” look when
actors turned with, and fired, a handgun.
It is essential to note that suspect hand movement – even that which seemed
tepid, as he/she turned toward the camera, was almost always too fast to determine the
nature of any object being held in the suspect’s hand. Suspect hand movement, as
viewed from the camera lens, almost always encompassed 4-5 feet of an arc of
movement toward the camera. With that much viewed distance of hand movement
being covered in one second or less to achieve (1) at least 90% completion of a full turn
or (2) the suspect firing his/her first shot at the camera, the latency period of saccadic
eye movement falls far behind the rapid movement of the hands. This visual
phenomenon becomes apparent when rapid movement is perceived as a blur or a
“smear” of motion. Subsequently, it was often difficult for participants to discern when a
suspect’s hands were not holding any objects until rapid hand movement terminated.
This tends to explain why some participants shot suspects who turned to “surrender”
with empty hands. The manner in which some suspects turned and the context in which
they were viewed (i.e., the type of crime that they were perceived committing), often
compelled participants to shoot before there was any certainty about whether the
suspect was armed.
In addition to movement of the hands, another substantive hand-variable was
noted after all scenarios were filmed and reviewed. Actors were not instructed when to
“open” their hands when turning toward the camera into a “surrender” position.
Subsequently, some actors initiated their turns toward the camera with open hands
while others initiated turns with clenched hands. Even with the rapid hand movement
evident in all scenarios, the difference was almost always readily apparent. And, the
difference is significant to trained eyes. An open hand is perceived as a much less
threatening hand since it is almost immediately recognized as an “empty hand.” An
empty hand is perceived as a weaponless hand. A clenched hand, especially when
viewed in rapid movement, exudes significant behavioral ambiguity. A rapidly moving
clenched hand is much less likely to be viewed innocuously if the situational context is
severe.
For the scenarios involving the (actor’s) use of deadly force (which was 1-in-3 of
all scenarios), we gave an orientation to the actors relative to using a Smith & Wesson
M640 revolver in .38 Special caliber. The revolver was being loaded with “full-flash”
Hollywood blanks for each scenario that involved the “suspect” using deadly force.
What this study did not originally account for was the manner in which actors
appeared “stimulated” whenever they acted out scenarios that involved them holding/
firing a handgun. They appeared to turn (toward the camera) at roughly the same pace,
but they more frequently turned in a crouching posture, with hands more clenched.

2008 Copyright The Police Policy Studies Council

14
Even facial expressions changed in substantive ways that might be perceived as being
“more aggressive.” Actors often appeared to grimace in anticipation of the loud report
and flash of the blank cartridges that were fired from the handgun. But, to a trained
observer, that same grimace might be perceived as an expression of defiance and
determination.
Empty-Hand vs. Object In-Hand Suspect Variations
In an attempt to determine the nature in which suspects might be shot while
unarmed, we filmed two variations of the unarmed scenarios. In all three types of
scenarios (burglaries, muggings, robberies), actor/suspects were filmed with firearms in
one variation, innocuous objects (cell phone, flashlight, police ID wallet) in another
variation and empty-handed in yet another variation. Due to the eye’s (saccadic) latency
period in tracking rapidly moving objects, it was believed that objects held like weapons
would in fact be confused for weapons when actors were moving rapidly, and under the
low light conditions in which all scenarios were filmed. Since at least 71% of all
“mistake-of-fact” police shootings occur at night (Aveni, 2002), this research element
was deemed critically relevant.
Participant Isolation
Each participant’s performance was kept confidential by isolating his/her
involvement behind closed doors. This provided the added benefit of keeping scenario
substance under wraps until individual participants could be debriefed about their
possible involvement in the project.
Unknown to participants, all research scenarios had common elements
embedded within them. Suspects, when confronted by police, initiate movement toward
the officer with their back initially facing the officer. This variable was purposely
embedded in scenarios to add to the ambiguous nature of each suspect’s veiled
movement. As each suspect’s back is facing the officer’s view, suspect hand movement
is obscured until he/she nearly completes a full rotational turn toward the officer. When
suspect role players turn toward officers at more brisk speeds, their hands appear to
blur due to the latency period of saccadic eye tracking of moving objects. Subsequently,
it is often difficult to determine an empty hand from one that isn’t. For the same reason it
is also extremely difficult to discern the nature of any object being held by the suspect
as he/she turns toward the officer. Suspects turning with cell phones, flashlights or
police ID were often shot by officers for this reason.
Some officers began recognizing the behavioral commonality after completion of
their first two scenarios. Others didn’t grasp this even after completing all three
scenarios. However, for the sake of the project’s integrity, each participant was asked
not to discuss the nature of what they had observed with anyone who hadn’t already
been tested. They were further advised that (participant) scenario assignment was
random and that no two participants from any agency were likely to have the same
combination of scenarios.
The Return-Fire Stimulus
Generally associated with the use of video-based training (“gunfight”) simulators
is the perceived lack of threat-induced stress. Officers seldom seem motivated at levels

2008 Copyright The Police Policy Studies Council

15
even remotely comparable to what they might experience in similar scenarios on the
street.
The IES-MILO system that was utilized in this study afforded a shoot-back
capability in the form of a slightly modified paintgun that fired .68 foam-rubber
projectiles. Though launched at velocities of roughly 300 fps (measured within 3’ from
the muzzle), the foam projectiles only cause mild discomfort when officers are struck by
them. Paintgun masks that afforded full wrap-around face protection were mandated for
wear by all participants.
The manner in which the return-fire stimulus was used was never fully explained
to participants. They knew that the device was fired remotely but were not told that the
device was deliberately aimed at the “cover” that they had been afforded – not at
participants. However, the possibility of being struck by a projectile was always present.
Perhaps as many as five participants were struck by projectiles – most were struck in
arm or shoulder areas that were exposed from behind cover. There were no resultant
injuries, of any kind.
First Shot Fired
In each scenario that involved a suspect’s shot being fired toward the camera
(officer), there was a specific frame of that video in which the first shot was identified by
way of the flash signature of that shot. By identifying when the first shot was fired in
each deadly force scenario we were able to then identify whether each participating
officer’s reactionary shot was fired before or after the “suspect’s” first shot. In was
anticipated that officers who resist the reflex to shoot at a rapidly turning suspect who
(later) is determined to be unarmed might also find themselves exhibiting that same
hesitation when the suspect is armed. If this projection were found to be true, we’d also
likely find that the officers who didn’t shoot an unarmed suspect were also hesitant
enough to allow the suspect to fire the first shot due to that additional time spent in
deciding to shoot.
Agency and Jurisdictional Characteristics
Prior research suggests that extreme police-citizen encounters typically occur
within the community in which both participants live and/or work, and characteristics of
that environment influence perceptions and behavior during such encounters (Kania
and Mackey, 1977; Fyfe, 1980, 1988; Geller and Karales, 1981a; Alpert, 1989). In
addition, Kania and Mackey (1977) concluded that there is a significant relationship
between police aggression and police exposure to violent crime. In this vein,
consideration was given in the developmental stages of this study to potential
differences in the way that participants from varying agencies might react to similar
scenarios. Initially, we had contemplated integrated comparisons of each agency’s
Calls-For-Service (CFS) data, along with a comparison of UCR data (e.g., homicides
per 100,000, aggravated assaults per 100,000, etc.). Though requested, CFS data was
not forthcoming from participating agencies. A fallback approach was utilized,
employing a rudimentary metric that was already embedded in the study.
All agency participants were categorized by their individual assignment type, into
four classifications; (1) Rural, (2) Suburban, (3) Urban and (4) Detective. Deputies from

2008 Copyright The Police Policy Studies Council

16
Kent, Ingham and Livingston counties were asked to describe the nature of their current
duty assignments to determine which classification was most appropriate. Officers from
Southfield and Grand Rapids were similarly classified by assignment, but either into
Urban or Suburban classifications.
This study was originally structured to omit any personnel not currently assigned
to “patrol” assignments. However, given the challenging nature of personnel scheduling
in summer months, two agencies provided a substantial number of investigators for
testing purposes. While all participating detectives/investigators indicated that they had
prior patrol experience, many were years removed from that experience and were
therefore classified as “Detectives.”
As our findings later seemed to suggest, assignment classification didn’t reflect
any significant statistical difference in participant decision-making.

