THE ROLE OF GENDER AND RISK PERCEPTION AMONG POLICE OFFICERS:
GENERAL LIFESTYLE RISKS AND OCCUPATION-SPECIFIC RISKS
By
Yi Ting Chua

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Criminal Justice

2012

ABSTRACT
THE ROLE OF GENDER AND RISK PERCEPTION AMONG POLICE OFFICERS:
GENERAL LIFESTYLE RISKS AND OCCUPATION-SPECIFIC RISKS
By
Yi Ting Chua
The role of gender has been shown to be important in numerous academic fields. This is the case
for the study of risk perception and the study of policing. Within the risk perception literature, it has been
consistently found that women tend to perceive higher levels of risks compared to men. Similar results
remain even when factors such as education and knowledge are controlled for; such as examining specific
occupation groups such as scientists, where a gendered driven risk perception gap still exists. The effect
of gender is less consistent when examining police officers’ behavior. However, there is a noticeable lack
of studies that focus on the study of risk perception among police officers with an emphasis on gender
effect. Thus, this research is an attempt to bridge the gaps between the two fields via the investigation of
whether gender difference in risk perception occurs within one particular occupation – police officers.
Two police departments from a city in the Mid-West were selected for this study. Surveys, both by paper
and electronically, were administered to the departments. The survey examined officers’ perception of
risk in relation to work related hazards and hazards encountered in daily civilian life and related factors
(such as worldviews). The survey was adapted from previous studies on risk perception and the “White
Male Effect”. The results of the research will expand existing literature from a theoretical and practical
viewpoint: theoretically, it will expand the knowledge on how risk perception is viewed in occupational
groups working under a higher-risk environment and the role that gender plays in this process. Practically,
it will provide new information to aid policing agencies in addressing potential gender gap in risk
perception related to work activities.

Copyright by
YI TING CHUA
2012

TABLE OF CONTENTS

LIST OF TABLES ························································································ v
Chapter 1
Thesis Overview ··························································································
Introduction ································································································
Sex and Gender ···························································································
Risk and Risk Perception ················································································
Factors Affecting Risk and Risk Perception ··························································
Trust ·······························································································
Worldviews ·······················································································
Optimistic Bias ···················································································
Policing ····································································································
Research Objectives and Hypotheses ··································································

1
1
3
5
7
7
8
9
9
11

Chapter 2
Introduction ································································································
Literature Review ·························································································
Risk Perception ··················································································
Risks in Policing ·················································································
Gender and Policing ·············································································
Perception and Policing ·········································································
Research Question ························································································
Methodology ······························································································
Survey Administration···········································································
Survey Construction ············································································
Analysis and Findings ···················································································
Univariate and Bivariate Analysis ····························································
Multivariate Analysis ···········································································
Discussions ································································································
Conclusions ································································································

13
15
15
20
21
21
23
24
24
25
27
27
33
35
38

Chapter 3 ·································································································· 40
APPENDIX ······························································································· 43
REFERENCES ························································································· 53

iv

LIST OF TABLES

Table 1: Cronbach's Alphas for Scales prior and
after the Adjustment in Sample Sizes ·································································· 27
Table 2: Descriptive Statistics for Independent
and Dependent Variables ················································································ 29
Table 3: Mean Scores on General Risks and Occupation-Specific Risks
for Female and Male Officers ··········································································· 30
Table 4a: Bivariate Regression of Worldview
Index and General Risks ················································································· 31
Table 4b: Bivariate Regression of Worldview
Index and Occupation-Specific Risks ································································· 31
Table 5a: Bivariate Regression of Trust Index
and General Risks ························································································ 32
Table 5b: Bivariate Regression of Trust Index
and Occupation-Specific Risks ········································································· 32
Table 6a: Multivariate Regression
of General Risks ·························································································· 33
Table 6b: Multivariate Regression
of Occupation-Specific Risks ··········································································· 34

v

Chapter 1

Thesis Overview
This thesis project examines the connection of gender to police officers’ risk perception.
In this chapter, I will provide a broad overview of the concepts and theories that inform this work
from three distinct bodies of literature: 1) sex and gender, 2) risk and risk perception, and 3)
policing. Furthermore, I will examine the extent that these concepts have been studied in
conjunction with each other. Finally, I will provide a description of my research methodology,
and study area.
Chapter 2 of the thesis is an article that will be submitted for publication. Within the
article, there is a more specific discussion of the theoretical background of the study than found
in Chapter 1, an analysis of the survey data collected, practical and theoretical implications of the
work, and suggestions for future research directions. The thesis closes with Chapter 3, which is a
self-reflection on the research process.

Introduction
The effect of gender has been shown to be important in various fields. Particularly, the
effect of gender has been examined in the field of risk perception (Baldassare & Katz, 1992;
Bazley, Lersch, & Mieczkowski, 2007; Cutter, Tiefenbacher, & Solecki, 1992; Finucane, Slovic,
Mertz, Flynn, & Satterfield, 2000; Van Liere & Dunlap, 1980), and policing ( DeJong, 2004;
Maguire & Nolan, 2011; Rabe-Hemp, 2008; Savage, 1993; Schuck & Rabe-Hemp, 2005;
Sjoberg, 2003; Sun, 2007). The risk perception literature has consistently found that females
perceive higher risk compared to males (Barke, Jenkins-Smith, & Slovic, 1997; Slovic, Trust,

1

Emotion, Sex, Politics, and Science: Surveying the Risk-Assessment Battlefield, 1999). The
effect of gender on policing and police behavior is less consistent across various police officers’
behavior (DeJong, 2004; Rabe-Hemp, 2008; Riksheim & Chermak, 1993).
The importance of gender difference in risk perception can be profound due to gender’s
potential to alter and influence one’s behavior. Brody (1984) found that women show a lower
level of support to nuclear power is due to their greater concerns about the safety of nuclear
power plants, both at a general level and within local communities. Baldassare and Katz (1992)
found that the most significant predictor of environmental actions is perceived personal threats,
which women tended to view as more serious than men. Applying this finding and concept to
policing, it can provide some insight as to how officers’ risk perception can affect how they react
to specific situations. Although situational factors play an influential role on officers’ behavior,
the importance of individual-level factors has also been acknowledged (Riksheim & Chermak,
1993).
Despite the influence gender has on behavior, there is a noticeable lack of studies that
focus on the study of risk perception among police officers with an emphasis on the role of
gender. Recent policing studies have also included the examination of risk within the field of
policing, but with a central focus on the application of risk management at the organizational
level (Heaton, 2010). Risk perception processes, which function at a micro-level, are less of a
focus in the policing literature. Thus, the proposed research will investigate whether the gender
difference commonly found in risk perception literature occurs within one particular occupation
– police officers.

