mapping police stress

25
http://pqx.sagepub.com/ Police Quarterly http://pqx.sagepub.com/content/14/3/227 The online version of this article can be found at: DOI: 10.1177/1098611111413991 2011 14: 227 originally published online 18 July 2011 Police Quarterly Matthew J. Hickman, Jennifer Fricas, Kevin J. Strom and Mark W. Pope Mapping Police Stress Published by: http://www.sagepublications.com On behalf of: Police Executive Research Forum Police Section of the Academy of Criminal Justice Sciences can be found at: Police Quarterly Additional services and information for http://pqx.sagepub.com/cgi/alerts Email Alerts: http://pqx.sagepub.com/subscriptions Subscriptions: http://www.sagepub.com/journalsReprints.nav Reprints: http://www.sagepub.com/journalsPermissions.nav Permissions: http://pqx.sagepub.com/content/14/3/227.refs.html Citations: at SEATTLE UNIV LIBRARY on September 19, 2011 pqx.sagepub.com Downloaded from

Upload: seattleu

Post on 21-Apr-2023

2 views

Category:

Documents


0 download

TRANSCRIPT

http://pqx.sagepub.com/Police Quarterly

http://pqx.sagepub.com/content/14/3/227The online version of this article can be found at:

 DOI: 10.1177/1098611111413991

2011 14: 227 originally published online 18 July 2011Police QuarterlyMatthew J. Hickman, Jennifer Fricas, Kevin J. Strom and Mark W. Pope

Mapping Police Stress  

Published by:

http://www.sagepublications.com

On behalf of: 

Police Executive Research Forum Police Section of the Academy of Criminal Justice Sciences

can be found at:Police QuarterlyAdditional services and information for     

  http://pqx.sagepub.com/cgi/alertsEmail Alerts:

 

http://pqx.sagepub.com/subscriptionsSubscriptions:  

http://www.sagepub.com/journalsReprints.navReprints:  

http://www.sagepub.com/journalsPermissions.navPermissions:  

http://pqx.sagepub.com/content/14/3/227.refs.htmlCitations:  

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

Police Quarterly14(3) 227 –250

© The Author(s) 2011Reprints and permission: http://www. sagepub.com/journalsPermissions.nav

DOI: 10.1177/1098611111413991http://pqx.sagepub.com

PQX413991 PQX14310.1177/1098611111413991Hickman et al.Police Quarterly

1Seattle University, Seattle, WA, USA2RTI International, Research Triangle Park, NC

Corresponding Author:Matthew J. Hickman, Department of Criminal Justice, Seattle University, 901 12th Ave, Seattle, WA 98122 Email: [email protected]

Mapping Police Stress

Matthew J. Hickman1, Jennifer Fricas1, Kevin J. Strom2, and Mark W. Pope2

Abstract

Research on police stress has developed out of several theoretical frameworks, but the knowledge base is limited by a common reliance on self-report stress measures. This article describes an innovative approach to studying police stress that attempts to overcome some of these limitations by using direct, real-time, and spatially anchored measurement of an officer’s stress response (via heart rate) during shift work. A pilot study was conducted using a single officer to determine whether this methodology is feasible for future studies. The pilot study demonstrated that continuous heart rate measurement over the course of the test officer’s shift was possible and that these data could be placed in space-time context for purposes of exploring potential stress “hot spots.” Overall, the results indicate that the methodology is both feasible and suitable for systematic studies of police stress, with the potential to advance our understanding of when, where, and why officers experience stress. Potential benefits, limitations, challenges of implementation, and future directions are discussed.

Keywords

police stress, heart rate, measurement, mapping

IntroductionPolice occupational stress is a significant concern for police administrators, officers, and the public. In addition to officer health concerns (both physical and mental well-being), occupational stress has broader impacts on job performance, family life, and the quality of service provided to the public. The sources of police occupational stress are varied and can range from the seemingly benign and more cumulative forms of stress (such

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

228 Police Quarterly 14(3)

as administrative stressors) to stress related to responding to “critical incidents” in which officers’ and/or citizens’ lives are at risk.

The most extreme forms of officer stress are associated with the use of lethal vio-lence by and against the police. For example, the Pacific Northwest region recently experienced an unusual string of deadly attacks on police officers. Several officers were shot at over a 2-month period, resulting in six deaths.1 One of these incidents received a large amount of national media coverage: the killing of four officers as they met in a local coffee shop at the beginning of their shift. The stress that officers experience when they are exposed (either directly or indirectly) to the killing of other officers reinforces a broader and powerful truth about police occupational stress: officers face an ever-present and unpredictable potential for violence, injury, and death (Crank, 1998; Crank & Caldero, 1991; Territo & Vetter, 1981; Violanti & Aron, 1995).

Yet while these critical incidents are a source of acute stress, they constitute only a small proportion of the total activity during an officer’s patrol shift, which has been described anecdotally as “. . . eight hours of boredom interrupted by five minutes of terror” (an officer quoted in Peak, 2009, p. 133). Realistically, a larger proportion of patrol shift time consists of relatively less stressful events, with academic studies of patrol time allocation tending to reinforce this notion (Brooks & Piquero, 1998; Piquero, 2005; Swatt, Gibson, & Piquero, 2007). Nonetheless, these more “routine” aspects of patrol work, which involve responding to calls for service with little information about what to expect, are a potential source of chronic stress that can accumulate and have a significant impact on an officer over time (Crank, 1998; Regehr, LeBlanc, Jelley, & Barath, 2008). The consequences of this work environment are significant, including physical health problems, substance abuse, psychological problems, burnout, and rela-tionship problems. Well documented in the literature is the fact that police populations have higher rates of disease as compared with the general public (e.g., Franke, Collins, & Hinz, 1998; Violanti, Vena, & Marshall, 1986).

One widely accepted belief is that the higher disease rate among police is a result of the stress inherent to policing, and specifically to the routine generation of chemical stress products (such as cortisol) within the body resulting from repeated initiation of the body’s “fight-or-flight” response without subsequent action (a process popularized in the mainstream literature by primatologist Robert Sapolsky’s [1994] book, Why Zebras Don’t Get Ulcers). In brief, the fight-or-flight response produces chemical stress products thought to damage the body through a process of immunosuppression. The physiological effects of the body’s stress-response mechanisms do not change whether the stress that causes them is real or perceived, emotional or physical, or acted on or ignored. Researchers focused on the physiological measurement of stress in polic-ing have approached their work from this viewpoint (e.g., Anderson, Litzenberger, & Plecas, 2002; Violanti et al., 2007).