2008 Copyright The Police Policy Studies Council

17

DATA ANALYSIS & RESEARCH FINDINGS
No significant correlation exists between officer action and any officer attribute
(race, sex, age, experience, type of duty location), nor between officer action and order
of videos, subject race or subject sex.
Significant (Alpha = .05) correlation exists between officer action and action of
the subject (shoot, surrender with object in-hand, and surrender without object in-hand),
acting quotient, and video setting (burglary, robbery, and mugging). Also, significant
correlation exists between officer action and two attributes of the subject – subject’s age
and subject’s dress.
Officers/deputies are more likely to shoot when subject is young (rather than old),
in punk dress (rather than business dress), and acting quotient is high (rather than low).
Officer is also more likely to shoot with video setting as a robbery than as a mugging
and more likely to shoot with video setting as a mugging than as a burglary. Lastly,
officer is more likely to shoot when subject's action is “shoot” than when subject’s action
is “surrender without,” and more likely to shoot when subject’s action is “surrender
without” than when it is "surrender with."
Correlation for All Scenarios, Subjects Armed and Unarmed
Officer Shoots

Cor

Abs (Cor)

Num

Denom

tcalc

ttable

Sig?

Shoot
Act Quotient
Surrender W/O
Old Suspect
Robbery
Burglary
Punk Dress
Officer-Other
Second Video
Officer-Black
First Video
Urban Officer
Suspect Female
Officer Yrs. Exp.
Rural Officer
Officer-Male
Officer-Age
Suspect-Black
Officer-Hispanic
Suburban-Officer

0.7447
0.4832
-0.3176
-0.1278
0.0771
-0.0677
0.0653
0.0483
-0.0473
0.0340
0.0328
-0.0310
-0.0294
0.0202
0.0198
-0.0196
0.0165
0.0123
-0.0093
-0.0066

0.7447
0.4832
0.3176
0.1278
0.0771
0.0677
0.0653
0.0483
0.0473
0.0340
0.0328
0.0310
0.0294
0.0202
0.0198
0.0196
0.0165
0.0123
0.0093
0.0066

22.4262
14.5538
9.5638
3.8493
2.3234
2.0377
1.9663
1.4535
1.4245
1.0237
0.9873
0.9334
0.8866
0.6080
0.5962
0.5891
0.4957
0.3692
0.2805
0.1982

0.6675
0.8755
0.9482
0.9918
0.9970
0.9977
0.9979
0.9988
0.9989
0.9994
0.9995
0.9995
0.9996
0.9998
0.9998
0.9998
0.9999
0.9999
1.0000
1.0000

33.5997
16.6237
10.0859
3.8811
2.3303
2.0423
1.9705
1.4552
1.4261
1.0243
0.9878
0.9339
0.8870
0.6081
0.5963
0.5892
0.4958
0.3692
0.2805
0.1982

1.9626
1.9626
1.9626
1.9626
1.9626
1.9626
1.9626
1.9626
1.9626
1.9626
1.9626
1.9626
1.9626
1.9626
1.9626
1.9626
1.9626
1.9626
1.9626
1.9626

yes
yes
yes
yes
yes
yes
yes
no
no
no
no
no
no
no
no
no
no
no
no
no

2008 Copyright The Police Policy Studies Council

18

Correlation for Instances Where Subject was Unarmed
Officer Shoots

Act Quotient
Old Subject
Robbery
Punk Dress
Officer-Black
Burglary
Surrender W
Officer other
First Video
Black Subject
Female Subject
Urban Officer
Officer Male
Rural Officer
Second Video
Officer Yrs Exp.
Officer Age
Officer Hispanic
Suburban Officer

Cor

Abs(Cor)

Num

Denom

tcalc

ttable

Sig?

0.2263
-0.1953
0.1489
0.1363
0.1093
-0.1076
0.1006
0.0884
0.0837
-0.0543
-0.0534
-0.0511
-0.0465
0.0399
-0.0353
0.0325
0.0250
-0.0207
-0.0193

0.2263
0.1953
0.1489
0.1363
0.1093
0.1076
0.1006
0.0884
0.0837
0.0543
0.0534
0.0511
0.0465
0.0399
0.0353
0.0325
0.0250
0.0207
0.0193

5.5671
4.8035
3.6628
3.3535
2.6875
2.6467
2.4741
2.1745
2.0595
1.3350
1.3144
1.2578
1.1446
0.9810
0.8687
0.7990
0.6161
0.5095
0.4747

0.9740
0.9807
0.9889
0.9907
0.9940
0.9942
0.9949
0.9961
0.9965
0.9985
0.9986
0.9987
0.9989
0.9992
0.9994
0.9995
0.9997
0.9998
0.9998

5.7154
4.8978
3.7041
3.3851
2.7037
2.6622
2.4868
2.1830
2.0667
1.3370
1.3163
1.2594
1.1458
0.9818
0.8692
0.7994
0.6163
0.5096
0.4748

1.9639
1.9639
1.9639
1.9639
1.9639
1.9639
1.9639
1.9639
1.9639
1.9639
1.9639
1.9639
1.9639
1.9639
1.9639
1.9639
1.9639
1.9639
1.9639

Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
No
No
No
No
No
No
No

Action vs. Reaction
As one might anticipate, the lag-time between when the officer made his/her
decision to fire and when the initial shot is fired has critical implications. If an officer
finds him/herself behind the reactionary curve in a rapidly evolving situation, the lagtime associated with decision-making can allow the suspect to fire one or more shots at
the officer before fire can be returned by the officer. Another critical lag-time
consideration becomes manifest when an officer fires at a person who ultimately
becomes known to be an unarmed person. As suspects turn toward the camera (officer)
they may appear to be armed as the turn is initiated - the suspect’s hands may be
clenched and/or positioned at or near waist level. However, since the officer’s decision
to fire at the suspect predates the subject being shot .25 seconds or more, the officer
can (and easily does) shoot the suspect as he/she is raising his/her hands into a
“surrender” position. This was a frequent and somewhat unanticipated outcome in many
of the shootings that involved “unarmed” suspects; suspects getting shot while
“surrendering.” The officer typically has 1/3 of a second or less (from a critical juncture
in each scenario) to decide whether or not to employ deadly force, and then to apply
that force, before he/she risks being “shot.”
Those officers who managed to shoot armed suspects before the suspect was
able to fire seemed to have elected to use deadly force before it could be clearly
determined that the suspect did, in fact, have a handgun. This tendency to employ
deadly force “preemptively” was at the core of our objective to quantify when such
actions were objectively reasonable. As anticipated, most officers found themselves
firing after the suspect fired his/her first shot at the officer.

2008 Copyright The Police Policy Studies Council

19

Officer’s Reaction to When Suspect Fires
Officers shoots after subject shoots
Officer shoots before subject shoots

Frequency
61.44%
38.56%

As one might expect, officers seemed more vigilant and situationally postured
when they were informed that they were responding to a forcible felony-in-progress.
Participants typically had their handguns drawn sooner in the robbery scenarios and
were more inclined to utilize the cover that was afforded to them. When handguns were
drawn and used in conjunction with a verbal challenge in the robbery scenarios,
participant muzzle dispositions were notably higher (i.e., more elevated toward the
suspect). And yet, even this higher state of situational readiness seemed to offer little
more than a statistical “dead-heat” to the suspect’s first shot fired.
Officer Reaction Time
in Relation to 1st Shot Fired by Suspect
Measure
Burglary Robbery
Mugging
Mean

0.166323

0.001

0.035897

Median

0.166667

0

0.066667

Since the “mugging” scenarios were actually officer-initiated scenarios, with little
situational clarity, participants generally exhibited an intermediate level of readiness.
Handguns weren’t unholstered as quickly as they were in the robbery scenarios and
muzzle orientation appeared somewhat lower when handguns were drawn. Not
surprisingly, participant reaction times were noticeably slower than what they were in
the robbery scenarios. And, vigilance and situational readiness in the “burglary”
scenarios was measurably worse than in both the robbery and mugging scenarios. It
should be noted that the burglary scenarios were actually given to participants as “alarm
activation” scenarios. It was only until after the mid-point in the burglary scenarios that it
became apparent that they were actually burglary-in-progress incidents. Situational and
behavioral ambiguity was purposely embedded in these research scenarios to enable
measurement of each participant’s interpretation of their value. (See “Reaction Time
Histograms” in Appendix)
Inter-Agency Shooting Variations
The 307 officers/deputies participating in this study shot a total of 117 unarmed
suspects – a frequency of 38% (0.3811). However, there was a significant difference
between one agency on the low end of the frequency scale and agencies on the higher
end. It should also be noted that the agency with the lowest frequency of participants
shooting unarmed suspects and the agency with the highest frequency of unarmed
suspects shot were both largely classified as being “urban” agencies. (See table below)