2

Sex and Gender
“Gender” and “sex”, although commonly used interchangeably, has specific definitions in
the field of psychology (Stoller, 1974). Sex is defined based on one’s biology and physiology,
such as the sex chromosomes and secondary sex characteristics. Gender, on the other hand, is
defined based on one’s social and cultural experiences. When one is categorized as a male or
female, it is referring to their sex. When one is described as being more masculine or feminine, it
is referring to their gender (Delphy, 1993; Unger, 1979). The distinction was made in attempts to
understand cross-gender individuals (Stoller, 1974). Unger (1979) later emphasized on using the
term “gender” to “… reduce assumed parallels between biological and psychological sex or at
least to make explicit assumptions of such parallels.” (p. 1086). Although the distinction between
gender and sex appears to be quite straightforward, the true dynamics between the two concepts
are far more complicated.
Since the distinction was made between the terms “sex” and “gender”, there has been a
debate on the relationship between the two concepts. The commonly held assumption is that
gender is a result of one’s sex (Delphy, 1993; Unger, 1979). The underlying logic of the
assumption is that one’s biological sex is natural and unchangeable, and thus it acts as a stimulus
variable, which in turn influences one’s sociocultural experiences. One’s responses and
internalization of these experiences then result in the formation of one’s gender identity. Gender
identity is one’s views on the properties that correspond either to being masculine, feminine, or
androgynous (Delphy, 1993; Unger, 1979). Recent argument to rethink this assumption of order
between sex and gender has been made. Delphy (1993) argued that gender actually precedes sex.
Her arguments mainly proposed that the concept of gender is another social method of creating
hierarchy and division of labors, and that the dichotomy of sex is simply a label for the

3

hierarchy. Generally, the main distinction between sex and gender still circle around biological
and sociocultural differences.
The dynamic between sex and gender is perceived as being correlated when examined as
a factor in risk and risk analysis studies. Explanations proposed on the differences between males
and females have been gender-related. Gender roles were one of the proposed explanations for
the differences in risk perception between males and females (Brody, 1984; Gustafson, 1998).
Females are viewed as nurturers and are more likely to value the safety of other individuals and
human lives in general. Thus, females tend to see greater risks and have greater worries about
these risks. Males, on the other hand, tend to value economic growth and thus perceive less risk
in regards to environmental concerns. Another explanation of the gender difference in risk
perception is the authority and power between men and women (Finucane et al., 2000; Flynn et
al., 1994; Gustafson, 1998). Men are more likely to be involved in realms such as politics and
education, which in turn allows them to attain greater power. Thus, they tend to see fewer risks
and have less concern about risk. This explanation is highly related to the “White Male Effect”,
which also suggests sociopolitical factors rather than biological factors to be the cause of
differences in risk perception (Flynn et al., 1994).
It has also been suggested that gender differences from studies in risk perception are due
to research design (Cutter et al., 1992; Gustafson, 1998). The main issue lies in the interpretation
of the word “risk”. Gustafson (1998) suggested that females and males might have different
conceptualizations of risks. Cutter and colleagues (1992) recommended that gender orientation
of activities and situations used to measure risk perception should be taken into consideration. In
general, all of the explanations for differences between men and women are mainly based on
sociocultural variation rather than biological dichotomy. However, the instruments used in the

4

measurement of gender have relied on the self-identification of participants as “men” or
“women” (Finucane et al., 2000; Flynn et al., 1994). Thus, it is safe to conclude that there is an
assumed dynamic between sex and gender in risk and risk analysis studies that recognizes the
difference between the two ideas. Attempts to incorporate the influence of each to some degree
have also been made.

Risk and Risk Perception
Definitions of risk, and risk perception, have been provided from a wide spectrum
ranging from broad points of views to more specific and discipline related definitions (Fischhoff,
Watson, & Hope, 1984; Kelman, 2003). The cause of the disagreement centers on the
subjectivity of risks, the complex dimensions of risks, and the quantification of risks (Fischhoff
et al., 1984; Slovic, 1999). For example, scholars distinguish risk as a socially constructed
concept, whereas danger is real and practical. Thus, it implies the lack of objectivity in risks as
whoever is in control of the society can easily manipulate what constitutes as risks (Slovic,
1999). In other words, it is almost impossible to construct what is known as objective risk
because of factors such as politics and culture in a society. Although it is assumed that risks, as
assessed by experts, appeared to be more objective, it is not the case. The similarity between
experts’ and laypersons’ perception and assessment of risks lies in the systematic processes
applied, and the difference lies in the actual processes selected (Fischhoff et al., 1984; Slovic,
1999). Thus, it is argued that there may not be true objective risks.
Similar to the dimensions of risks discussed above, the debate on the quantification of
risk stems from the wide applications of risk and risk assessment on various topics and fields
rather than in conceptual definitions. When assessing or measuring risks, the focus tends to lean

5

more toward specific potential consequences that the topic of interest (i.e. technology, an event)
can bring, rather than the potential benefits. Some scholars suggested clarification and a more
well-rounded consideration should be made when assessing and measuring risks, with a focus on
negatives and benefits (Fischhoff et al., 1984). Quantifying risks poses similar difficulties in the
method and wordings of the risks itself, and its relevant benefits and consequences (Fischhoff et
al., 1984; Slovic, 1999). The potential influence of administration methods and wordings can
illicit varying responses from participants. What one participant or one group of participants may
see as a benefit may not translate over to another individual or group of participant.
Nonetheless, definitions of risk contain three components (probability, exposure, and
outcome of exposure), commonly expressed formulaically as risk = likelihood x consequence.
The first addresses the possibility of an occurrence, which can be an event, a hazard, or a loss.
The second attends to the potential outcomes if the occurrence does happen; potential outcomes
can be either positive or negative (Kelman, 2003). It is important to note that both aspects of the
risk formula imply uncertainty in terms of the possibility of exposure and the potential outcomes.
On the other hand, scholars defined risk perceptions as “intuitive risk judgment” (Slovic, 1987,
p. 280). In other words, it is an individual’s initial response to a risk, which is often not based on
the complete risk equation.
In the field of risk and risk perception, studies have consistently found that systematic
gender differences occur in how risk is perceived. Gender differences are found when examining
risks in general, or when examining specific components of risk. The risks that have been
examined include daily activities, accidents, environmental risks and concerns, and advanced
technology. The samples of most studies are diverse in terms of age, income, race, education,
and mainly drawn from metropolitan areas (Baldassare & Katz, 1992; Cutter et al., 1992;

6

Savage, 1993). Despite variance in the significant of difference across studies, gender difference
is consistently found across different types of risks. The robustness of this effect is demonstrated
by gender difference in risk perception still being present among practitioners in highly technical
fields (such as chemistry) concerning hazards that affect that field (Barke et al., 1997; Slovic,
Malmfors, Mertz, Neil, & Purchase, 1997). Consistently, studies have shown that men tend to
perceive less risk from an array of hazards, activities and technology, when compared to women.