Although much work has been conducted to better understand police stress (Anderson et al., 2002; Bergen & Bartol, 1983; Brooks & Piquero, 1998; Buker & Wiecko, 2007; Crank, 1998; Crank & Caldero, 1991; Davidson & Veno, 1978; Evans, Coman, & Stanley, 1992; Franke et al., 1998; Gershon, Barocas, Canton, Li, & Vlahov, 2009;

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

Hickman et al. 229

Kroes, 1976; Kroes & Hurrell, 1977; Morash, Haarr, & Kwak, 2006; Newman & Rucker-Reed, 2004; Pendleton, Stotland, Spiers, & Kirsch, 1989; Piquero, 2005; Regehr et al., 2008; Roberg, Hayhurst, & Allen, 1988; Storch & Panzarella, 1996; Swatt et al., 2007; Territo & Vetter, 1981; Violanti, 2007; Violanti & Aron, 1995; Violanti et al., 1986; Wallace, Roberg, & Allen, 1985), the bulk of this research has tended to rely on perceptual methodologies that do not directly assess through physio-logical measurements the stress experienced by officers in the midst of a patrol shift. Survey questionnaires and observational research, although generating important and insightful information, have limitations that can only be overcome by direct physio-logical measurement. Given the unique nature of police work, directly measuring offi-cer physiological response during a shift would provide a more accurate assessment of the stress they encounter.

In addition, there is a need to address the issue of police stress more rigorously within a spatial context. Although ample research documents how different parts of a city generate sizeable proportions of police “activity” (in terms of calls for service, crime incidents, arrests, etc.; e.g., Sherman, Gartin, & Buerger, 1989; Sherman & Weisburd, 1995), police stress research has tended to examine how average self-reported stress levels vary across police organizational units. This aggregation masks much finer varia-tion in the spatial context of police stress and does not capture the actual physiological response to discrete incidents within that context. Understanding the spatial distribu-tion of police stress could potentially inform decisions regarding deployment and dispatch strategy as well as occupational health programs targeted at managing and reducing stress.

In sum, we believe it is important for researchers and police executives to understand the actual stress that officers experience on a daily basis, in the course of their routine patrol activities. This requires direct physiological measurement during regular shift work and the ability to anchor those measurements in time and space. The purpose of the cur-rent study is to describe a methodology for quantifying the physiological response to calls for service by means of bio-feedback devices that record an individual’s heart rate in space-time. The pilot study reported herein used GPS-enabled wrist-watch recorders manufactured by Garmin to collect heart-rate data during a test officer’s shift. These devices use a wireless chest-strap heart-rate monitor and simultaneously record heart rate as well as latitude and longitude in real time. We describe how these data can be linked with calls-for-service data to model the average physiological response to differ-ent calls. We also describe future directions for more systematic study, including using calls-for-service data as a proxy for ride-alongs, using subsamples of officers for obser-vational ride-alongs, and collecting questionnaire data and other baseline measurements prior to collecting field-based heart-rate data.

The next section reviews relevant literature on stress, law enforcement occupa-tional stress, and the physiological measurement of stress. We then present the pilot study and results and turn to a discussion of issues associated with a more system-atic approach. We conclude with a discussion of potential implications for policy and practice.

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

230 Police Quarterly 14(3)

Literature ReviewDefining StressAlmost all discussions of stress physiology begin with the work of Selye (1956), who developed the idea of general adaptation syndrome (GAS). GAS is induced by any “nonspecific stressor” that alters homeostasis (or equilibrium) in the body. There are three stages of GAS, including an acute alarm stage (i.e., the fight-or-flight response, in which the body prepares for action by providing the necessary energy resources), a more chronic resistance stage (in which the body copes with the alarm stage and seeks a return to homeostasis), and a chronic exhaustion stage (in which energy resources are exhausted and the body is unable to return to homeostasis). Selye (1974) is also credited with the idea of positive forms of stress (eustress) or stress that is actually pleasurable (such as might be encountered in elective physical exercise).

A more psychosocial model of stress rejects the nonspecific stressor in favor of an evaluative calculus in which an actor perceives a task and weighs it against their ability to manage that task (e.g., Lazarus, 1981; Sarason & Sarason, 1979). Anderson et al. (2002) draw on the psychosocial model in their study of police officer stress, in viewing stress as a combination of stressor and stress reactivity. We take a similar approach here by explicitly defining police stress as an officer’s physiological response to the per-ceived imbalance between situational demands and his or her capacity to meet or over-come those demands, induced by task-related stressors common to the law enforcement occupation.

Police Occupational StressIt is difficult to identify the origins of research on police occupational stress and its effects on officers. It has long been recognized that policing is a stressful occupation, but research attention appears to have grown substantially since the 1970s in concert with government-sponsored symposia and studies reporting that police officers had higher rates of stress-related illnesses. There was also an increase in studies describing the nature of police stressors and exploring stress-reduction techniques (Davidson & Veno, 1978; Kroes, 1976; Kroes & Hurrell, 1977) as well as research focused on officer burnout (Maslach & Jackson, 1979; Roberg et al., 1988; Wallace et al., 1985).

This literature began to identify several stressors both considered unique to policing (e.g., death of a fellow officer, killing a civilian in the line of duty, pursuit driving, use of force) and not unique to policing (e.g., rigid bureaucracy, high workload, limited promo-tional opportunities, shift work; e.g., Territo & Vetter, 1981). Subsequent research on police stress has tended to distinguish between organizational or administrative stressors as compared to job/task-related stressors (Brooks & Piquero, 1998). For example, Violanti and Aron (1995) surveyed approximately 100 officers in a depart-ment in New York and asked the respondents to rank-order various organizational and police work–related stressors. Among the top-ten (mean) ranked items were killing some-one in the line of duty, fellow officer killed, physical attack, battered child, high-speed

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

Hickman et al. 231

chases, shift work, use of force, inadequate department support, incompatible partner, and accident in patrol car. They also reported variation in stressor level by officer length of service (with more seasoned officers evidencing higher mean scores).

Some research, however, indicates that aspects of policing commonly thought to be high stressors (e.g., killing someone in the line of duty, officer killed) may be overstated as a stressor. Crank and Caldero (1991) surveyed officers from eight departments in Illinois and found that the majority of respondents identified aspects of the organization (including their supervisors) as the main source of stress to offi-cers while citizen contact and the potential for danger were ranked lower. Similar findings (emphasizing organizational stressors over job/task-related stressors) have been reported in other survey-based research studies (Buker & Wiecko, 2007; Morash et al., 2006).

Measuring Officer StressThe measurement of police stress has typically revolved around self-report scales consisting of a series of statements with Likert-type response options or scored check-lists. These scales generally seek to quantify exposure to police stressors (e.g., Beehr, Johnson, & Nieva’s [1995] 25-item Police Stress Scale) and/or perceptions of work-related stress. Researchers then correlate scores on these measurements with officer demographics or with other measures designed to tap expected stress outcomes such as depression, substance abuse, and burnout, and intermediaries such as coping skills (e.g., Gershon et al., 2009). Spielberger, Gorsuch, Lushene, Vagg, and Jacobs’s (1983) State-Trait Anxiety Inventory has been used fairly extensively in the police stress literature (Bergen & Bartol, 1983; Evans et al., 1992; Newman & Rucker-Reed, 2004; Pendleton et al., 1989; Storch & Panzarella, 1996). The State-Trait Inventory seeks to capture both state anxiety (i.e., anxiety in the present state or present conditions) and trait anxiety (i.e., static anxiety embedded in one’s personality). Trait anxiety is an important control measure in understanding state anxiety, with variability in state anxiety somewhat contingent on trait anxiety.