2008 Copyright The Police Policy Studies Council

20

UNARMED
FREQUENCY
SUSPECTS SHOT
1
76
18
24%
2
46*
18
39%
3
51
20
40%
4
38*
16
42%
5
54*
24
44%
6
43
21
49%
*Indicates that individual participants shot more than one unarmed suspect
AGENCY

PARTICIPANTS

The question will undoubtedly arise; “What noted differences were there between
the agency with the lowest frequency of shootings (of unarmed suspects) and those
with the highest frequency?” The answer, simply put; “It was a difference in training.”
(See “Training Implications”)
The Suspect “Surrender” Phenomenon
The action-reaction disparity previously mentioned was instrumental in a
substantial number of suspects being shot while in a “surrender position” (i.e., empty
hands raised overhead). All non-deadly force scenarios that we filmed culminated in a
raised-hand, “surrender” position, which we defined as having the (suspect’s) hands
held at sternum height or higher, palms facing forward, fingers pointed mostly upward.
By this definition, 92.04% of the unarmed suspects who were shot by officers were shot
while in the surrender position. This was an oblique yet wholly predictable outcome of
this study. Due to the highly compressed timeframe in which officers had to initiate
and/or halt action to raise/fire their handgun, or to terminate a firing impulse if it became
apparent that the suspect’s actions didn’t warrant a deadly force response, the
difference between shooting an armed suspect and shooting an unarmed suspect was
usually delineated in fractions of a second. .
On average, each scenario had a critical decisional juncture in which the officer
had about 1/3 of a second to the suspect’s aggressive movement (i.e., 180° turn toward
the officer). In after-action review of each participant’s video playback, it became
apparent that those officers who managed to shoot unarmed suspects before the
suspect was able to “surrender” seemed to have elected to use deadly force either (1)
before the suspect initiated his/her turn toward the officer, or (2) at the earliest possible
juncture in which the initiation of a turn toward the officer was perceived. Another issue
worth noting was a common tendency for officers to shoot or continue shooting after an
unarmed subject completed his/her turn. There are two plausible explanations for this
phenomenon. The first is that it takes time to “apply the brakes” of a neuromuscular
response (i.e., firing a handgun). Recent research (Lewinski, 2003) suggests that
officers can, on average, fire two or more shots after they’ve been given a visual ceasefire stimulus. In addition, if an officer is engaged in a multiple shot firing string, he/she
may have greater difficulty in terminating a firing sequence. Relevant to that, many of
the officers in this study fired 4-12 times in apparent response to the on-screen “threat”
remaining erect and not going down.
In this study, since scenarios did not have a branching capability, subjects shot
by officers did not go down when shot. Any officer trained to “fire until your foe falls”

2008 Copyright The Police Policy Studies Council

21
would likely continue firing until he/she saw the subject projected on the screen go down
– which wasn’t a design characteristic in this study. Therefore, a longer shot-string could
have been anticipated. In addition to the participant’s perception of a persistent threat
(i.e., one that doesn’t fall when shot), there is another important firing phenomenon to
consider; when the handgun of the officer/deputy is brought up to point-shoulder-aim, it
occludes vision of hand movement until the (suspect’s) hands are raised to
approximately shoulder height. In effect, the participant literally couldn’t see whether the
suspect’s hands were empty – once the officer/deputy had committed to firing from
point-shoulder-aim.
If a follow-up study were conducted, in pursuit of similar research objectives, the
use of branching scenarios would be beneficial for a number of reasons. At least three
of the participating agencies employ video-based simulators with branching capabilities
and many officers/deputies fully expect to see an on-screen “suspect reaction” to verbal
commands and shots fired.
Cover from Return-Fire
In an effort to ascertain whether participants would utilize available cover within
each scenario’s context and whether participants were decisionally influenced by their
use or non-use of cover, each participant was provided a vertical cardboard “barricade”
that was 24” wide by 70” tall. The cardboard was partially reinforced by strapping tape
to inhibit penetration by the .68 rubber balls that were frequently fired at participants
when scenario-subjects fired at them. While most of the participating officers used the
cover provided, there was a wide degree of variance in how early (or late) in each
scenario they elected to use cover and to what degree they used cover effectively.
Many participating officers/deputies exposed far too much of themselves when using
cover.
If this study were to be replicated in the future, it is recommended that the use of
a “return-fire” capability be structurally modified so that cover could be used more
intuitively. Our use of the return-fire device had it positioned to the right of the projection
screen. The “cover” structure was positioned toward the left side of the screen. The
“threat” imagery projected onto the screen was centered on the screen and yet the
actual threat (projectiles fired from the return-fire device) was emanating from the
participant’s right side. This seemed to skew the manner in which cover was used and
may have also influenced shot placement to some degree.
“Death by Defiance™”
One of the least documented, perhaps least understood and yet most
problematic concerns that entered into this study was a phenomenon that injects an
incendiary set of cultural variables into citizen confrontations with police. For the better
part of two decades, if not more, pop culture in general and hip-hop culture in particular
has glamorized defiance and resistance against authority figures. While sociologists
have looked extensively at race as a predictor of the police use of deadly force, they
may have to consider whether culturally influenced nuance that police perceive as
“defiant” suspect behavior serves as a more reliable predictor than race alone.

2008 Copyright The Police Policy Studies Council

22
One of the most consistently influential factors elicited from participants in the
after-action debriefing process was that their decision to shoot was based on subject
non-compliance with verbal commands. The most common verbal commands were;
“police – don’t move,” and, “show me your hands.” It is from that frame of reference that
subsequent suspect actions were apparently perceived as being threatening.
In robbery scenarios, 49% of participants said non-compliance to verbal
commands was a determining factor in the use of deadly force, 42% stated that it was a
factor in burglary scenarios and 41% stated that it was so in mugging scenarios. And,
non-compliance was likely the perceptual filter through which a suspect’s concealed
hands, and turns toward the officer were mentally processed. (See APPENDIX for a full
tally of responses gleaned from the Officer Debrief Forms.)
Definition – “Death by Defiance”
A justifiable homicide that occurs after a flagrant level of furtive, suspicious or
otherwise reckless behavior is viewed concurrently with a subject exhibiting
non-compliance to stated and/or exhibited police authority. When such behavior
is confronted within substantive situational context, and the officer senses that
the subject’s (non-compliant) actions appear consistent with the initiation of a
lethal threat, an officer might reasonably use deadly force in a pre-emptive
manner. Thus, the subject’s defiant behavior becomes a compelling
determinant of the officer’s use of deadly force.
To expound upon this theorem, DBD might best be described as a symbiotic
behavioral phenomenon occurring (1) where and when non-compliant behavior (2) is
viewed as being hostile within what appears to be a felonious context (3) eliciting the
police use of deadly force when a threat of death or serious bodily harm is perceived as
being imminent to that officer. An officer’s perception of imminent danger can be
reasonably construed when the officer has issued concise audible commands (e.g.,
“Police – don’t move!” or “Police, show me your hands!”) yet is met with threatening
behavior construed from direct defiance and contradiction of the commands given. If,
given substantive situational context, the officer senses (non-compliant) furtive
movements that appear consistent with the initiation of a lethal threat, he/she might
reasonably use deadly force in a preemptive manner.
Mugging Scenarios and the Bystander Influence
Every participant in this study was subjected to one scenario in which a deadly
assailant was encountered with a victim/bystander directly behind the assailant. This
element was integral to the “mugging” scenario that each participant was exposed to.
Additionally, the nature of this scenario placed an apparent “victim” directly in the
participant’s line-of-fire when he/she elected to use deadly force. Most officers fired
when fired upon under these circumstances, but some did not. Many of the officers who
did fire at their assailant under these circumstances had one or more of their errant
rounds hit the “innocent bystander.” However, the frequency with which officers shot
bystanders while in the line-of-fire of perceived adversaries was not quantified in this
study.