Factors affecting Risk and Risk Perception
Trust
In addition to the conceptual issues discussed above, risk research has identified variables
that have an effect on risk and risk perceptions. Slovic (1999) has listed factors that need to be
considered when measuring and assessing risks. These factors include trust, emotion, sex, and
worldviews. Trust is measured in terms of confidence toward authorities. The authorities tend to
vary depending on the topic of interest. Siegrist and Cvetkovich (2000) found that trust toward
authorities significantly and negatively correlates with risk perception. The effect became more
prominent when controlling for one’s knowledge about the risks in questions. In other words,
trust in authorities become more prominent under the situations where one lacks knowledge on a
specific risk. A re-examination of three studies on food technologies from the United Kingdom
found that the association between trust and perceived risk is mediated by an individual’s
acceptance of the technology (Eiser, Miles, & Frewer, 2002). It reasonably follows that the
association between trust and risk perception does exist, although the strength and dynamics of
the association are not simple or straightforward. However, an emphasis is placed on the

7

inequality between trust and distrust in terms of its manifestation and affect on an individual’s
responses to risk.
Worldviews
The definition of worldviews was derived from the concept of social relations in culture
theory, which reflects the types of interpersonal relationships prevalent within a culture (Dake,
1991). Although initially measured as an aggregate variable, social relations is suggested to be
the source of influence on an individual’s view of the world and identities (Dake, 1992). Thus,
research on worldviews and risk perception has operationalized worldviews in terms of the
various forms of social relations. There are four basic forms of worldviews: hierarchical,
fatalistic, individualistic, and egalitarian. The hierarchical worldviews endorsed rank and order.
Authority resides in individuals of higher rank, and obedience is expected from individuals of
lower rank. Egalitarian worldviews are quite the opposite from hierarchical. It emphasizes on
equality among individuals. Individualistic worldviews valued individual rights and autonomy,
but allow acceptance of authorities in matters that infringe on rights and autonomy. Fatalistic
worldviews are more isolated and tend to focus on individuals rather than groups (Dake, 1991).
In a study on the relationship between social relations and risk perception, Dake (1991)
has found differences in societal concerns among different worldviews. In general,
egalitarianism correlated positively with items measuring societal concerns whereas the
correlation of hierarchy and individualism was less consistent. In terms of technology and
environment, the difference is more drastic. Moderate positive correlation was found among
egalitarianism and technological and environmental danger, while moderate negative correlation
was found between hierarchy and individualism. Peters and Slovic’s later research (1996),
focusing on the effect of worldviews and affect on risk perception, replicated these results. The

8

study examined different risks such as health, technological, and environmental health risks.
Specifically, Peters and Slovic (1996) examined the role of affect (the feeling or experience of
emotions) and worldviews on support toward nuclear power, and found interactions between the
two factors. The consistency in results illustrates the importance of worldviews in how an
individual responds to risks.
Optimistic Bias
Another variable shown to affect risk perception is optimism. Optimism, or optimistic
bias, is when one unrealistically perceives oneself as having lower risk than others (Weinstein,
1987). Several factors, such as egocentrism and perceived controllability, have been proposed as
the cause of optimistic bias. In a study examining the relationship between optimism and health
problems, Weinstein (1987) discovered that participants had optimistic bias when asked about
their risks of 32 hazards. When examining the effects of optimism on health risks with emphasis
on sexually transmitted diseases, a significant association was found but the strength of the
association was moderate (Van der Velde, Hooykaas, & Van der Pligt, 1992). Similar results are
seen among college-aged individuals and their perception of traffic accident risk (DeJoy, 1989).
The results of these findings are consistent with the potential effects of optimism on one’s risk
perception.

Policing
Historically, women have often been viewed as nurturers and caretakers (Garcia, 2003).
Occupations such as education and social work have been the traditional professional home of
females, whereas fields such as engineering and politics were dominated by males. Interestingly,
females have had a long history within the criminal justice field, compared to other male-

9

dominated fields. For instance, females have been assigned to positions such as prison guards for
females in custody or female prisoners (Brown & Heidensohn, 2000; Wells & Alt, 2005).
Eventually, females were hired into the police field, but without any arrest power at first. Most
often, these females had secretarial duties or duties involving women and children. In the postCivil War era, women were hired as “police matrons” that involved duties similar to that of a
probation officer today (Wells & Alt, 2005). The first policewoman was introduced into the
policing field in 1910, and the first female Police Chief was appointed in 1914 in the United
States. Similar patterns occurred in European countries during the same time period (Brown &
Heidensohn, 2000).
However, women are still one of the minority groups in police organizations. Individuals
in a minority group often face more obstacles in their career paths than individuals of the
majority group. Sexism, harassment, discrimination, and higher stress levels are not uncommon
among policewomen (Brown & Heidensohn, 2000; Burke & Mikkelsen, 2005; Greene & del
Carmen, 2002; Morash, Kwak, & Haarr, Gender Differences in the Predictors of Police Stress,
2006; Wells & Alt, 2005). Studies have also focused on examining differences in behavior and
attitudes between female and male police officers (Gerber, 1996; Greene & del Carmen, 2002;
Hoffman & Hickey, Use of Force by Female Police Officers, 2005; Poteyeva & Sun, 2009).
Gender differences can occur in various aspects of a police officers’ work – attitudes, stress
levels, use of force, perceptions of certain types of crimes and so forth. More recent studies on
gender and police behavior have found gender differences in police behaviors ranging from the
use of force to comforting behaviors toward victims (Bazley et al., 2007; DeJong, 2004; Maguire
& Nolan, 2011; Rabe-Hemp, 2008; Schuck & Rabe-Hemp, 2005; Sun, 2007). In general, the
effect of gender varies among different police officers’ behavior and attitude.

10

Research Objectives and Hypotheses
Given the relative lack of occupation-specific risk perception study, the proposed
descriptive research attempts to shed new light on this particular subfield. This will be achieved
via the inclusion of detailed questions designed to increase understanding of the perception of
both occupation-specific risk and general risk. In addition, the proposed research attempts to
bridge the fields of risk perception and policing, which is a relatively new approach in the field
of criminal justice. The central research goal is the examination of police officers’ perception of
risk (general risk and occupation-specific risk), with a focus on gender difference. In particular,
the proposed research will examine both the direct and indirect effect of gender on risk
perceptions among police officers. The format of the proposed research will bear some similarity
with other risk perception studies, yet it will include elements and knowledge from the field of
policing. The research is descriptive in nature.
There are mainly two hypotheses for this research. A previous study has found no gender
difference when individuals are exposed to a highly stressed environment (Greenberg &
Schneider, 1995). The definition of a stressed environment is an environment that consists of
multiple hazards and risks. The nature of police work is highly uncertain and involves great risks
(Homant et al., 2000; Mayhew, 2001; Uchida et al., 1987). With the nature of police work
resembling the definition of a stressed environment, the proposed finding is the lack of gender
difference in the perception of the occupation-specific risks.
With general risk, gender differences are found when examining risks that are more
general and happen in daily lifestyle (Baldassare & Katz, 1992; Brody, 1984; Cutter et al., 1992;
Finucane et al., 2000; Savage, 1993; Sjoberg, 2003; Van Liere & Dunlap, 1980). However,
studies focusing on specific occupation groups, especially those that are considered to be more
male-dominant, have found gender differences (Barke et al., 1997; Slovic et al., 1997). With
11

police work considered as a more male-oriented occupation, it is predicted that female officers
would still perceive higher risks when compared to male officers in terms of general risks.