In general, the literature on police occupational stress reflects a traditional progression of scientific discovery. Earlier work tended to focus on defining and conceptualizing police stress and the categorization of specific stressors (e.g., organizational/administrative stressors vs. stressors associated with police work), whereas more recent work has tended to focus on the measurement of police stress and the development and testing of theory (see Swatt et al., 2007). A limitation is that much of this body of research relies on survey-based methodology, with very little direct measurement of stress in real-world settings. That is, most police stress research has been designed within a perceptual framework. This is an important limitation, especially when viewed in light of consistent survey research findings which show that organizational/administrative stressors are stronger predictors of stress and related outcomes than job/task-related stressors such as perceived danger (Buker & Wiecko, 2007; Crank & Caldero, 1991; Gershon et al., 2009; Morash et al., 2006; Storch & Panzarella, 1996).

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

232 Police Quarterly 14(3)

Physiological Measurement of Stress

The direct physiological measurement of stress is a challenging task. Stress, when broadly defined as a tension that alters homeostasis in the human body, manifests itself in a wide variety of physiological indicators. Acute stress situations (in which the human fight-or-flight response is activated in the hypothalamus) trigger the sympathetic ner-vous system (SNS) and the hypothalamic-pituitary-adrenal (HPA) axis. The stimulants epinephrine and norepinephrine are released into the bloodstream, resulting in elevated heart rate, elevated blood pressure, and increased blood flow to the muscles. The hor-mones cortisol and thyroxin are also released, which help to support the production of glucose. In essence, this process is the body’s way of preparing for action in response to the stressful stimuli.

Cortisol has been a subject of research attention as an indicator of stress response and due to its potential harmful long-term effects insofar as it suppresses the immune system. One research approach has been to simulate stressful conditions and use mouth swabs to measure salivary cortisol levels. For example, Regehr et al. (2008) reported on a simulation study in which 84 police recruits went through a Firearms Training Systems (FATS) scenario designed to induce acute stress. They measured heart rate and cortisol levels at different time points, in addition to collecting questionnaire-based measures of state anxiety and demographic information. Likewise, Violanti et al. (2007), as part of their larger study of baseline biomarkers for stress in police officers, studied self-reported posttraumatic stress disorder (PTSD) symptoms in relation to salivary corti-sol samples taken over a 3-day period from 100 police officers in Buffalo, New York.

While it would be ideal to be able to measure several different indicators (e.g., heart rates, blood pressure, perspiration, cortisol levels, and other indicators) during actual police work, as opposed to simulated or clinical conditions, it is technically unfeasible to monitor invasively (e.g., taking mouth swabs for cortisol levels) under these actual work circumstances. However, it is feasible to monitor heart rates during actual police work with minimal impact on the officer. In such a study, heart rate would be used as a proxy for nervous system activation—part of the body’s stress response.

Heart-rate measurements as a proxy for physiological reactions to stressful events have long been used in both laboratory and field experiments on the effects of stress (Johnston, 2002; Johnston, Anastasiades, & Wood, 1990; Johnston, Tuomisto, & Patching, 2008; Kamarck, Debski, & Manuck, 2000; Kamarck, Schwartz, Janicki, Shiffman, & Raynor, 2003). Heart rate is a reliable indicator of immediate physiological adjustments to stress as mediated through the sympathetic nervous system (part of the autonomic nervous system). In addition, heart-rate variability (HRV) is also used, in combination with baseline (usually resting, nonstress) heart-rate measurements, as a means of expressing changes in heart rate in response to various simulated or real-life scenarios expected to cause sympathetic arousal.

In contrast, direct measures of HPA axis arousal in response to stress—such as sali-vary cortisol measurement—are known to lag in time behind the appearance of and

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

Hickman et al. 233

reaction to the stressful scenario (Looser et al., 2010). However, several studies (Iellamo et al., 2003; Looser et al., 2010; Samel, Vejvoda, & Maass, 2004) have demonstrated that the HPA axis and the autonomic nervous system work relatively independent of one another during the stressors that occur as part of normal, everyday life. During significant stressful events—such as those which may be anticipated or occur during a police shift—the physiological responses of the HPA axis and the autonomic nervous system synchronize (Looser et al., 2010). This allows for significant anticipation of the physiological markers which may change and be available for measurement under “routine” vs. “significant” circumstances. These facts, in combination with the rela-tively more intensive methods needed for real-life measurement of increases in cortisol (such as saliva swabs) makes heart rate both a convenient and reliable measure of phys-iological field stress reactions when interpreted while controlling for other factors known to influence heart rate.

Controlling for the various individual and situational factors which can influence heart rate and HRV is complex. LeBlanc, Regehr, Jelley, and Barath (2008) have shown that age influences heart rate in general but not HRV in response to stress situations. In addition, Looser et al. (2010) succinctly state, “HR is regulated both by sympathetic and parasympathetic tone . . . and increases in response to physical demands as well as psychological stressors” (p. 281). Given that police officers are often required to under-take physical demands as well as psychological stress during any given shift, and that these demands can at best only be partially anticipated, it is important to account for these factors which can cause variation. Demographic factors, such as age, relation-ship status, exercise activities and fitness, and body mass index, have been shown to correlate with heart rate and overall cardiac health (Hamer et al., 2006; LeBlanc et al., 2008; Looser et al., 2010). Second, several personality and coping styles have also been shown to influence HRV in response to stressful situations (Johnston et al., 2008; LeBlanc et al., 2008; Zellars, Meurs, Perrewè, Kacmar, & Rossi, 2009). Finally, although less clearly delineated in the literature, the amount of coffee intake may influence the size of HRV in response to stress situations (Hamer et al., 2006).

The first—and apparently only—study to directly measure officer heart rates in the course of their regular duties was conducted during 1998-1999 in 12 municipal police departments of British Columbia, Canada (Anderson et al., 2002). As part of a larger study of the physical requirements of policing in Canada, Anderson et al. conducted ride-along observations with a subsample of 121 officers. During these ride-along observations, officers were fitted with heart monitors and heart rates were recorded during the entire 12-hr shifts. The time interval was minutes, with each minute of observational data linked to each minute of heart-rate monitoring. Heart-rate data were downloaded at the end of each shift and subsequently merged with the observa-tional data for analysis. Unfortunately, technical difficulties with the monitoring devices required replacement with an improved device midway through the study, resulting in a total of 76 usable cases.

Heart rates were reported as both “above-resting” (i.e., heart rate expressed as the difference between a given measurement and the officer’s lowest heart rate for the shift)

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

234 Police Quarterly 14(3)

and as “percentage of heart-rate reserve” (i.e., heart rate expressed as a proportion of the heart-rate range, where maximum heart rate was defined as 220 minus the officer’s age). Above-resting heart rates standardize for differences in resting heart rate across officers, whereas the percentage of heart-rate reserve standardizes for the effects of age and allows exploration of “working” heart rate vs. maximum potential (Anderson et al., 2002).