2008 Copyright The Police Policy Studies Council

23
The Black Officer Phenomenon
When preliminary results of this study were released, one issue became an
almost immediate lightning-rod. Though only nine of the of the 307 officers that
participated in this study were self-identified as being black/African-American, seven of
those nine black officers shot unarmed suspects in this study. One of the nine black
participants shot both of the two unarmed suspects that he was exposed to.
Consequently, the regression and correlation tables (See: “Correlation for Instances
Where Subject was Unarmed”) reflect a significant relationship between black
participants and deadly force usage against unarmed suspects. In essence, the (.777)
frequency in which participating black officers used deadly force against unarmed
subjects in this study was twice that of the aggregate participant average. If we were to
factor in the one black officer who shot two unarmed subjects, the (unarmed suspect)
shooting frequency (.888) rises even more.
However, with so few black officer/deputy participants in this study, interpretation
of such limited data would be extremely tenuous, at best. Nevertheless, the numbers
are what they are and probably warrant greater emphasis in any future study structured
to replicate this project. Obviously, any such effort would require a much greater
participation of black officers and deputies. Presumably, this objective might best be
pursued by eliciting the involvement of urban law enforcement agencies that reflect a
higher composition of African-American officers.

2008 Copyright The Police Policy Studies Council

24

TRAINING IMPLICATIONS
As many would seem to sense, there is a downside to allowing officers too much
unrestrained discretionary latitude in the use of deadly force. Restraint isn’t just a
manifestation of agency policy parameters. It likely has more to do with the degree and
manner in which officers have been trained. Officers lacking in such ingrained restraint
might exhibit difficulty in controlling impulses that proper training often tends to mitigate.
Defining what proper training is has always been a monumental challenge. While we
can say that proper training involves task-oriented training that best reflects the nature
of an officer’s duties, further delineations become arguable. Officer proclivity to use
deadly force seems to be mostly derived from the perceived nature of each scenario
(burglary/robbery/mugging), and the actor’s furtive movements and non-compliance with
verbal commands (See Officer Debrief Summaries).
As was previously noted (“Inter-Agency Shooting Variations”), the distinct interagency differences relevant to shooting unarmed suspects seem directly attributable to
training. There were no substantive differences in agency policies pertinent to the use of
deadly force. The one agency that required its personnel to complete a “Use of Force
Report” whenever unholstering their handguns had a 44% rate of frequency in
engagement of unarmed suspects. Its participants did typically unholster their handguns
more slowly than participants from other agencies, but that didn’t seem to influence their
overall judgment in the research scenarios. The agency with the lowest frequency of
unarmed suspects shot (24%), judging from all informal participant debriefs, had the
most rigorous scenario-based training regimen. Virtually every participant interviewed
from that agency stated that he/she had had one or more force-on-force training
sessions in the last 12 months. In itself, this might not seem evidence adequate to
suggest that training was the most influential factor, but it is the only factor that clearly
stood out from all others. We were impressed by the overall professionalism exhibited
by participants afforded to us by all agencies. Scenario-based training was evident to
some degree in all participating agencies. However, in all but one agency, it seemed
much more intermittent rather than routine.
Another issue relevant to what was observed throughout the implementation of
this study was that of officer/deputy muzzle discipline. Many participants were seen
“covering-down” on suspects with their muzzles pointing directly at “center mass” of the
suspect. This may diminish reaction time by about one-tenth of a second, but it also
demands some serious trade-offs. A handgun presented to eye level occludes vision of
almost everything from the suspect’s sternum, down. A suspect’s hand and arm
movement are then difficult-to-impossible to discern. This would seem to suggest that
there might be serious threat identification issues corresponding with this approach.
Another possible trade-off lies within the fact that by truncating reaction time (by
elevating the muzzle before committing to fire), you also truncate the amount of time
needed to bring cessation to an erroneous “threat reflex” impulse. Many participants
were noted elevating their muzzles when they sensed a threat was imminent, only to

2008 Copyright The Police Policy Studies Council

25
“brake” their reaction string just as they had initiated their trigger-pull. Indeed, this
phenomenon was captured on many of the participant videos. In essence, truncated
reaction time is a double-edged sword.
Recent trends in active shooter training have provided somewhat of a trickledown effect of SWAT tactics to patrol officers. As an example of this trend, many patrol
officers are being taught a “ready-gun” position that has the handgun muzzle pointed at
center-mass of a suspect that they may not yet be willing to shoot. This runs contrary to
universally-embraced firearms safety protocol that mandates that one, “never point your
weapon at anything you’re not willing to destroy.” Consequently, the “muzzledominance” techniques commonly employed by SWAT personnel are becoming
manifest in patrol encounters, sometimes with critical unintended consequences. Even
where quasi-military (proactive) tactics have appropriate law enforcement applications,
weapon handling doctrine should be held to strict safety standards.
One final perspective should be made regarding the value of scenario-based
training – it can be part of the problem or part of the solution. Scenario-based training
should truly be geared toward “conflict resolution,” not merely gun-fighting skills. In
addition, a disproportionate number of “aggressive” scenarios may begin influencing
reactions akin to “fear-biting” in canines. A recent buzzword in the police training lexicon
has been that of “stress inoculation.” Approach this concept with caution. Your
officers/deputies may never be “warriors,” nor may you want them to be, but they must
be rational decision-makers. If your agency’s scenario-based training proportionately
reflects what duties and conflicts your officers/deputies are most likely to encounter on
the street, your training is likely where it needs to be.

2008 Copyright The Police Policy Studies Council

26

OIS INVESTIGATIVE IMPLICATIONS
In virtually all scenarios filmed and utilized, participants found themselves having
to decide whether to fully commit to using deadly force within a window in time that
generally averaged less than .35 seconds. In each scenario, participants seem to have
difficulty remembering everything they had been exposed to in such compressed
periods of intense decision-making. One might assume that what the officer is able to
process consciously (and then recall) might be a fraction of what he/she processes
subconsciously.
A contemporary and critical debate regarding this issue has been whether
officers should be permitted to view available dash-cam video of their own shooting
incident before being compelled to give a statement about that incident. One of the
peripheral issues noted in this study was the prevalence of so-called “threat-focus.”
Many of the shots fired by officers were shot at, or in very close proximity to, the
suspect’s gun-hand. This suggests that the officer’s visual focus is often directed toward
the weapon by which he/she feels most threatened. Subsequently, many other
peripheral (behavioral and situational) cues are likely filtered out of the officer’s
perceptual process. If one were to accept that premise, one might also accept the
premise that what is suppressed or filtered out of the “input” (perceptual) process will be
unavailable (or unreliable) for retrieval at a later time. For that reason, this study allowed
participants to review their own performance via the video-capture system that was
employed with each scenario that participants were subjected to. We would suggest
that officers/deputies be afforded the same opportunity if dash-cam video of a shooting
is available – prior to compelling that officer/deputy to give pertinent incident
statements.
All scenarios scripted and filmed for this study were “low light” scenarios. This
variable was engineered into the scenarios to increase realism and incident ambiguity.
After-action debriefing forms indicated that 20-23% of participants, in all three types of
scenarios, stated that, “The time of day and/or existing lighting conditions may have
influenced my decision to shoot.” To what extent officers/deputies can discriminate a
handgun from a cell phone, flashlight or wallet at night has become a source of concern
in recent years. Investigators should seriously consider taking detailed light
measurements wherever a low light officer-involved shooting has taken place. When in
doubt about this process, consult with people who specialize in low light visual issues
and who are experienced at taking low light measurements for forensic applications.
Keep in mind that our previous research indicates that 71-75% of “mistake-of-fact”
shootings occur under low light conditions and that an officer's or deputy’s lack of visual
acuity may have direct bearing on what he/she was able to visually discern during an
extreme encounter.

2008 Copyright The Police Policy Studies Council

27

POLICY CONSIDERATIONS
As many aforementioned findings suggest, many police shootings are “grayarea” events that often reflect split-second differentials between in-policy and out-ofpolicy shootings. The almost universal embrace of the “imminent threat” standard, in our
opinion, is warranted and court-defensible. For clarification purposes, “immediate threat”
scenarios are “must-shoot” incidents, and therefore generate the least amount of postincident controversy. When an officer responds to an immediate threat, he/she is either
reacting to someone lunging with an impact or edged weapon, looking down the barrel
of a suspect’s gun, or dodging bullets that have already been fired. We don’t require
that officers wait until a threat has manifested to that level before he/she can react with
deadly force. An insightful operational definition of what “imminent danger” is was
expressed by the New Jersey Division of Criminal Justice in 2000;
Imminent Danger:
•
Threatened actions or outcomes that may occur during an encounter
•
Threatened harm does not have to be instantaneous
This concept affords officers a large degree of latitude in their interpretation of
events – as does the Graham v. Connor “reasonable officer” standard. In fact, this
standard has become so universally accepted that even the ACLU now seems to
embrace the imminent threat standard for the police use of deadly force.
(http://www.aclu.org/police/gen/14614pub19971201.html)
However, many agencies have been seeking ways to diminish the frequency of
“may-shoot” incidents that arise from an imminent threat standard. The widespread
implementation of TASERs has been credited with substantially reducing the frequency
of “may-shoot” incidents. Policy modifications that attempt to narrow the parameters in
which officers may use deadly force have become a preferred “remedy.” However,
attempts to further define or restrict when an officer is authorized to use deadly force
often create a countervailing dynamic pertinent to what a “reasonable officer” might be
expected to do.
One agency encountered in this study had a requirement that its officers report
incidents in which handguns are unholstered operationally. In effect, an unholstered
handgun becomes a reportable “use of force.” Observationally, this did seem to
influence how early in a scenario participants from that agency drew their handguns –
even in robbery scenarios. A few participants literally waited to draw their handguns
until they came under fire. When debriefed about the delay that these participants
exhibited in drawing their handguns, a common response from younger, less
experienced individuals was that they were concerned about having their personnel files
reflecting frequent usage of force when in reality “force” was never used. Perhaps, in
retrospect, even the best intentions have demonstrable occupational safety implications.