The following chapter contained an article that includes a thorough description of
theoretical backgrounds and previous studies, sampling method and survey instrument.
Foundational studies that have specifically examined gender effect on risk perception in police
work and other occupations are examined in relevance to the current study. The construction of
the survey instrument also relies on previous studies on risk perception, especially studies
focusing on the "White Male Effect" (Flynn et al., 1994; Rivers, Arvai, & Slovic, 2010).
Adaptations have also been made from existing survey items with literary support. The article
also provides an in-depth description of the statistical analysis for the study. Results are
discussed and explained within the contexts of the theoretical backgrounds. It also includes a
discussion on the limitations of the study, indiciations for practical applications, and suggestions
for future studies.

12

Chapter 2

Introduction
It has been almost a century since the first waves of women entered the criminal justice
and law enforcement field (Brown & Heidensohn, 2000; Wells & Alt, 2005). Yet, the
marginalization and unequal treatments of female officers are not addressed until about 30 years
ago (Garcia, ''Difference'' in the Police Department : Women, Policing, and ''Doing Gender'',
2003). Despite the long history, the debate on women's capabilities and effectiveness to be good
police officers persist. The origin of the debate is largely due to the perceived masculine nature
associated with police work and police organizations, and the devaluation of characteristics
traditionally attributed to females (Garcia, 2003). Rather than recognizing the potential benefits
that these feminine characteristics may bring to the nature of police work, the debate commonly
focuses on the absence of masculine characteristics.
There are indeed large discrepancies in the perception of female officers' competencies in
the use of force, violence, and other characteristics associated with masculinity. Male police
officers tend to have less favorable attitudes toward female officers' capabilities in carrying out
patrol duties (Vega & Silverman, 1982). This is especially true with specific patrol duties that
have higher probabilities of violence. Similar attitudes are also found among male criminal
justice students (Austin & Hummer, 1994). On the other hand, female officers tend to view
themselves as highly competent in carrying out patrol duties (Vega & Silverman, 1982), and
administrative and supervisory tasks (Kakar, 2002).
Recent studies have begun to explore the potential benefits that women may bring to
police work (DeJong, 2004; Miller, 1999; Sun, 2007). Yet, Hoffman and Hickey (2005) found a

13

lack of gender difference in traditionally masculine activities such as use of force. Exception to
the finding is female officers' less frequent use of extreme force (Rabe-Hemp, 2008). In general,
the debate has subsided. The focus has shifted toward understanding the influence of gender on
policing behavior, and how such influence can be more beneficial to current policing styles.
What is missing from this debate and its recent shift in emphasis is the concept of risk
perception. This concept is important to consider as it may have an effect on how an individual
responds to a stimulus, situation, or context (Baldassare and Katz, 1992). In addition, it is
relatively well-established that females perceive a higher level and frequency of risks than their
male counterparts (Slovic, 1999). In other words, it is possible that gender difference can exist in
how male and female police officers perceive an event encountered while on-duty.
The gender difference in risk perception is also prevalent when examining other maledominated occupations, such as scientists and toxicologists (Barke, Jenkins-Smith, & Slovic,
1997; Slovic, Malmfors, Mertz, Neil, & Purchase, 1997). However, the effect of gender can be
moderated by external factors such as the stress level in the environment (Greenberg &
Schneider, 1995). Thus, it is crucial to determine if such a pattern of gender difference is present
in police work. The presence of a pattern can assist police organizations in the development of
training processes, while the absence of gender difference can change how others perceive
female officers.
In the following section, I will discuss the existing literature on risk perception, gender,
and policing. The literature review is followed by a brief description on the gap that exists
between the three fields. This study attempts to address the gap by examining the effect of
gender on officers' risk perception. The subsequent sections provide an in-depth description of

14

the instruments, the distributions of the instruments, and the findings. The paper concludes with
a discussion of the findings.

Literature Review
Risk Perception
Definitions of risk, and risk perception are varied and are often discipline specific
(Fischhoff, Watson, & Hope, 1984; Kelman, 2003). In general, most definitions of risk contain
three components (probability, exposure and outcome of exposure), and this can be expressed
formulaically as risk (r) = likelihood (l) x consequence (c). The first component, likelihood,
indicates the possibility of an occurrence, which can be an event, a hazard, or a loss. The second
component, consequence, refers to the potential outcomes if the occurrence does happen, and
potential outcomes can be either positive or negative (Kelman, 2003). It is important to note that
both aspects of the risk formula imply uncertainty in terms of the possibility of exposure and the
potential outcomes. In terms of risk perception, scholars define the concept as “intuitive risk
judgment” (Slovic, 1987, p. 280). In other words, it is an individual’s initial response to a risk,
which is often not based on the complete risk equation.
Studies in the risk and risk perception literature have shown systematic gender
differences in how risk is perceived. "Gender", although generally measure as the male-female
dichotomy, is seen as interactive and correlated with the concept of "sex" in risk perception
studies. Hence, the terms "sex" and "gender" are sometimes use interchangeably. Gender
differences are found when examining risks in general, or when examining specific components
of risk. The risks that have been examined include daily activities, accidents, environmental risks
and concerns, and advanced technology. The sample populations of most of these studies are

15

diverse in terms of age, income, race, education, and mainly drawn from metropolitan areas
(Baldassare & Katz, 1992; Cutter et al., 1992; Savage, 1993). Despite variance in the
significance of the difference across studies, gender differences are consistently found across
different types of risks. The robustness of this effect is demonstrated by gender difference in risk
perception still present among practitioners in highly technical fields (such as chemistry)
concerning hazards that affect that field (Barke et al., 1997; Slovic, Malmfors, Mertz, Neil, &
Purchase, 1997). Repeatedly, studies have shown that men tend to perceive less risk from an
array of hazards, activities and technology, when compared to women.
The gender difference seems to hold true despite age. When examining driving behavior
of young adults, gender difference was lacking in terms on their perceived frequency of risky
driving behavior and the probability of being caught (DeJoy, 1989). However, with regard to the
probability and severity of accidents, males considered risky driving behavior as less likely to
result in accidents and less serious. These results reflect on the generalization of gender
difference in risk perception across different types of contexts and populations.
However, there are exceptions to the well-established gender difference in risk
perception. When the studies are restricted to certain contexts, little to no gender difference in
risk perception is found (DeJoy, 1989; Greenberg & Schneider, 1995). For example, there is an
absence of gender difference in the perception of environmental risk under the conditions of a
stressful environment (Greenberg & Schneider, 1995). In this case, a stressful environment is
defined as an environment with multiple hazards such as landfills, crime, and major highways.
The main theoretical backgrounds that support the lack of gender differences in environments
with higher stress have its origin in the studies of victims of natural disasters rather than risk