The average resting heart rate was 59 beats per minute, increasing to an average of 82 during shift work (the average above-resting rate was thus 82 – 59 = 23). Above-resting rates were highest at the beginning of the shift (25 during the first hour) and lowest at the end of the shift (19 during the last hour) to which Anderson et al. attribute anticipatory stress at the start of shifts. As might be expected, average above-resting rates were lowest when sitting (M = 20) and highest when engaged in pushing or pull-ing activity (M = 49). Use of force actions, such as wrestling with suspects (M = 65) and handcuffing (M = 45), generated the highest above-resting rates. Also of note were elevated rates during incidents with variations of officer “hand on gun” in the presence of suspects (means ranged from 40 to 49).

Calls for service generated higher above-resting rates that corresponded with the call’s priority code; assigned “Code 3” calls (i.e., high priority) had an average above-resting rate of 28 (compared with 19 for normal driving), and officers providing back-up on these calls had an even higher above-resting rate (M = 41). Another key finding was that although heart rates dropped following “critical incidents,” they did not return to preincident levels, with officers maintaining elevated rates for 30 to 60 min afterward, varying with postincident activities (e.g., when talking to suspects). Anderson et al. took this as evidence of officers “. . . maintaining a state of hyper-vigilance” (Anderson et al., 2002, p. 414).

In sum, Anderson et al. reported the first physical evidence of physiological responses to actual police work, in the form of elevated heart rates. The authors offer their results as evidence of fight-or-flight response initiation and suggest that when this response is not acted on, the lingering stress products can negatively affect the body. This is offered as a potential link between stress and disease. Two policy suggestions are provided, including programs to help officers manage their perceptions of control and debriefing sessions following critical incidents.

As the only study to monitor officer heart rates in the course of their regular duties, the Anderson et al. (2002) study provides important insight into the physiological response to police work activities. However, the study has some limitations that leave room for improvement. For example, Anderson et al. did not report any statistical sig-nificance testing, so it is unknown whether the differences they report are meaningful or possibly due to chance variation. Technical problems with the monitors limited the available sample and raise questions about bias and generalizability. The heart-rate data were linked to ride-along observations, but no attempt was made to link to calls-for-service databases, which would permit an additional real-time connection between the call and the officers response. Finally, Anderson et al. did not examine the spatial

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

Hickman et al. 235

context of officer stress, which as noted earlier may be important given the nonrandom patterning of crime and calls-for-service within cities.

Present StudyThere are four primary goals of the methodological approach we describe herein. The first goal is to improve the measurement of police stress by focusing on the real-time measurement of physiological response (via heart rates) as well as linkage to the dis-patch system. This approach will allow an assessment of stimulus (call dispatch) and response (elevated heart rate). The judicious use of observational ride-alongs would enable assessment of the extent to which the dispatch database can effectively serve as a proxy for observational ride-alongs, which are often deemed too obtrusive to employ on a large scale.

The second goal is to provide basic descriptive information about the physiological stress associated with police patrol. For example, what is the average physiological response to calls in the study area? Can we chart a typical “day in the life” of an officer using heart rates? What proportion of a shift is spent at resting heart rate? What proportion of a shift is spent in activity with elevated heart rates?

The third goal is to enable formal testing of research hypotheses, several of which are consistent with extant criminological theories such as general strain theory (e.g., Swatt et al., 2007). The following are examples of possible hypotheses: (a) higher priority calls for service will be followed by higher average stress response; (b) officers on the late shift (3rd watch) will exhibit higher average stress levels than officers on day shifts; (c) officers with greater trait anxiety will have greater stress response; (d) officers in one-officer cars will have higher average stress response than those in two-officer cars; and (e) offi-cers with greater years of service will have lower stress response.

A final goal is to place the measurement of police stress in a space-time context. As the recording units simultaneously record heart rate as well as latitude and longi-tude, the point data can be mapped using common mapping software and aggregated to areal units as desired for tests of spatial clustering. With careful sampling, it would be possible to generate valid “hot-spot” style maps, although such maps would depict aver-age officer heart rates (in density terms) instead of crimes.

For some time now, athletes have been using GPS-enabled heart-rate monitors and other sensing equipment to enhance their training regimens (Fleming et al., 2010). The benefits of this approach include accurate tracking of pace, elevation change, and other factors that contribute to a more comprehensive understanding of performance. Advances in technology have brought these tools to mass market, such that affordable and accu-rate devices for the real-time measurement of heart rate, coupled with spatial location, are now widely available.

To further the study of police stress and in particular to expand our ability to develop data relevant to when, where, and why officers experience stress, we tested a product manufactured by Garmin using a volunteer officer from the Seattle Police Department

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

236 Police Quarterly 14(3)

(SPD). We sought to measure stress response, via heart rate, in real-time during the course of actual shift work and to be able to map these data. The next section describes our methods and the results of that test, along with a demonstration of the potential utility of mapping stress over a study area and examining stress in a space-time context.

MethodTo determine the feasibility of this methodology, a volunteer test participant (an officer in the East Precinct of the SPD) was recruited to wear a Garmin GPS-enabled wrist-watch monitor equipped with a wireless heart-rate monitor chest strap during his regular shift.2 The test participant was a 33-year-old male patrol officer with approximately 10 years of service in the SPD. The test participant was accompanied by the lead author as a ride-along observer. This test was conducted in early 2010, during a weekend evening shift (1930-0430).

The data recorded by the Garmin watch were subsequently downloaded for analysis. The watch records four key variables at each point of measurement: (a) date and time (in a combined format), (b) latitude, (c) longitude, and (d) heart rate (beats per minute [bpm]). The raw data were imported into an Excel spreadsheet, and various formulae were used to extract the specific date and time variables to express the results in con-tinuous local time.

We proceed with a descriptive and conceptual analysis, first reviewing the series of events occurring during the shift, presenting the heart-rate data and then mapping the data in a series of conceptual maps. We also provide a more detailed review of a particu-larly serious event which corresponded with the peak heart rate achieved during the shift. We then turn to a discussion of limitations and improvements for future study.

ResultsThe shift proceeded with a series of calls, including a traffic accident, 911 hang-up, traffic stops pursuant to vehicle record checks, responding to a hit/run call (which evolved into a DUI arrest at gunpoint), a call to assist a “repo man” confronted by an angry car owner wielding a baseball bat, and responding to a call reporting scream-ing and a possible knife within a Seattle Housing Authority apartment building. The shift also included a number of officer-initiated activities beyond traffic stops, such as talking to individuals on the street to obtain information or to build on existing relationships.