2008 Copyright The Police Policy Studies Council

28
As was previously noted (See: Inter-Agency Shooting Variations), there were
some substantive differences between agencies on the lower and upper spectrum of
(unarmed suspect) shooting frequency. The differences noted seemed to be attributable
to training – not policy. As much as policy has been touted as a means of moderating
undesirable behavior, it is apparently investment in training that yields the best results.
A not-unexpected outcome of this study is the prevalence in which participants
found themselves firing at “suspects” after the suspect had already turned and fired at
them. A practical and altogether reasonable interpretation of what a “reasonable” officer
might do when, for instance, he/she confronts a non-compliant robbery suspect, would
be for that officer/deputy to shoot the suspect as he/she initiates a turning motion
toward that officer/deputy. This will likely be construed as being “controversial” in some
circles, but this study’s findings certainly seem to suggest that such officer/deputy
latitude is both reasonable and necessary.
Agencies should strive to provide frequent experiential interpretations of deadly
force policy. Often, policy directives are difficult to reconcile operationally. Scenariobased training is the most effective means of providing officers a practical frame of
reference with agency policy. Since training resources tend to be limited, effort should
be focused upon training that embeds critical decision-making activity within the most
common tasks associated with an officer’s job assignment. Vehicle stops, “suspicious
person” calls and domestic disturbance scenarios may not afford the sexiest fodder for
training, but for most beat officers, such calls represent a persistent occupational safety
problem.

2008 Copyright The Police Policy Studies Council

29

REFERENCES
Aveni, T. (2003), “Following Standard Operating Procedure,” Law & Order, Vol. 51, No. 8, pg. 79-87.
Correll, J., Park, B., Judd, C.M., & Wittenbrink, B. (2002). The police officer’s dilemma: Using ethnicity to
disambiguate potentially threatening individuals. Journal of Personality and Social Psychology, 83, 1314–
1329.
Fyfe, J.J. (1979), ``Administrative interventions on police shooting discretion: an empirical examination’’,
Journal of Criminal Justice, Vol. 7, pp. 309-24.
Fyfe, J.J. (1980), ``Geographic correlates of police shooting: a microanalysis’’, Journal of Research in
Crime and Delinquency, Vol. 17, pp. 101-13.
Fyfe, J.J. (Ed.) (1981a), ``Toward a typology of police shootings’’, in Contemporary Issues in Law
Enforcement, Sage Publications, Beverly Hills, CA, pp. 136-51.
Fyfe, J.J. (1981b), ``Who shoots? A look at officer race and police shooting’’, Journal of Police Science
and Administration, Vol. 9, pp. 367-82.
Fyfe, J.J. (1982), ``Blind justice: police shootings in Memphis’’, Journal of Criminal Law and Criminology,
Vol. 73, pp. 702-22.
Fyfe, J.J. (1986), ``The split-second syndrome and other determinants of police violence’’, in Campbell,
A.T. and Gibbs, J.J. (Eds), Violent Transactions, Basil Blackwell, Oxford.
Geller, W.A.(1982). Deadly force: What we know. Journal of Police Science and Administration,10,151–
177.
Geller, W. and Karales, K.(1981a),Split-Second Decisions: Shootings of and by Chicago Police, Chicago
Law Enforcement Study Group, Chicago, IL.
Greenwald, A.G., Oakes, M.A.,& Hoffman, H.G.(2003). Targets of discrimination: Effects of race on
responses to weapons holders. Journal of Experimental and Social Psychology, 39, 399–405.
Jacobs, D. and O’Brien, R.M. (1998), ``The determinants of deadly force: a structural analysis of police
violence’’, American Journal of Sociology, Vol. 103, pp. 837-62.
Kania, R.E. and Mackey, W.C. (1977), ``Police violence as a function of community characteristics,”
Criminology, Vol. 15, pp. 27-48.
Kobler, A.L. (1975a), ``Figures (and perhaps some facts) on police killing of civilians in the United States,
1965-1969’’, Journal of Social Issues, Vol. 31, pp. 185-91.
Lewinski, W. and Hudson, W. (2003), Police Marksman, “Reaction Times In Lethal Force Encounters,”
Vol. XXVIII, No. 5, 26-29
Meyer, M.W. (1980), ``Police shootings of minorities: the case of Los Angeles’’, Annals of the American
Academy of Political and Social Science, Vol. 452, pp. 98-110.
Milton, C., Halleck, J.S., Lardner, J. and Albrecht, G.L. (1977), Police Use of Deadly Force, Police
Foundation, Washington, DC.
Payne, B.K. (2001). Prejudice and perception: The role of automatic and controlled processes in
misperceiving a weapon. Journal of Personality and Social Psychology, 81, 181–192.
Plant, E.A., & Peruche, B.M. (2005). The consequences of race for police officers’ responses to criminal
suspects. Psychological Science, 16, 180–183.

2008 Copyright The Police Policy Studies Council

30
Plant, E.A., Peruche, B.M.,& Butz, D.A. (2005). Eliminating automatic racial bias: Making race nondiagnostic for responses to criminal suspects. Journal of Experimental and Social Psychology, 41, 141–
156.
Reisig, M.D., McCluskey, J.D., Mastrofski, S.D., & Terrill, W. (2004). Suspect disrespect toward the
police. Justice Quarterly, 21, 241–268.
Skolnick, J.H. and Fyfe, J.J. (1993), Above the Law: Police and the Excessive Use of Force, Free Press,
New York, NY.
Smith, B.W. (2004). Structural and organizational predictors of homicide by police. Policing: An
International Journal of Police Strategies and Management, 27, 539–557.
Sorenson, J.R., Marquart, J.W. and Brock, D.E. (1993), ``Factors related to killings of felons by police
officers: a test of the community violence and conflict hypotheses’’, Justice Quarterly, Vol. 10 No. 3, pp.
417-40.
Sunshine, J., & Tyler, T. R. (2003). The role of procedural justice and legitimacy in shaping public support
for policing. Law & Society Review, 37, 513–548.
Tennenbaum, A.N. (1994), ``The influence of the Garner decision on police use of deadly force’’, Journal
of Criminal Law and Criminology, Vol. 85, pp. 241-60.
Weitzer, R., & Tuch, S. A. (2004). Race and perceptions of police misconduct. Social Problems, 51, 305–
325.
White, M.D. (2002), “Identifying situational predictors of police shootings using multivariate analysis.”
Policing: An International Journal of Police Strategies & Management, Vol. 25 No. 4, pp. 726-751.
White, M.D. (2000), ``Assessing the impact of administrative policy on use of deadly force by on and offduty police’’, Evaluation Review, Vol. 24 No. 3, pp. 295-318.
White, M.D. (2001), ``Controlling police decisions to use deadly force: reexamining the importance of
administrative policy’’, Crime and Delinquency, Vol. 47 No. 1, pp. 131-51.

2008 Copyright The Police Policy Studies Council

31

APPENDICES

2008 Copyright The Police Policy Studies Council

32

Biographical Information

Thomas J. Aveni, MSFP
Thomas J. Aveni has over 20 years of law enforcement experience, having served on
the local and state levels in three states (NJ, UT, NH). His police career began in 1978,
and he has served as a police trainer since 1983. From 1990 to 2001 Tom served as a
police "Training Coordinator" with the once prestigious Smith & Wesson Academy.
There he was instrumental in training over 12,000 police and military personnel from
across the United States and 23 other countries. He has an undergraduate degree in
Criminal Justice from Westfield State College (MA) and a Masters Degree in Forensic
Psychology from American International College (MA). He is a co-founder of the Police
Policy Studies Council and a frequent lecturer at police training conferences.
Mr. Aveni conceptualized the need and material essence of this study, based on his
years of experience training law enforcement officers, and reviewing hundreds of
officer-involved shooting narratives. He also served as the primary architect, participant
proctor and data gatherer of this study.