16

perception.. These studies tend to show no gender difference for victims who had experienced
natural disasters.
Several explanations have been proposed for the gender differences found in risk
perception. One set of researchers has suggested that the origin of the differences lies in gender
roles. Females are viewed as nurturers and are more likely to value the safety of other individuals
and human lives in general. Thus, females tend to perceive greater risks and have greater worries
about these risks. Males, on the other hands, tend to value economic growth and thus perceive
less risk with regards to environmental concerns (Brody, 1984; Gustafson, 1998). Another
explanation of the gender difference is the difference in authority and power between genders
(Finucane et al., 2000; Flynn et al., 1994; Gustafson, 1998). Men are more likely to be involved
in politics and education, which allows them to attain greater power. Thus, they tend to perceive
fewer risks and have fewer concerns about risk. This explanation is highly related to the “White
Male Effect” (Flynn et al., 1994), which suggested that the difference in risk perception is
derived from sociopolitical rather than biological factors.
It has also been suggested that gender differences in risk perception are due to research
design (Cutter et al., 1992; Gustafson, 1998). The main issue lies in the interpretation of the word
“risk”. Gustafson (1998) suggested that females and males may have different conceptualizations
of risks, while Cutter and colleagues (1992) recommended that the gender orientation of
activities and situations used to measure risk perception should be taken into consideration.
Although the explanations require additional supporting evidence, study results have illustrated
that gender does make a difference in the perception of risk.
Some explanations of gender difference in risk perception have been investigated
between specific occupations (Barke et al., 1997; Slovic et al., 1997). Among professional

17

scientists, females still perceive greater risk in terms of nuclear technology, and environmental
and health hazards despite their levels of education (Barke et al., 1997). However, an
individual’s area of training in science can affect his or her risk perception. For example,
physical scientists perceive less risk compared to life scientists, but within each field, gender
difference is still evident. The ranking of sub-groups with the least perceived risk to the greatest
is as follows: male physical scientists, female physical scientists, male life scientists, and female
life scientists (Barke et al., 1997). The gender difference within life scientists was significant,
and there is virtually no difference between the risk perception of female physical scientists and
male life scientists.
Similar results were found when surveying toxicologists from Britain (Slovic et al., 1997).
Female toxicologists are more likely to rate 27 out of 29 items as having moderate or high risk.
In addition to risk perception, gender difference was also found in attitudes toward chemical
risks. Female toxicologists have more negative feelings about chemicals than male toxicologists.
The survey mainly focused on health risks of various types of hazards such as technology,
lifestyle, and crime. The survey had five sections: (1) risk perception, (2) attitudes and opinions,
(3) worldviews, (4) toxicological evaluation, and (5) personal characteristics and professional
experience. These two studies provide a framework for this study, but the population of interest
is a new occupation group – police officers.
In addition to gender, risk perception can be influenced by other individual-level factors.
Slovic (1999) has listed factors that need to be considered when measuring and assessing risk.
These factors include trust, emotion, and worldviews. Trust is measured in terms of confidence
toward authorities. The authorities tend to vary depending on the topic of interest. Siegrist and
Cvetkovich (2000) found that trust toward authorities significantly correlates with risk

18

perception. The effect became more prominent when controlling for one’s knowledge about the
risks in questions. In other words, trust in authorities become more prominent under the
situations where one lacks knowledge on a specific risk. A re-examination of three studies on
food technologies from the United Kingdom found that the association between trust and
perceived risk is mediated by an individual’s acceptance of the technology (Eiser, Miles, &
Frewer, 2002). It reasonably follows that the association between trust and risk perception does
exist, although the strength and dynamics of the association are more complex. There appear to
be other factors that can intervene with the correlation.
Worldview has also been previously identified as a factor influencing one's perception of
risk. The definition of worldviews was derived from the concept of social relations in culture
theory, which reflects the types of interpersonal relationships prevalent within a culture (Dake,
1991). Although initially measured as an aggregate variable, the concept of social relations was
suggested to be the source of influence on an individual’s view of the world and identities (Dake,
1992). Thus, later research on worldviews and risk perception has operationalized worldviews in
terms of the various forms of social relations. There are four basic forms of worldviews:
hierarchical, fatalistic, individualistic, and egalitarian (Dake, 1991). Dake (1991) has found
differences in societal concerns among different worldviews. Peters and Slovic’s later research
(1996), focusing on the effect of worldviews and affect on risk perception, replicated these
results. The study examines different risks such as health, technological, and environmental
health risks. Specifically, Peters and Slovic (1996) examined the role of affect (the feeling or
experience of emotions) and worldviews on support toward nuclear power, and found
interactions between the two factors. The consistency in results illustrates the importance of
worldviews in how an individual responds to risk.

19

Risks in Policing
To examine risk perception among police officers, it is necessary to understand the
specific risks encountered in policing. The nature of policing and police work has its unique
features. Skolnick (2010) discusses the "working personalities" of policeman. In general,
working personalities describe the impacts that one's occupations have on one's view of the
world. In the context of policing, these personalities are developed due to two outstanding
features of police work: danger and authority. These two features interact to create a sense of
social isolation and heightened suspicions toward others among police officers (Skolnick, 2010).
Although danger and risks are two distinct concepts, the unique nature of police work results in
police officers being exposed to risks that are specific to their field.
In addition, police work involves a high level of uncertainty (Wilson, 1968). For
example, police officers often times respond to a call or respond to a citizen with little
information. In addition to uncertainty, situations encountered during work can become
dangerous and life-threatening (Homant, Kennedy, & Hupp, 2000; Mayhew, 2001; Uchida,
Brooks, & Kopers, 1987). Examples of common police encounters feature a high level of
uncertainty include domestic situations (Rabe-Hemp & Schuck, 2007; Uchida et al., 1987), and
encounters where quick judgment and use of force are required (Homant et al., 2000). The risk of
being assaulted during work is only one of many risks that officers potentially face. Some
examples of other risks include death, infectious diseases, fatigue, and injuries (Corrigan, Lester,
& Loftus, 1980; Mayhew, 2001). To better reflect the wide range of risks involved in policing,
there has been an effort to widen the definition of danger faced by police officers and to include
injuries and deaths that occur as a result of accidents (Brandl & Stroshine, 2003).

20

Gender and Policing
Much recent studies on gender and police behavior have found gender differences in
some police behaviors such as the level of force used (Bazley et al., 2007; Rabe-Hemp, 2008;
Schuck & Rabe-Hemp, 2005), comforting behaviors toward victims (DeJong, 2004; Sun, 2007),
and attitude on work assignments (Maguire & Nolan, 2011). Gender differences are also found
when examining the likelihood of police officers themselves experiencing victimizations. With
domestic situations, Rabe-Hemp and Schuck (Rabe-Hemp & Schuck, 2007) found that female
officers were more likely to be victimized when handling family violence calls. However, no
gender difference was found for police-citizen encounters in general.
There are other instances where the influence of gender on police behavior seems to be
lacking. For example, it has been suggested that gender only plays a role when examining the
types of weapons used in incidents and arrests where force was used (Hoffman & Hickey, 2005).
In general, factors that have an effect on gender differences in officers’ behavior include the
gender of paired officers (Schuck & Rabe-Hemp, 2005), assignments (Rabe-Hemp, 2008), trust
toward fellow colleagues (Maguire & Nolan, 2011), and situational contexts such as the type of
crime and suspect behavior (Bazley et al., 2007; DeJong, 2004). Thus, it is in the interest of the
researcher to examine whether there are patterns of officers’ perceptions, using gender as the
main variable. Locating the pattern, if any, is important since risk perception has an effect on
one’s behavior (Baldassare & Katz, 1992; Brody, 1984).
Perception and Policing
Plenty of research has examine how gender plays a role in the actions and behavior of
police officers. An early study by Corrigan and his colleagues (1980) found officers who
perceived higher level of danger were more likely to exert their authority. The perception of