Heart RateThe trace of the test participant’s heart rate during the course of the shift appears in Figure 1. One feature of the Garmin watch allows the programmer to (a) record data using a “smart” technology that makes a record every time there is a significant change in heart rate or direction or (b) to record every second. If using the latter option, the

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

Hickman et al. 237

user is guaranteed the most accurate record but sacrifices available recording time (up to 3.5 hr). For the pilot study, we opted for the former “smart” mode to be able to obtain recordings over the course of the entire shift. As can be seen in Figure 1, how-ever, this results in unequal time intervals, and an interpolation strategy may be necessary if this is determined to be the optimal method for recording heart rate. Without interpo-lation, this could also lead to potentially inaccurate statistics (such as the average heart rate during a shift). Interpolation of latitude and longitude would also be necessary to accurately represent these data in spatial context.

The heart-rate monitor recorded an average heart rate of 96 bpm and a maximum heart rate of 165 bpm.3 The maximum heart rate was reached at the point of the officer responding to a noncompliant DUI suspect. The initial call was for a “hit-and-run” by a pick-up truck at approximately 21:36. In this incident, the driver had apparently hit eight vehicles and run over two stop signs. The test participant located the vehicle and pursued. The driver did not respond to lights and siren. After some distance, the driver finally stopped his vehicle by means of hitting a curb. At approximately 21:42, the test participant issued repeated commands over the PA system to the driver to turn off the ignition and throw his keys on to the pavement. The driver was nonresponsive. The test participant exited the cruiser and drew and pointed his service weapon at the suspect vehicle’s driver compartment. The test participant advanced on the vehicle, repeatedly issuing commands to the driver to put his hands out of the window. The test partici-pant’s heart rate peaked at 165 bpm at 21:43. When the test participant arrived at the driver’s window, he issued a command to unlock and open the door. The driver was again nonresponsive. Finally, the test participant opened the door and assisted the driver

Heart Rate During Shift

020406080

100120140160180

19:1

6:58

20:0

3:27

20:2

7:22

20:4

5:40

20:5

8:56

21:2

0:20

21:3

4:57

21:4

5:32

23:0

4:30

0:00

:17

1:13

:05

1:35

:59

1:51

:57

2:06

:30

2:18

:35

2:31

:26

2:43

:26

2:57

:25

3:10

:26

3:33

:57

TIME

BPM

Figure 1. Heart-rate recording of test participant

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

238 Police Quarterly 14(3)

out of the vehicle, at which point the driver—obviously intoxicated—simply collapsed onto the ground. The test participant then rolled the driver onto his stomach and hand-cuffed him.

With regard to the comfort of the equipment, toward the end of the shift (after about 7 hr), the test participant reported that the chest-strap was starting to feel slightly uncomfortable. The watch was reported to be bulky but not to the point of interfering with standard operations. However, the test participant reported that if he was to be involved in a foot chase, jumping over fences, fighting with suspects, and so on, that the watch could possibly get stripped off in the course of these more strenuous activities.

Mapping of Heart-Rate DataThe geo-coordinates and associated data can easily be imported into GIS software packages, such as ArcGIS. Doing so clearly demonstrates the value of placing the heart-rate data in spatial context. As can be seen in Figure 2, the incidents described above can now be understood in space-time. The ellipse drawn within the map identi-fies the DUI gunpoint arrest incident. Moving from left to right within that ellipse, as the officer realizes that he is dealing with a noncompliant and potentially dangerous, intoxicated suspect, his heart rate steadily increases, maximizing at the point of drawing his sidearm and advancing on the vehicle.

Other locations and incidents also evidence higher heart rates. For example, the incident involving a repo-man confronted by an angry car owner wielding a baseball bat occurred due east of the previously described DUI arrest incident later in the shift. The test participant’s heart rate peaked at 132 bpm during this incident. Lower “resting” rates were observed during routine patrol activities along major thoroughfares and side streets.

Going beyond the mapped point data, different aggregation techniques can be used to examine the statistical validity of clustering through calculation of LISA statistics. As an example, Figure 3 (offered for conceptual purposes only) demonstrates how a cho-ropleth map could be generated using a conventional study grid consisting of approxi-mately 200 sq. ft cells. The point data were aggregated to the study grid cells and average values computed.

If LISA statistics confirm significant clustering, it may then be possible to generate meaningful density surfaces—in essence generating “hot spot”–style maps of police stress. As an example, Figure 4 (offered for conceptual purposes only) presents a kernel density estimation surface generated from the point data.

Linking Heart-Rate Data With Calls-for-ServiceLinking the officer’s heart-rate data to the calls-for-service database can be accomplished using data elements common to each source. The SPD calls-for-service database, like many other computer-aided dispatch (CAD) systems, includes several data fields that are generated with each dispatch. These include (a) an event number assigned to the

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

Hickman et al. 239

Figure 2. Map of point data collected via GPS-enabled heart-rate monitor

call, (b) the date and time the call came in the CAD system and the date and time the call was finished, (c) the “x” and “y” coordinates from the call taking system (null if non-911 generated), (d) the census tract and block, (e) an incident report code indicating the

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

240 Police Quarterly 14(3)

nature of the call according to the officer clearing the call, (f) the disposition of the call, (g) the nature of the call according to the 911 operator taking the call, (h) a numeric indicator of how important the call is (0, E, and 1 are emergency calls; 2 is important;

Figure 3. Conceptual map demonstrating aggregation of heart-rate data to study grid for identifying potential stress “hot spots”

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

Hickman et al. 241

3 and 4 are routine; 7 is a traffic stop; 8 is an officer-initiated call), and (i) the car number of the primary unit dispatched to the call. Of particular importance is the officer/car identifier to delineate which unit responded to the call. This field, in addition to the date,

Figure 4. Conceptual map demonstrating kernel density estimation surface to help visualize potential stress “hot spots”

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

242 Police Quarterly 14(3)

time, and geo-coordinates, allows linkage of the CFS data to the officer heart-rate data via administrative data that can be collected prior to an officer’s shift (see Figure 5).

By systematically collecting data from several officers within a study area, linking these data will permit the estimation of average heart-rate responses to calls (e.g., the average heart rate following high-priority calls versus low-priority calls, on different shifts, in different sectors, etc.) and enable basic hypothesis testing. Given the continu-ous stream of data, it may also be possible to use conventional time-series methods, such as interrupted ARIMA models.

Limitations and Improvements for Systematic StudyAlthough the data generated by the GPS-enabled heart-rate monitor (and linked with the CFS database) is a potentially powerful research tool, it will be both necessary and desirable to collect additional data to realize its fullest potential. First, it will be impor-tant to collect baseline measurements on officers, including heart rate and blood pres-sure, prior to shift work. This would enable the standardized “above resting” measures used by Anderson et al. (2002) as well as provide additional control variables for any preshift conditions. Postshift heart-rate and blood pressure measurements would also be desirable to collect because they provide data on prolonged elevations that could be correlated with stress incidents from the preceding shift. Such prolonged elevations could have negative health effects.