2008 Copyright The Police Policy Studies Council

33

Reaction Time Histograms

Officer Shot Lag Histogram
50
45
35
30
25
20
15
10
5

2008 Copyright The Police Policy Studies Council

1.
9
M
or
e

1.
7

1.
5

1.
3

1.
1

0.
9

0.
7

0.
5

0.
3

0.
1

0
-0
.9
-0
.7
-0
.5
-0
.3
-0
.1

Frequency

40

34

Measure

Burglary

Robbery

Mugging

Mean

0.166323

0.001

0.035897

Median

0.166667

0

0.066667

2008 Copyright The Police Policy Studies Council

35

First Shot Fired Scenario Summary
Columns in RED reflect the Acting Quotient and the time in which the actor fired his first shot in that specific scenario. Scenario
numbers are found in the leftward column with the “IES #” header.

IES
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40

Acting
Quotient
Features

V

Suspect
Begins
Turn

T-LH-U-CH
E-LH-PC-CH
E-LH-U-CH
E-LH-U-HC
E-LH-PC-CH
E-LH-PC-OH
E-LH-PC-CH
E-LH-PC-CH
E-LH-U-OH
E-LH-PC-CH
E-LH-PC-CH
E-LH-U-OH
E-LH-PC-CH
E-HH-PC-CH
E-LH-U-OH
E-HH-U-CH
E-LH-U-CH
T-LH-U-OH
T-LH-U-CH
E-LH-FC-CH
T-HH-U-OH
T-LH-U-CH
E-HH-FC-CH
E-LH-PC-OH
E-LH-PC-CH
E-LH-FC-CH
E-LH-PC-CH
E-LH-FC-CH
E-LH-FC-CH
E-HH-U-OH
E-LH-U-CH
E-LH-PC-CH
E-LH-U-OH
E-HH-PC-HC
E-LH-PC-CH
E-LH-U-CH
E-LH-U-OH
E-LH-U-CH
E-LH-U-OH
E-LH-U-OH

2
3.5
3
2.5
3.5
2.5
3.5
3.5
2
3.5
3.5
2
3.5
2.5
2
2
2
1
2
4
0
2
3
2.5
3.5
4
3.5
4
4
1
3
3.5
2
2
3.5
3
2
3
2
2

18:27
17:16
19:22
18:15
21:03
21:04
19:25
18:21
19:27
20:25
20:27
19:22
18:26
20:04
20:11
20:25
20:18
21:02
23:35
22:05
21:16
21:20
21:12
20:00
20:08
21:11
20:25
20:19
20:20
22:18
24:07
24:04
24:12
23:22
25:12
25:07
25:17
26:01
27:10
25:15

Suspect’s
1ST
Shot Fired
20:15
22:00

19:14

22:03

22:07

21:21

23:07

23:01

22:02

21:12

24:17

26:10

26:18

IES
#
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80

Acting
Quotient
Features

V

Suspect
Begins
Turn

E-LH-U-CH
E-LH-U-OH
E-HH-U-CH
E-LH-FC-CH
E-HH-U-OH
T-HH-U-CH
E-LH-PC-CH
T-HH-U-OH
T-HH-U-CH
E-LH-FC-CH
T-HH-U-OH
E-LH-U-CH
E-LH-PC-CH
E-LH-PC-OH
E-LH-PC-CH
E-LH-PC-CH
T-LH-PC-OH
E-LH-U-HC
T-LH-U-CH
E-LH-U-OH
E-LH-U-CH
E-LH-PC-CH
E-LH-U-OH
E-LH-PC-CH
E-LH-PC-CH
E-LH-U-CH
E-HH-U-CH
E-LH-PC-CH
E-LH-U-OH
E-LH-PC-CH
E-LH-PC-CH
E-LH-U-OH
E-HH-U-CH
E-LH-FC-CH
T-HH-U-OH
E-HH-PC-CH
E-LH-PC-CH
E-LH-U-OH
E-HH-U-CH
T-HH-PC-CH

3
2
2
4
1
1
3.5
0
1
4
0
3
3.5
2.5
3.5
3.5
1.5
2.5
2
2
3
3.5
2
3.5
3.5
3
2
3.5
2
3.5
3.5
2
3
4
0
2.5
3.5
2
2
1.5

27:17
24:02
24:07
28:05
24:22
25:15
28:10
22:18
22:23
23:00
26:08
21:28
24:24
24:19
21:25
22:26
23:27
2308
22:22
24:00
22:03
22:07
23:25
22:22
22:14
20:18
20:26
22:25
20:10
22:24
20:15
26:18
26:04
27:02
23:03
25:01
25:15
24:28
23:28
23:22

Suspect’s
1ST
Shot Fired
28:05

28:25

29:10

23:26

25:14

23:27

24:12

23:01

23:06

23:29

21:04

28:07

26:10

25:17

Suspect’s turn toward Officer/Camera: T = Tepid E = Energetic
st
“Energetic” turns = at least 90% completed within 1 sec. or in which the suspect fired the 1 shot within 1 sec. of turn initiation.
st
“Tepid” = turns are those that take more than 1 second to complete at least 90% of the turn toward the camera or fire 1 shot at the camera.
LH = Turn w/ Low-Hands were turns initiated with “gun-hand” at or below the navel area
HH = Turn w/ High-Hands were turns initiated with “gun-hand” above height of navel area
U = Upright-turn
PC = Partial-Crouch: Shoulders slightly forward of the hips with minimal forward lean
FC = Crouch: Knees bent - shoulders noticeably forward of hips with feet planted at least shoulder-width apart
CH = Clenched Hands; suspect turns with clenched hand or hands. This CAN include scenarios where suspect turned with object in hand.
HC = Half-Closed Hand
OH = Open Hands; suspect turned with his/her hands open/unclenched

2008 Copyright The Police Policy Studies Council

36

Participant DeBrief Form Questions

DEBRIEF
QUESTION
1
2
3
4
5
6
7
8
9
10
11
12
12A
12B
12C
12D
12E
13
14
15
16
17
18
19
20
21
22

PARTICIPANT DEBRIEF FORM QUESTIONS
I used cover and that influenced my decision to
I did not use cover and that influenced my decision to
I used verbal commands and the suspect’s non-compliance influenced my decision to
When the subject’s hand or hands were concealed from view it influenced my decision to
I saw the suspect look over his/her shoulder at me and that influenced my decision to
I saw the suspect’s shoulder dip as his/her back was to me and that influenced my decision to
I saw the suspect’s shoulders start to rotate abruptly toward me and that influenced my decision to
As the suspect turned toward me I saw his/her right hand at waist height and that influenced my decision to
As the suspect turned toward me, I saw something in his/her right hand that influenced my decision to
As the suspect turned toward me, I saw nothing in his/her right hand that influenced my decision to
As the suspect turned toward me, I couldn’t tell if anything was in his/her right hand, influencing my decision to
The suspect’s outward physical appearance influenced my decision to
AGE
ATTIRE
ETHNICITY
GENDER
SIZE
I believe the suspect’s outward appearance may have caused me to hesitate in my decision to
I believe the suspect’s outward appearance may have caused me to hasten my decision to
The nature of the call, as I perceived it, influenced my decision to
The suspect’s proximity to what appeared to be a crime-in-progress may have influenced my decision to
The time of day and/or existing lighting conditions may have influenced my decision to
The time of day and/or existing lighting conditions may have caused me to hesitate in my decision to
If I had believed that I had back-up officers nearby that may have influenced my decision to
The suspect’s proximity to the crime-in-progress may have influenced my decision to
The suspect’s initial possession of a pry-bar may have influenced my decision to
The suspect’s decision to drop the pry-bar may have influenced my decision to