21

danger is different than the perception of risk, but it illustrates the impacts perception can have
on an individual's behavior. This study, however, provided no insight regarding gender
differences since the sample only included male police officers.
Another study examining perception of physical danger in the context of police
departments found gender difference, but in the opposite direction expected (Jermier et al.,
1989). Specifically, female employees in a police department, regardless of their occupational
groups, perceived less physical danger than their male counterparts. The authors provided two
explanations for the finding. The first explanation was that female officers and employees were
less likely to be assigned to tasks involving physical danger. The second explanation was
females’ lower frequency in constructing images of danger. Although somewhat related to this
research, the focus on the study was to understand organizational behavior. It relied heavily on
organizational theories, and placed its emphasis on the role of physical danger within
organizational settings where employees were exposed to a varying level of danger and risk.
There have also been attempts to understand the perception of risk in policing. This is
done through the examination of police officers’ cognitive decision-making processes (Flin,
Pender, Wujec, Grant, & Steward, 2007). Officers were first asked to group 19 situations on
different themes. In the second part of the study, the same group of officers were asked to assess
each situation factors known to affect the decision-making process. These factors include amount
of time available, risks to oneself, risks to others, familiarity to the situation, and level of stress
involved. In addition to the first group of officers, a new group of officers were recruited for this
part of the study. The descriptions of the situations used in both parts were written by
experienced officers. However, gender was not a factor that was examined in the study.

22

Research Question
Although situational factors play an influential role on officers’ behavior, the importance
of individual-level factors has been acknowledged (Brown & Frank, 2006; Riksheim &
Chermak, 1993; Rydberg & Terrill, 2010; Sherman, 1980). Thus, the central goal of this study is
to examine police officers’ perception of risk (general risk and occupation-specific risk), with a
focus on potential gender differences. Given the relative lack of occupation-specific risk
perception studies in the field of criminal justice, this analysis attempts to bridge the fields of
risk perception and policing, which is relatively unique for both fields. In addition, this approach
should provide some insight into how officers’ perception of risk can affect their reactions to
specific situations. Also, the research examines the effects of known factors on risk perceptions
relationships among police officers. In general, the research is descriptive in nature.
There are two central hypotheses for this research. The first hypothesis is that there will
be a lack of gender difference in the perception of occupation-specific risks (H1). The main
reason behind the hypothesis is based on the existing literature. A previous study has found no
gender difference when individuals are exposed to a highly stressed environment (Greenberg &
Schneider, 1995). The definition of a stressed environment is an environment that consists of
multiple hazards and risks. The nature of police work is highly uncertain and involves great risks
(Homant et al., 2000; Mayhew, 2001; Uchida et al., 1987). With the nature of police work
resembling the definition of a stressed environment, it is logical to assume that the same patterns
will be found in police officers' risk perception.
The second hypothesis, on the other hand, is that female officers would still perceive
higher risks when compared to male officers in terms of general lifestyle risks(H2). With general
risk, gender differences are found when examining risks that are more general and happen in

23

daily lives (Baldassare & Katz, 1992; Brody, 1984; Cutter et al., 1992; Finucane et al., 2000;
Savage, 1993; Sjoberg, 2003; Van Liere & Dunlap, 1980). However, studies focusing on specific
occupation groups, especially those that are considered to be more male-oriented, have found
gender differences (Barke et al., 1997; Slovic et al., 1997). With police work considered as a
more male-dominant occupation, it is predicted that there will be a gender difference in the
perception of general lifestyle risks among police officers.

Methodology
Survey Administration
The research design for this study is a non-experimental descriptive study. The method of
data collection was through survey administration, which began in May 2012 and ended in
August 2012. Convenience sampling was utilized for this research due to its preliminary nature.
Initially, a total of four departments located around or within a university city in a Midwestern
state were approached through e-mails to participate in the research. The departments in towns
adjacent to this university city either declined or did not response. The two other departments,
which were both located within the university city, agreed to participate in the survey. The police
chiefs of both departments were given the proposal of the research project during an initial faceto-face meeting.
Electronic distribution of the survey has been preferred over the more common method of
face-to-face survey administration during roll call due to time and budget constraints. Although
electronic administration of surveys has been known for having a low completion rate, there are
several possible methods of administration that can help increase completion rate (Evans &
Mathur, 2005; Sax, Gilmartin, & Bryant, 2003; Vicente & Reis, 2010). One of the methods is to

24

increase interactivity within the survey, using radio buttons instead of drop-down menus, and
keeping the survey at a moderate length (Vicente & Reis, 2010). Another method to increase
response rate is to provide subjects with a hard-copy of the survey, but also provide the option to
complete the survey online (Sax et al., 2003). This method has been found to be effective with
college students. For the current study, a modified version of the mentioned method was adopted
due to specific preference of the departments. Thus, rather than providing both electronic and
hard-copy surveys to each department, only the department's preferred method was used.
Department A is a university police department with 69 sworn police officers.
Department B is a local police department with 58 sworn officers and 43 non-sworn personnel.
The method of survey administration differed between the two departments. For Department A,
electronic distribution of surveys was preferred. For Department B, printed versions of the
surveys was preferred. Due to the difference in administration methods, the two departments had
different means of notifying the potential participants: sworn police officers. Several e-mails
containing a link to the survey were sent through the period of data collection for Department A.
For Department B, the survey was left with the Chief to be distributed at the beginning of the
study. Completed surveys were collected periodically. There were periodic pick-up of completed
surveys. The purpose of the research and approximated time of completion were stated at the
beginning of each survey.
Survey Construction
There are four sections of the survey. A copy of the survey used can be found in
Appendix A. On average, the time necessary for completing the survey was between 15 to 30
minutes. The first section consists of two lists of risk. The first list includes activities and risks
that have been commonly used in risk perception studies. These activities includes

25

environmental risks and health concerns (Brody, 1984; Flynn et al., 1994; Savage, 1993). The
second list includes activities and situations that police officers could encounter during their
work shift. The list of activities and situations are derived from the literature (Flin et al., 2007;
Homant et al., 2000; Mayhew, 2001; Uchida et al., 1987) and through interviews with experts.
These activities and situations include domestic violence, situations involving use of force,
situations involving narcotics, situations involving hazardous materials, terrorism and bomb
threats, and traffic stops and accidents. The complete lists can be found in Appendix A. For this
section, subjects were asked to rank these activities and situations on a scale from 1 to 10, which
range from risky/dangerous to least risky/dangerous.
In the second section, subjects were asked to estimate the frequency of encountering
situations and activities listed in the first section. These questions provide greater insights as to
why certain situations are viewed as more risky by an individual. For these questions, a Likerttype scale is used. The third section includes various statements from indexes and scales that
measure the following variables: trust, worldviews, dispositional optimism, and control. These
variables have been shown in previous studies to correlate with one’s perception of risk
(Finucane et al., 2000; Flynn et al., 1994; Rivers et al., 2010). The specific scales and items in
this survey are modified and adopted from those in previous studies (Finucane et al., 2000; Flynn
et al., 1994; Rivers et al., 2010). The last section collects information on the subjects’
demographic information. Specifically, it includes questions with regards to officers’ income,
education, rank, length of service, and sex.