It will also be important to conduct short surveys with the officers to collect addi-tional explanatory and control variables, including responses to standard perceptual police stress and/or state-trait anxiety scales. We should be clear that although we are advocating the use of physiological measures, we are not suggesting that self-report perceptual measures should be discarded. Rather, a combined approach would yield a more comprehensive assessment and allow researchers to tease out the relative strengths and weaknesses of each method. A questionnaire should include, at a minimum, demo-graphic variables, and basic self-reported data on exercise frequency, coffee consump-tion, medications known to affect heart rate, and selected personality or coping-style inventories. Baseline physiological measurements should include heart rate, blood pressure, height, and weight. A brief measurement of resting heart rate and blood

Call For Service DataDateTimeX coordinateY coordinateBeatCall typeOfficer/car identifier

Shift Specific DataOfficer/car identifierDateShift timeStress assessment questionsBaseline heart rate

Officer Stress Data (Heart Rate)DateTimeX coordinateY coordinateHeart rateOfficer identifier

Figure 5. Links between calls for service, officer, and officer heart-rate data

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

Hickman et al. 243

pressure prior to the start and at the conclusion of each shift would be beneficial for comparative purposes (Johnston et al., 2008). In this way, resting and recovery data can be collected with each shift and incorporated into analyses along with baseline and experimental (shift) data. Finally, Johnston et al. (2008), also showed that self-reported feelings of “tense arousal” (defined as feelings of rushing, anxiety or tension, and irrita-tion or anger) were positively associated with heart rate increases in the field. However, the potential benefit of these data must be weighed against the importance of a minimally invasive research design in keeping with the nature of police field work.

Using multiple data sources underscores the importance of a unique identifier with which to link across the various data sources (see Figure 5). The identifier must be capable of linking in-field heart-rate data to calls-for-service data and to pre- and post-shift physiological measurements and survey questionnaires as well as ride-along observations. As such, having a well-thought-out data-management process is critical to being able to produce valid and meaningful results.

While ride-alongs would provide the richest data source on stressful stimuli during the course of an officer’s shift, they are likely too obtrusive and costly as a primary means of collecting these data. Using calls-for-service data as a proxy is a unique way to obtain stressful stimuli data, particularly when the call’s comments are incorporated to provide additional context regarding the call. While ride-along observations are not feasible for every shift/officer included in a systematic study, they could realistically be conducted with a subsample of officers. This would serve two important purposes: First, to ensure that calls-for-service data are serving as a valid proxy for observations of calls to which officers respond and, second, to collect time-marked qualitative data about the incidents to which officers respond and the officer’s reactions to relate that information to the heart-rate data.

The overall burden on departmental resources could also be quite significant. Data-collection efforts would likely need to be split by a combination of sector and shift to logistically manage meeting with officers, administering the baseline measurements, and providing the heart-rate monitor to them. For example, data could be collected from officers assigned to a particular sector for 1 week and then on officers in remain-ing sectors on following weeks. This staggered data-collection approach could be completed at different time points to account for seasonality, so that data are generated for each officer for a week of shifts for a few distinct weeks.

The GPS data can be imported into suitable software (such as ArcGIS) so that the heart-rate data can be mapped. As demonstrated by the pilot study, this is a fairly straightforward process. However, given the volume of data that would be generated by a larger, systematic study, it will be necessary to develop an optimal database struc-ture to efficiently process these data.

Perhaps the most important aspect of a more systematic approach is having the commitment and support of both the police department and the police union. This is especially true with regard to the use of GPS monitoring and the collection of health-related data. Police unions in some jurisdictions are quite vocal in their opposition to GPS monitoring, and in other jurisdictions the GPS issue has already been resolved.

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

244 Police Quarterly 14(3)

There are also potential liability issues associated with the many “what if” questions (e.g., what if an officer becomes involved in an incident that results in legal action—will the research data be seized and/or otherwise accessible?). Ultimately, the research-er’s responsibility is to the research participants and not to the police department, police union, or to the courts. This is an important discussion to have with all research stake-holders prior to engaging in any data collection, and institutional review boards (IRBs) may have some difficulty with the associated liability to the researcher and/or institution. Having thorough and documented stakeholder discussions (perhaps in the form of a memorandum of understanding) prior to submission and as part of the research protocol will facilitate IRB decision making. An IRB will also have questions associated with the possible risks to officers that may accrue from the monitoring equipment as well as the security of the health-related data and plans for managing the confidentiality and ultimate anonymity of the data. A clear plan outlining the use of scrambled identifiers for purposes of data linkage, along with a final randomly generated case identification number (implemented as soon as practical) would be essential in this regard.

Potential Implications for Criminal Justice Policy and PracticePatrol officers play a foundational role as decision makers in the criminal justice process. If the health and safety of these officers is not understood, monitored, and responded to, the quality of their decisions will likely suffer. In addition, data on the long-term health consequences that police officers experience related to their occu-pational stress exposure is an important policy issue, given the higher rates of health problems that police experience and the service officers provide to society. We believe the methodological approach we advocate here has several potential implications for criminal justice policy and practice, including (a) improved understanding of officer exposure to stressful situations, (b) beat and/or district modifications, (c) deployment strategies, (d) improved officer job satisfaction, (e) police training, and (f) future research directions.

One of the key implications of this pilot study is its potential for serving as a model for replication in police stress research. Past research in the area of police stress has focused primarily on officer self-reports of stress, as opposed to a physiological indi-cator of stress. Collecting a real-time physiological indicator of stress (i.e., heart rate) will allow for a more accurate analysis of officer stress over a variety of time and space dimensions. By pairing the officer stress data with calls for service data, we can better understand the types of incidents that contribute to higher levels of officer stress.

Flowing directly out of the above implication is the potential to use data for beat or district modification. If analyses show that specific beats exhibit a propensity for officers to regularly encounter highly stressful situations, it could lead to a redrawing of beat boundaries to ensure better load balancing across officers in a department. Redistricting could be based in part on stress-level patterns, rather than call volume alone (as is typi-cally done). In Seattle, the proportion of officers assigned to each precinct is balanced with the proportion of workload, where workload is measured as service hours spent

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

Hickman et al. 245

on calls for service and officer-initiated activity. For example, the East precinct has 18% of the workload and 17% of all officers are assigned there; the North precinct carries 32% of the workload and 31% of all officers are assigned there. However, while the workload-to-officer ratios are balanced at the precinct level, it is not the case that specific call types are balanced.

Similar to beat/district modification, a department could also change how officers are deployed to minimize repeated exposure to highly stressful situations. For example, if a particular shift or time period is found to have higher rates of stress among officers or particular hot spots of stress, a department may want to ensure that officers are not repeatedly assigned to that shift or time period to reduce prolonged exposure to stress. While these practices would of course depend on staff resources available within the department, a rotation method could be enacted to move officers through other sectors, even within shifts.