2008 Copyright The Police Policy Studies Council

CONSIDER SHOOTING
BURG
MUGG
ROBB
31.72% 22.40% 32.79%
8.44%
5.83%
5.84%
54.72% 48.70% 54.58%
73.11% 72.28% 72.79%
63.19% 60.46% 70.39%
55.41% 53.97% 60.66%
64.38% 67.43% 63.16%
57.98% 50.49% 52.61%
34.74% 26.07% 19.16%
2.29%
2.61%
1.96%
9.12%
6.54%
7.17%
14.24% 14.71% 15.58%
7.44%
6.82%
7.82%
14.89%
9.80%
13.73%
4.85%
3.26%
4.58%
10.36%
8.14%
10.78%
10.68%
8.50%
10.13%
9.39%
10.49%
5.88%
10.36% 14.10% 12.38%
68.75% 49.67% 68.42%
71.01% 59.74% 69.84%
52.29% 51.96% 40.52%
17.48% 20.13% 16.01%
11.73%
9.84%
9.18%
66.34% 33.11% 63.19%
67.54% 38.03% 54.07%
15.26%

37

DEBRIEF
QUESTION
1
2
3
4
5
6
7
8
9
10
11
12
12A
12B
12C
12D
12E
13
14
15
16
17
18
19
20
21
22

PARTICIPANT DEBRIEF FORM QUESTIONS
I used cover and that influenced my decision to
I did not use cover and that influenced my decision to
I used verbal commands and the suspect’s non-compliance influenced my decision to
When the subject’s hand or hands were concealed from view it influenced my decision to
I saw the suspect look over his/her shoulder at me and that influenced my decision to
I saw the suspect’s shoulder dip as his/her back was to me and that influenced my decision to
I saw the suspect’s shoulders start to rotate abruptly toward me and that influenced my decision to
As the suspect turned toward me I saw his/her right hand at waist height and that influenced my decision to
As the suspect turned toward me, I saw something in his/her right hand that influenced my decision to
As the suspect turned toward me, I saw nothing in his/her right hand that influenced my decision to
As the suspect turned toward me, I couldn’t tell if anything was in his/her right hand, influencing my decision to
The suspect’s outward physical appearance influenced my decision to
AGE
ATTIRE
ETHNICITY
GENDER
SIZE
I believe the suspect’s outward appearance may have caused me to hesitate in my decision to
I believe the suspect’s outward appearance may have caused me to hasten my decision to
The nature of the call, as I perceived it, influenced my decision to
The suspect’s proximity to what appeared to be a crime-in-progress may have influenced my decision to
The time of day and/or existing lighting conditions may have influenced my decision to
The time of day and/or existing lighting conditions may have caused me to hesitate in my decision to
If I had believed that I had back-up officers nearby that may have influenced my decision to
The suspect’s proximity to the crime-in-progress may have influenced my decision to
The suspect’s initial possession of a pry-bar may have influenced my decision to
The suspect’s decision to drop the pry-bar may have influenced my decision to

2008 Copyright The Police Policy Studies Council

BURG
5.83%
3.57%
17.92%
10.49%
9.45%
10.49%
17.65%
19.22%
30.84%
0.00%
2.93%
1.62%
0.32%
0.97%
0.00%
0.97%
0.65%
5.50%
3.88%
8.88%
14.66%
9.80%
5.18%
3.58%
14.71%
14.75%
4.22%

SHOOT
MUGG
10.71%
5.50%
19.48%
7.92%
6.86%
8.94%
18.09%
19.67%
34.98%
0.98%
2.94%
1.63%
0.65%
1.31%
0.65%
0.98%
0.65%
4.26%
2.62%
4.93%
17.16%
11.11%
8.44%
4.26%
11.15%
9.51%

ROBB
8.77%
4.87%
26.47%
15.08%
13.49%
14.75%
25.66%
25.49%
37.01%
0.65%
8.14%
3.57%
1.95%
2.94%
0.33%
1.96%
1.31%
6.86%
2.93%
25.99%
24.59%
13.73%
3.59%
2.30%
28.99%
24.76%

38

DEBRIEF
QUESTION
1
2
3
4
5
6
7
8
9
10
11
12
12A
12B
12C
12D
12E
13
14
15
16
17
18
19
20
21
22

PARTICIPANT DEBRIEF FORM QUESTIONS
I used cover and that influenced my decision to
I did not use cover and that influenced my decision to
I used verbal commands and the suspect’s non-compliance influenced my decision to
When the subject’s hand or hands were concealed from view it influenced my decision to
I saw the suspect look over his/her shoulder at me and that influenced my decision to
I saw the suspect’s shoulder dip as his/her back was to me and that influenced my decision to
I saw the suspect’s shoulders start to rotate abruptly toward me and that influenced my decision to
As the suspect turned toward me I saw his/her right hand at waist height and that influenced my decision to
As the suspect turned toward me, I saw something in his/her right hand that influenced my decision to
As the suspect turned toward me, I saw nothing in his/her right hand that influenced my decision to
As the suspect turned toward me, I couldn’t tell if anything was in his/her right hand, influencing my decision to
The suspect’s outward physical appearance influenced my decision to
AGE
ATTIRE
ETHNICITY
GENDER
SIZE
I believe the suspect’s outward appearance may have caused me to hesitate in my decision to
I believe the suspect’s outward appearance may have caused me to hasten my decision to
The nature of the call, as I perceived it, influenced my decision to
The suspect’s proximity to what appeared to be a crime-in-progress may have influenced my decision to
The time of day and/or existing lighting conditions may have influenced my decision to
The time of day and/or existing lighting conditions may have caused me to hesitate in my decision to
If I had believed that I had back-up officers nearby that may have influenced my decision to
The suspect’s proximity to the crime-in-progress may have influenced my decision to
The suspect’s initial possession of a pry-bar may have influenced my decision to
The suspect’s decision to drop the pry-bar may have influenced my decision to

2008 Copyright The Police Policy Studies Council

BURG
43.04%
83.12%
18.89%
1.97%
19.87%
30.49%
14.05%
14.98%
27.27%
71.24%
81.11%
55.34%
36.89%
30.10%
43.04%
36.89%
36.57%
79.29%
79.94%
16.78%
12.70%
35.29%
71.20%
74.92%
17.32%
7.21%
16.23%

NA
MUGG
47.73%
82.20%
19.81%
3.96%
24.18%
31.79%
11.18%
22.95%
29.37%
71.34%
85.95%
59.80%
36.36%
33.33%
42.02%
36.48%
36.60%
78.69%
78.69%
33.55%
18.15%
33.66%
64.29%
74.75%
40.66%
45.57%

ROBB
39.94%
85.06%
11.44%
0.98%
12.17%
21.31%
8.55%
15.69%
36.36%
67.65%
80.13%
52.60%
34.53%
30.39%
41.18%
33.66%
36.27%
80.07%
75.24%
3.95%
4.92%
43.46%
74.84%
84.26%
7.17%
20.20%

39

DEBRIEF
QUESTION
1
2
3
4
5
6
7
8
9
10
11
12
12A
12B
12C
12D
12E
13
14
15
16
17
18
19
20
21
22

PARTICIPANT DEBRIEF FORM QUESTIONS
I used cover and that influenced my decision to
I did not use cover and that influenced my decision to
I used verbal commands and the suspect’s non-compliance influenced my decision to
When the subject’s hand or hands were concealed from view it influenced my decision to
I saw the suspect look over his/her shoulder at me and that influenced my decision to
I saw the suspect’s shoulder dip as his/her back was to me and that influenced my decision to
I saw the suspect’s shoulders start to rotate abruptly toward me and that influenced my decision to
As the suspect turned toward me I saw his/her right hand at waist height and that influenced my decision to
As the suspect turned toward me, I saw something in his/her right hand that influenced my decision to
As the suspect turned toward me, I saw nothing in his/her right hand that influenced my decision to
As the suspect turned toward me, I couldn’t tell if anything was in his/her right hand, influencing my decision to
The suspect’s outward physical appearance influenced my decision to
AGE
ATTIRE
ETHNICITY
GENDER
SIZE
I believe the suspect’s outward appearance may have caused me to hesitate in my decision to
I believe the suspect’s outward appearance may have caused me to hasten my decision to
The nature of the call, as I perceived it, influenced my decision to
The suspect’s proximity to what appeared to be a crime-in-progress may have influenced my decision to
The time of day and/or existing lighting conditions may have influenced my decision to
The time of day and/or existing lighting conditions may have caused me to hesitate in my decision to
If I had believed that I had back-up officers nearby that may have influenced my decision to
The suspect’s proximity to the crime-in-progress may have influenced my decision to
The suspect’s initial possession of a pry-bar may have influenced my decision to
The suspect’s decision to drop the pry-bar may have influenced my decision to