26

Analysis and Findings
Univariate and Bivariate Analysis
Fifty-eight printed versions of the surveys were distributed in Department B, and all 69
sworn officers of Department A were contacted. The total number of collected surveys is 64,
with 51 surveys collected electronically from Department A and 13 surveys collected physically
from Department B. The total return rate is approximately 50.4%. The return rates for
Departments A and B are respectively 73.9% and 22.4%. However, for the purpose of this
research, cases with missing data on the variables of interests are excluded. Thus, the final
sample size is smaller (N=51). Thirty-eight surveys, which accounts for 74.5% of the final
sample size, comes from Department A. Department B has 13 surveys, which is 25.5% of the
total sample.
In the final sample, the majority of the participants reported their ranks as " patrol
officers". The highest education level achieved by most officers in this study is a four-year
college degree, followed by a Master's degree. The mean of participants' age is about 40 years
old and the range is 37 years. Length of service ranges from 1 year to 39 years, with a mean of
16.3 years.

Table 1. Cronbach's Alphas for Scales Prior and After the Adjustment in Sample Sizes
Names of Scales

Prior to Adjustment

After Adjustment

General Risks

0.839

0.831

Occupation-Specific Risks

0.867

0.876

Worldview

0.647

0.669

Trust

0.851

0.858

27

Table 1 includes the Cronbach's alphas for each of the scale created or adopted for this
research. As shown, the majority of the scales (General Risks, Occupation-Specific Risks,
Worldview Index, and Trust Index) have relatively high Cronbach's alphas (α > 0.831), with the
exception of the Worldview Index, both before and after the adjustment in sample sizes (α >
0.647). The reason for the lower alpha is the measurement of the Worldview Index, which is a
composite measurement of six sub-indexes with a total of 16 items. The composite measurement
can hinder the reliability of the scale since each index attempts to measure different and specific
concepts. The high Cronbach's alphas for the other scales indicate that these scales have high
reliability.
For this research, the main independent variable is the officer's gender. Female officers
are coded as "1", while male officers are coded as "0". Approximately 25.5% of the sample
consists of female officers, and 74.5% of the sample members are male. For Department A, there
are 27 male officers and 11 female officers, which account for 71.1% and 84.6% of each
respective gender. For Department B, there are 11 male officers and 2 female officers, which are
respectively 28.9% and 15.4%.
The second independent variable is the Worldview Index. As mentioned, this variable
was coded from six different sub-indexes. The total possible scores for the Worldview Index
range from -32 to 32. For the current data, the range of the Worldview Index is between -18 and
27. The mean is approximately equal to -3.78. The standard deviation for the Worldview Index is
about 7.45, suggesting a high level of dispersion in the data.
The third variable of interest is the Trust Index. Since this index is measured with 11
items, the total possible scores range from -22 to 22. For the current data, the range of the Trust
Index is from -21 to 12. The mean of the variable is approximately equal to -0.33. The wide
28

range of scores and low mean value suggest a lower of level of trust in officials to manage
among the respondents. The high level of standard deviation (7.28) also suggests a high level of
dispersion in trust level.

Table 2. Descriptive Statistics for Independent and Dependent Variables
Variable
Female

Minimum

Maximum

Mean

Standard Deviation

0

1

0.25

0.44

Worldview

-18

27

-3.78

7.45

Trust

-21

12

-0.33

7.28

General Risks

17

100

55.94

19.63

OccupationSpecific
Risks

12

116

82.67

19.13

Age

23

60

39.83

8.23

1

39

16.33

8.59

Length of
Service

There are two main dependent variables for the research: general risks and occupationspecific risks. The total score for each respondent on the General Lifestyle Risks Scale is
calculated by summing respondents' scores across 12 items. The total possible range of scores is
from 12 (perception of low risk) to 120 (perception of high risk). The range of scores for the
current data is from 17 to 100. The mean and standard deviation were 55.94 and 19.63
respectively. This suggests moderate dispersion within data. In general, officers perceive
moderate levels of general risks.

29

Each respondent's total score for the Occupation-Specific Risks Scale is calculated by
summing respondents' scores on 12 items. The total possible range of scores is from 12
(perception of low risk) to 120 (perception of high risk). For the current data, the scores for the
scale ranges between 12 and 116. The mean is 82.67, with a standard deviation of 19.13, which
suggests some dispersion within the data. It also suggests that most officers perceive occupationspecific risks as more risky when compared to their perception of general lifestyle risks.

Table 3. Mean Scores on General Lifestyle and Occupation-Specific Risks for Female and Male
Officers
Dependent Variable

Male Officers

Female Officers

Total

General Risks*

54.29

60.77

55.94

Occupation-**
Specific
Risks

81.26

86.77

82.67

* t=1.028, Significance=0.309
** t=0.849, Significance=0.376

To test the hypotheses, the Student's t-Test is conducted for the three main independent
variables (Officers' gender, worldview, and trust) and the two main dependent variables (general
risks and occupation-specific risks). The results are presented in Table 3. For officers' gender and
general risks, pooled variance estimates are used since the F-test did not reach significance. The
absolute mean difference between female and male officers' perception of general risks is equal
to 6.48, with female officers having higher risk perception. However, the difference is
insignificant. This finding does not support the second hypothesis.
When specific items on the General Risks Scale are analyzed by gender, none of the
items reach significance. Although the differences are insignificant, female officers generally
30

have higher mean on perceived risks compared to their male counterparts on the specific items,
with the exception of three items: vaccines, medical X-rays, and drinking sugar-sweetened
beverages. The range of the absolute difference between female and male officers on each item is
wide. It ranges from 0.07 ("Eating fatty food") to 2.15 ("Cigarette smoking").
When examining the specific items, none of the 12 items on the Occupation-Specific
Risks Scale reach significance. However, like the General Risks Scale, female officers have
higher means on the majority of items. There are two exceptions where male officers have higher
means: vehicle accidents and being assaulted by suspects. The range of item-specific means is
between 0.18 ("Investigating a methamphetamine laboratory") and 1.18 ("Arresting a suspect
surrounded by bystanders").

Table 4a. Bivariate Regression of Worldview Index and General Risks
Regression coefficient (B)

Standard Error

Y-Intercept

55.04

3.10

Worldview

-0.24

0.38

Beta

-0.09

p=0.530

Table 4b. Bivariate Regression of Worldview Index and Occupation-Specific Risks
Regression coefficient (B)

Standard Error

Y-Intercept

81.45

3.02

Worldview

-0.32

0.36

Beta

-0.13

p=0.382

The Worldview Index is negatively correlated with the General Lifestyle Risks Scale and
the Occupation-Specific Risks Scale. The correlation with general lifestyle risks is much weaker
31

than the correlation with occupation-specific risks. The negative correlation suggests that as
one's worldview score increases, there is a decrease in risk perception. However, since
worldview is a composite score from multiple sub-indexes, it is impossible to untangle which
specific aspects of one's worldview is correlated with risk perception. Both of the regression
models failed to explain any variations in risk perception since the p-values for each risk did not
reach significance.