A related area potentially informed by this approach is stress inoculation training (SIT). Much like the inoculations delivered in the medical field to prevent the onset of more serious forms of disease, the idea here is that by exposing an individual to mild forms of stress you can improve their ability to cope with more severe forms of stress (e.g., Meichenbaum & Novaco, 1978; Novaco, 1977). Police training is designed to some extent with this process in mind; simulated stressful conditions are a common part of academy training that prepares recruits for real-world stress-ful conditions (often referred to as “survival stress” training), particularly with regard to use of force (Hickman, 2005; Reaves, 2009). SIT models view stress as the product of the interaction between individual and environment; as such, knowledge of the micro-geography of police stress may assist with the design of survival stress training for individual police departments (as well as individual SIT treatments delivered by clinicians).

By nature, police work is stressful but the ability to identify factors or situations that make it more stressful will enable departments to enact proactive strategies to counter these factors or situations and improve officer health. Future research could employ this methodology in more departments across the country, further examin-ing how specific officer characteristics are related to officer stress (and how officer characteristics might mediate the effect of workload) and investigating how offi-cer stress changes over time and across departments. There may also be meaningful research questions surrounding the ecological correlates of police stress. Are higher stress levels associated with particular community characteristics? Are certain call characteristics associated with higher levels of stress? Are crime hot spots correlated with stress hot spots?

We believe the results of this initial pilot study support the need for more systematic efforts, and we encourage researchers to take up our call for exploring more direct and comprehensive approaches to understanding police stress. Understanding and respond-ing to the occupational health consequences of policing could ultimately lead to improved officer job satisfaction and better long-term health outcomes for police officers.

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

246 Police Quarterly 14(3)

Acknowledgments

The authors thank Chief John Diaz and Captain James Dermody of the Seattle Police Department and Sergeant Richard O’Neill of the Seattle Police Officers’ Guild for their support of this research; Stephen Rice and Alex Piquero for their comments on earlier drafts; and an anonymous Seattle police officer who served as the test participant and made this research possible.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Notes

1. These include Officer Timothy Brenton of the Seattle Police Department (October 31, 2009), Officers Greg Richards, Mark Renninger, Tina Griswold, and Ronald Owens of the Lakewood Police Department (November 29, 2009), and Deputy Kent Mundell of the Pierce County Sheriff’s Office (December 21, 2009).

2. The SPD was only willing to let us use a single test officer. This was due in part to the fact that our research followed shortly after the killing of an SPD officer as well as several other officers in the region. The SPD subsequently issued a policy restricting all citizen ride-alongs, and we had to obtain special permission from the Chief’s office for a single test. In addition, SPD was concerned about any potential impact of the monitoring equipment on the perfor-mance of duties. In the event that the equipment interfered with officer duties, it was thought prudent to limit that to a single officer rather than several officers. While this was obviously limiting, we believe that the single test officer was an important and sufficient first step for purposes of demonstrating the feasibility and suitability of the method for more systematic studies of police stress.

3. Note that without interpolation of data points during periods of less activity, the average here (96 bpm) is quite a bit higher than the average 82 bpm reported in Anderson et al. (2002). Over the course of the shift, the average measurement interval was 10 s. With interpolation, our average would probably be lower. Of course, the pilot test involved only one officer who experienced some fairly stressful events during the test shift, and the finding of a higher average than reported in Anderson et al. (2002) could easily be due to normal variation.

References

Anderson, G., Litzenberger, R., & Plecas, D. (2002). Physical evidence of police officer stress. Policing, 25, 399-420.

Beehr, T., Johnson, L., & Nieva, R. (1995). Occupational stress: Coping of police and their spouses. Journal of Organizational Behavior, 16, 3-25.

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

Hickman et al. 247

Bergen, G., & Bartol, C. (1983). Stress in rural law enforcement. Perceptual and Motor Skills, 56, 957-958.

Brooks, L., & Piquero, N. (1998). Police stress: Does department size matter? Policing, 21, 600-617.

Buker, H., & Wiecko, F. (2007). Are causes of police stress global? Testing the effects of common police stressors on the Turkish National Police. Policing, 30, 291-309.

Crank, J. (1998). Understanding police culture. Cincinnati, OH: Anderson.Crank, J., & Caldero, M. (1991). The production of occupational stress in medium-sized police

agencies: A survey of line officers in eight municipal police departments. Journal of Criminal Justice, 19, 339-349.

Davidson, M., & Veno, A. (1978). Police stress: A multicultural, interdisciplinary review and perspective. Abstracts on Police Science, 6, 187-199.

Evans, B., Coman, G., & Stanley, R. (1992). The police personality: Type A behavior and trait anxiety. Journal of Criminal Justice, 20, 429-441.

Fleming, P., Young, C., Dixon, S., & Carre, M. (2010). Athlete and coach perceptions of technology needs for evaluating running performance. Sports Engineering, 13, 1-18.

Franke, W., Collins, S., & Hinz, P. (1998). Cardiovascular disease morbidity in an Iowa law enforcement cohort, compared with the general Iowa population. Journal of Occupational and Environmental Medicine, 40, 441-444.

Gershon, R., Barocas, B., Canton, A., Li, X., & Vlahov, D. (2009). Mental, physical, and behav-ioral outcomes associated with perceived work stress in police officers. Criminal Justice and Behavior, 36, 275-289.

Hamer, M., Williams, E., Vuononvirta, R., Gibson, E., & Steptoe, A. (2006). Association between coffee consumption and cardiovascular function during mental stress. Journal of Hypertension, 24, 2191-2197.

Hickman, M. (2005). State and local law enforcement training academies, 2002. Washington, DC: Bureau of Justice Statistics.

Iellamo, F., Pigozzi, F., Parisi, A., Di Salvo, V., Vago, T., Norbiato, G., . . . Pagani, M. (2003). The stress of competition dissociates neural and cortisol homeostasis in elite athletes. Journal of Sports Medicine and Physical Fitness, 43, 539-545.

Johnston, D. (2002). Acute and chronic psychological processes in cardiovascular disease. In K. Schaie, H. Leventhal, & S. Willis (Eds.), Effective health behavior in older adults (pp. 55-64). New York, NY: Springer.

Johnston, D., Anastasiades, P., & Wood, C. (1990). The relationship between cardiovascular responses in the laboratory and in the field. Psychophysiology, 27, 34-44.

Johnston, D., Tuomisto, M., & Patching, G. (2008). The relationship between cardiac reactivity in the laboratory and in real life. Health Psychology, 27, 34-42.

Kamarck, T., Debski, T., & Manuck, S. (2000). Enhancing the laboratory-to-life generalizability of cardiovascular reactivity using multiple occasions of measurement. Psychophysiology, 37, 533-542.

Kamarck, T., Schwartz, J., Janicki, D., Shiffman, S., & Raynor, D. (2003). Correspondence between laboratory and ambulatory measures of cardiovascular reactivity: A multilevel modeling approach. Psychophysiology, 40, 675-683.

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

248 Police Quarterly 14(3)

Kroes, W. (1976). Society’s victim—The policeman: An analysis of job stress in policing. Springfield, IL: Charles C. Thomas.