2008 Copyright The Police Policy Studies Council

BURG
0.65%
0.97%
0.65%
8.52%
0.65%
0.33%
0.65%
0.33%
0.00%
1.96%
2.61%
21.36%
48.54%
47.57%
49.84%
48.22%
49.51%
0.32%
0.65%
0.99%
0.65%
0.65%
1.29%
1.30%
0.98%
2.95%
2.92%

UNKNOWN
MUGG
0.65%
2.27%
0.97%
9.24%
1.31%
1.66%
0.66%
0.66%
0.33%
0.65%
0.98%
18.95%
51.62%
51.96%
53.09%
52.44%
51.96%
0.66%
0.33%
0.99%
1.32%
0.65%
0.65%
1.64%
0.66%
0.33%

ROBB
1.62%
2.60%
1.31%
8.85%
0.66%
1.31%
0.99%
0.98%
1.95%
1.96%
1.30%
22.40%
51.14%
48.37%
52.61%
51.31%
50.65%
1.63%
2.28%
0.99%
0.66%
0.98%
1.63%
0.66%
0.33%
0.33%

40

DEBRIEF
QUESTION
1
2
3
4
5
6
7
8
9
10
11
12
12A
12B
12C
12D
12E
13
14
15
16
17
18
19
20
21
22

PARTICIPANT DEBRIEF FORM QUESTIONS
I used cover and that influenced my decision to
I did not use cover and that influenced my decision to
I used verbal commands and the suspect’s non-compliance influenced my decision to
When the subject’s hand or hands were concealed from view it influenced my decision to
I saw the suspect look over his/her shoulder at me and that influenced my decision to
I saw the suspect’s shoulder dip as his/her back was to me and that influenced my decision to
I saw the suspect’s shoulders start to rotate abruptly toward me and that influenced my decision to
As the suspect turned toward me I saw his/her right hand at waist height and that influenced my decision to
As the suspect turned toward me, I saw something in his/her right hand that influenced my decision to
As the suspect turned toward me, I saw nothing in his/her right hand that influenced my decision to
As the suspect turned toward me, I couldn’t tell if anything was in his/her right hand, influencing my decision to
The suspect’s outward physical appearance influenced my decision to
AGE
ATTIRE
ETHNICITY
GENDER
SIZE
I believe the suspect’s outward appearance may have caused me to hesitate in my decision to
I believe the suspect’s outward appearance may have caused me to hasten my decision to
The nature of the call, as I perceived it, influenced my decision to
The suspect’s proximity to what appeared to be a crime-in-progress may have influenced my decision to
The time of day and/or existing lighting conditions may have influenced my decision to
The time of day and/or existing lighting conditions may have caused me to hesitate in my decision to
If I had believed that I had back-up officers nearby that may have influenced my decision to
The suspect’s proximity to the crime-in-progress may have influenced my decision to
The suspect’s initial possession of a pry-bar may have influenced my decision to
The suspect’s decision to drop the pry-bar may have influenced my decision to

2008 Copyright The Police Policy Studies Council

NOT SHOOT
BURG
MUGG
ROBB
18.77% 18.51% 16.88%
3.90%
4.21%
1.62%
7.82%
11.04%
6.21%
5.90%
6.60%
2.30%
6.84%
7.19%
3.29%
3.28%
3.64%
1.97%
3.27%
2.63%
1.64%
7.49%
6.23%
5.23%
7.14%
9.24%
5.52%
24.51% 24.43% 27.78%
4.23%
3.59%
3.26%
7.44%
4.90%
5.84%
6.80%
4.55%
4.56%
6.47%
3.59%
4.58%
2.27%
0.98%
1.31%
3.56%
1.95%
2.29%
2.59%
2.29%
1.63%
5.50%
5.90%
5.56%
5.18%
4.26%
7.17%
4.61%
10.86%
0.66%
0.98%
3.63%
0.00%
1.96%
2.61%
1.31%
4.85%
6.49%
3.92%
8.47%
9.51%
3.61%
0.65%
14.43%
0.33%
7.54%
6.56%
0.65%
61.36%

41

Participant Consent Form

CONSENT TO PARTICIPATE IN RESEARCH
MMRMA DEADLY FORCE PROJECT
You are being asked to participate in a research study conducted by Thomas J.
Aveni, under the auspices of the Police Policy Studies Council. This study has been funded in part by the
Michigan Municipal Risk Management Authority, the Police Policy Studies Council, and all law
enforcement agencies participating in this study. If you have any questions or concerns about the
research, please feel free to contact your agency liaison for this study: _____________.
PURPOSE OF THE STUDY
This study has been designed to determine what an objectively reasonable officer might do given
exposure to certain incident variables. There are no “right” or “wrong” responses in this study and your
individual decisions will not be judged as a result.
PROCEDURES
If you volunteer to participate in this study, we would ask you to do the following things:
•

Complete the “Officer Profile” form that you are given and then return to the study proctor.

•

Please utilize the safety equipment that you are issued and do not remove it until told to do so.

•

Assist us in assuring that you have no live ammunition on your person when you arrive for
testing.

•

Judge each scenario as if you observed or were being dispatched to a similar incident while on
duty.

•

Use verbal commands where, when and as often as you deem them necessary.

•

Deploy your service pistol where and when you deem it to be necessary – re-holstering when you
believe it’s appropriate.

•

After each scenario culminates, fill out the “Debrief Sheet” that you’ll be issued for eliciting your
observations.

•

When you have completed your three scenarios and you’ve been released to resume your official
responsibilities, PLEASE REFRAIN FROM DISCUSSING THE NATURE OF THE STUDY WITH
YOUR PEERS UNTIL THIS STUDY HAS BEEN COMPLETED!

Your time with us, while participating in this study will likely encompass no more than twenty (20) minutes,
which includes completion of preliminary and observation report forms.
POTENTIAL RISKS AND DISCOMFORTS
The risks associated with this study are being kept as minimal as possible. No live ammunition will be
employed. However, in effort to inject some degree of situational stress into the decision-making process,
we’ll be employing a “Return Fire Cannon” that may fire a foam rubber projectile at you. You’ll be given
protective equipment that will minimize risk of injury.
POTENTIAL BENEFITS TO PARTICIPANTS AND/OR TO SOCIETY
This project will likely provide direct benefits to you, your agency and those that you serve, as it has been
designed to provide valuable insights about the critical decision-making process as it pertains to deadly
force. From a scientific standpoint, this study will be collecting and evaluating valuable data never before
gathered. We believe this will be a ground-breaking study.

2008 Copyright The Police Policy Studies Council

42

PAYMENT FOR PARTICIPATION
This study is being undertaken on a small MMRMA grant and upon the cumulative good will of all
participating agencies. There will be no compensated participation in this study and involvement is
voluntary.
CONFIDENTIALITY
Every effort will be made to ensure confidentiality of any identifying information that is obtained in
connection with this study.
Upon your initial contact with the study’s proctor, you’ll be assigned a roster number. The roster number
that you’ll be assigned is for control and auditing purposes only. Upon completion of this study, any and
all publication of the results will be devoid of any personal references of those that participated. The
participant roster will be kept for a period of 12-24 months for the purpose of facilitating any review or
audit of the methodology or analysis involved. Any and all data shared with the participating agencies will
be communicated in aggregate and/or categorized formats. The names of participants will not be divulged
within the context of individual performance.
All research activities will be videotaped for review and auditing purposes by the PPSC research staff.
Video footage will be kept strictly confidential. Each participant has a right to review the taped footage.
Taped footage will be kept for 12-24 months for the purpose of facilitating any review or audit of the
methodology or analysis involved.
RIGHTS OF RESEARCH PARTICIPANTS
You may withdraw your consent at any time and discontinue participation without penalty. You are not
waiving any legal claims, rights or remedies because of your participation in this research study. You
may exercise the option of removing your data from the study. You may also refuse to answer any
questions you don’t want to answer and still remain in the study. The investigator may withdraw you from
this research if circumstances arise that warrant doing so. If you have questions regarding your rights as
a research participant, contact:

Thomas J. Aveni, MSFP
The Police Policy Studies Council
P.O. Box 475
Spofford, NH 03462
877-267-7772 Ext. 2
tom@theppsc.org
SIGNATURE OF RESEARCH PARTICIPANT/LEGAL REPRESENTATIVE
I have read the information provided for the study “MMRMA Deadly Force Project” as described
herein. My questions have been answered to my satisfaction, and I agree to participate in this
study. I have been given a copy of this form.

_____________________________
Name of Participant (please print)

___________________________
Signature of Participant

____________________________
Name of Witness (please print)

__________________________
Signature of Witness

DATE _________________________

2008 Copyright The Police Policy Studies Council