Kroes, W., & Hurrell, J. (1977). Job stress and the police officer: Identifying stress reduction techniques. Washington, DC: U.S. Government Printing Office.

Lazarus, R. (1981). The stress and coping paradigm. In C. Eisendorfer, D. Cohen, A. Kleinman, & P. Maxim (Eds.), Models for clinical psychopathology. New York, NY: Spectrum.

LeBlanc, V., Regehr, C., Jelley, R., & Barath, I. (2008). The relationship between coping styles, performance, and responses to stressful scenarios in police recruits. International Journal of Stress Management, 15, 76-93.

Looser, R., Metzenthin, P., Helfricht, S., Kudielka, B., Loerbroks, A., Thayer, J., & Fischer J. (2010). Cortisol is significantly correlated with cardiovascular responses during high levels of stress in critical care personnel. Psychosomatic Medicine, 72, 281-289.

Maslach, C., & Jackson, S. (1979, January/February). Burned out cops and their families. Psychology Today, 20-21.

Meichenbaum, D., & Novaco, R. (1978). Stress inoculation: A preventative approach. In C. Spielberger & I. Sarason (Eds.), Stress and anxiety (Vol. 5). Washington, DC: Hemisphere.

Morash, M., Haarr, R., & Kwak, D. (2006). Multilevel influences on police stress. Journal of Contemporary Criminal Justice, 32, 631-641.

Newman, D., & Rucker-Reed, M. (2004). Police stress, state-trait anxiety, and stressors among U.S. Marshals. Journal of Criminal Justice, 32, 631-641.

Novaco, R. (1977). A stress inoculation approach to anger management in the training of law enforcement officers. American Journal of Community Psychology, 5, 327-346.

Peak, K. (2009). Policing America: Challenges and best practices (6th ed.). Upper Saddle River, NJ: Prentice Hall.

Pendleton, M., Stotland, E., Spiers, P., & Kirsch, E. (1989). Stress and strain among police, firefighters, and government workers. Criminal Justice and Behavior, 16, 196-210.

Piquero, N. (2005). Understanding police stress and coping resources across gender: A look towards general strain theory. In H. Copes (Ed.), Policing and stress (pp. 126-139). Upper Saddle River, NJ: Prentice Hall.

Reaves, B. (2009). State and local law enforcement training academies, 2006. Washington, DC: Bureau of Justice Statistics.

Regehr, C., LeBlanc, V., Jelley, R. B., & Barath, I. (2008). Acute stress and performance in police recruits. Stress and Health, 24, 295-303.

Roberg, R., Hayhurst, D., & Allen, H. (1988). Job burnout in law enforcement dispatchers: A comparative analysis. Journal of Criminal Justice, 16, 385-393.

Samel, A., Vejvoda, M., & Maass, H. (2004). Sleep deficit and stress hormones in helicopter pilots on 7-day duty for emergency medical services. Aviation, Space, and Environmental Medicine, 75, 935-940.

Sarason, I., & Sarason, B. (1979). The importance of cognition and moderator variables in stress (Technical Report SCS-LS-009). Arlington, VA: Office of Naval Research.

Sapolsky, R. (1994). Why zebras don’t get ulcers. New York, NY: Henry Holt & Co.Selye, H. (1974). Stress without distress. Philadelphia, PA: Lippincott.

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

Hickman et al. 249

Selye, H. (1956). The stress of life. New York, NY: McGraw-Hill.Sherman, L., Gartin, P., & Buerger, M. (1989). Hot spots of predatory crime: Routine activities

and the criminology of place. Criminology, 27, 27-55.Sherman, L., & Weisburd, D. (1995). General deterrent effects of police patrol in crime hot spots:

A randomized, controlled trial. Justice Quarterly, 12, 635-648.Spielberger, C., Gorsuch, R., Lushene, R., Vagg, P., & Jacobs, G. (1983). State-trait anxiety

inventory for adults. Redwood City, CA: Mind Garden.Storch, J., & Panzarella, R. (1996). Police stress: State-trait anxiety in relation to occupational

and personal stressors. Journal of Criminal Justice, 24, 99-107.Swatt, M., Gibson, C., & Piquero, N. (2007). Exploring the utility of general strain theory in

explaining problematic alcohol consumption by police officers. Journal of Criminal Justice, 35, 596-611.

Territo, L., & Vetter, H. (1981). Stress and police personnel. Journal of Police Science and Administration, 9, 195-208.

Violanti, J., Andrew, M., Burchfiel, C., Hartley, T., Charles, L., & Miller, D. (2007). Post-traumatic stress symptoms and cortisol patterns among police officers. Policing, 30, 189-202.

Violanti, J., & Aron, F. (1995). Police stressors: Variations in perception among police personnel. Journal of Criminal Justice, 23, 287-294.

Violanti, J., Vena, J., & Marshall, J. (1986). Disease risk and mortality among police officers: New evidence and contributing factors. Journal of Police Science and Administration, 14, 17-23.

Wallace, P., Roberg, R., & Allen, H. (1985). Job burnout among narcotics investigators: An exploratory study. Journal of Criminal Justice, 13, 549-559.

Zellars, K., Meurs, J., Perrewè, P., Kacmar, C., & Rossi, A. (2009). Reacting to and recovering from a stressful situation: The negative affectivity–physiological arousal relationship. Journal of Occupational Health Psychology, 14, 11-22.

Bios

Matthew J. Hickman is an assistant professor in the Department of Criminal Justice at Seattle University, Seattle, Washington. His general research interests are in policing, forensic science, and quantitative methods. His recent work has been published in a variety of journals including Criminology & Public Policy, Journal of Quantitative Criminology, and Crime & Delinquency. He received his doctoral degree in criminal justice from Temple University in 2005.

Jennifer Fricas is a faculty member in the College of Nursing at Seattle University. She is a registered nurse specializing in community and public health, and teaches courses in commu-nity health theory and clinical, basic physical assessment as well as global health. She has worked with a variety of unique communities over the course of her career, as a professional nurse, an educator, and a collaborator committed to the intersection of academic research and community-defined relevance. She received her bachelor of science in nursing from James Madison University and master of public health in Global Health Policy from The George Washington University.

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from

250 Police Quarterly 14(3)

Kevin J. Strom is a senior research scientist at RTI International. His research interests include the impact of forensic science on the criminal justice system and law enforcement responses to community violence. His recent work has been published in journals that include Criminology & Public Policy, the Journal of Quantitative Criminology, and Crime & Delinquency. He received his doctoral degree in criminology from the University of Maryland, College Park in 2001.

Mark W. Pope is a research analyst in RTI International’s Crime, Violence, and Justice Program. He has more than 10 years of experience conducting research in the areas of policing, homeland security, prisoner reentry, and large-scale evaluations. His research interests include using informa-tion technology to develop data-driven solutions to crime, text mining/text analysis, and homeland security. He received his master of science in information science from University of North Carolina–Chapel Hill in 2007.

at SEATTLE UNIV LIBRARY on September 19, 2011pqx.sagepub.comDownloaded from