Table 5a. Bivariate Regression of Trust Index and General Risks
Regression coefficient (B)

Standard Error

Y-Intercept

56.08

2.75

Trust

-0.42

0.38

Beta

0.16

p=0.271

Table 5b. Bivariate Regression of Trust Index and Occupation-Specific Risks
Regression coefficient (B)
Y-Intercept

Standard Error
2.67

0.44

Trust

82.81

0.37

Beta

0.17

p=0.244

The models for Trust Index have also failed to address any variations since the p-values
for both models are greater than the level of 0.05. The Trust Index is positively correlated with
both scales. The correlation is relatively weak. In general, it appears that as one's level of trust
toward officials increases, the level of risks perceived increases. Specifically, with an increase of

32

one point on the Trust Index, there is an increase of 0.42 in the General Risks Scale and an
increase of 0.44 in the Occupation-Specific Risks Scale.
Multivariate Analysis
To determine the effect of all three independent variables on both of the dependent
variables, it is necessary to conduct a multivariate analysis. To do so, the sample size has to be
further reduced to exclude cases where there are missing information on some of the control
variables (officers' rank, length of service, age, etc.). The final sample size for the multivariate
analysis diminishes to 45. Most of the demographic variables are not affected by this reduction in
sample size; however, the proportion of female and male officers has changed. Male officers
now account for 73.3% of the sample, while female officers account for 26.7% of the sample.
Due to the small sample size, the results of the multivariate analysis are to be taken as
preliminary and exploratory rather than conclusive.

Table 6a. Multivariate Regression of General Risks
Regression coefficient (B)

Standard Error

28.25

30.69

Female

3.39

6.91

0.21

Worldview

0.06

0.43

0.02

Trust

0.21

0.39

0.08

Race/
Ethnicity

3.69

5.28

0.11

Age

1.27

1.20

0.53

-9.02

4.99

-0.27

Y-Intercept

Highest
Education

33

Beta

Table 6a. (cont'd)
Regression coefficient (B)
Current
Rank
Length of
Service
p>.05, r2=.255

Standard Error

Beta

2.29

2.26

0.19

-0.62

1.20

-0.27

For both the Occupation-Specific Risks and the General Lifestyle Risks, none of the
variables included in the regression models reach significance. Hence these regression models
failed to address the variance in these dependent variables. A collinearity diagnostic is performed
for both models to determine if there were any correlations among the variables, which can
account for the insignificance of the regression models. The results of the diagnostics showed the
variance inflation factors for participants' age and participants' length of service are higher than
other variables. Subsequent regression models were conducted where participants' age,
participants' length of service, or both variables were omitted. In these subsequent analyses, the
models and the variables remained insignificant.

Table 6b. Multivariate Regression of Occupation-Specific Risks
Regression coefficient (B)

Standard Error

85.60

32.58

0.82

7.34

0.02

-0.36

0.46

-0.14

Trust

0.47

0.42

0.18

Race/
Ethnicity

7.26

5.60

0.21

Y-Intercept
Female
Worldview

34

Beta

Table 6b. (cont'd)
Regression coefficient (B)

Standard Error

Beta

Age

-0.35

1.28

-0.15

Highest
Education

-2.85

5.30

-0.09

Current
Rank

3.16

2.40

0.26

Length of
Service
p>.05, r2=.1

0.47

1.27

0.20

Discussions
Based on the results from the statistical analyses, H1 is supported in this analysis. There
was a lack of gender difference in the perception of occupation-specific risks. The results
indicate that female officers did have higher risk perception in comparison to male officers.
When looking at specific occupation-specific risk, female officers perceived higher risks on most
items with the exception of two items. In general, however, the univariate analyses showed these
differences among male and female police officers in terms of the total or individual scores on
each risk to be insignificant. This result replicates previous findings of no gender difference in
risk perceptions in high-stressed environments (Greenberg & Schneider, 1995). Police work has
been recognized as an occupation that has a high level of uncertainty and risks (Homant et al.,
2000; Mayhew, 2001; Uchida et al., 1987). The results from this research imply that the
definition of a high-stressed environment can be applied not only to daily encountered
circumstances, but also the work environment.
Surprisingly, the results from the statistical analyses did not support the second
hypothesis. No significant gender differences were found among officers on their perception of
35

general lifestyle risks. However, the differences for the total and individual score on each risk
were in the expected direction. In other words, female officers tend to have higher mean scores
than male officers. There were only three risks where male officers had higher mean scores than
female officers, which were vaccines, medical X-rays and drinking sugar-sweetened beverages.
This is contrary to the results from previous risk and risk perception studies (Baldassare & Katz,
1992; Brody, 1984; Cutter et al., 1992; Finucane et al., 2000; Savage, 1993; Sjoberg, 2003; Van
Liere & Dunlap, 1980). One possible explanation is the effect of the small sample size on the
statistical analyses. The small sample size could have masked or eliminated the potential
differences. Another potential explanation is a spill-over effect. The lack of gender difference in
risk perception on general lifestyle risks could have been a result of officers' necessary
adaptation and changes in perceptions for their high-stress job environments. The adaptations
and changes may be only for job-related risks, or may be so fundamental that it affects how risks
are perceived in general. This effect may have been more significant for female officers because
of the unique and possibly additional sources of stresses and challenges that originate from
working in a male-dominated occupation (He, Zhao, & Archbold, 2002; Morash & Haarr,
Gender, Workplace Problems, and Stress in Policing, 1995).
The bivariate analysis on worldview and trust indexes and the two dependent variables
are preliminary attempts to examine the findings in previous literature (Dake, 1991; Eiser et al.,
2002; Peters and Slovic, 1996; Siegrist and Cvetkovich, 2000). Negative correlations were found
between worldview index and both types of risks. The trust index was negatively correlated to
the general risks but positively correlated with the occupation-specific risk. One explanation for
the positive correlation between trust and occupation-specific risk is the hierarchical structure of
police work. Officers may be required to obey and place trust not only in their superiors.

36

However, police officers typically encounter these risks when they are out on-duty with their
partners, and not their superiors. Hence, it is possible that the correlation is not meaningful. In
general, the findings from these results are consistent with results from previous studies yet the
correlations between these variables are generally weak. The correlations between the trust index
and both types of risks are slightly stronger than the correlations of worldview index. Since the
two variables are composite measures, it was hard to conclude which specific components of the
index was correlated more strongly to the two types of risks.
The findings of the multivariate analysis did not indicate any of the independent or
control variables to be strongly related to either of the dependent variables. This finding suggests
that the dynamics involved in shaping an individual's perception of risk are much more complex
than is currently known. The results are to be understood with caution. The applicability and
generalizability of the results from this research are limited due to the focus on one specific
occupation group. The dynamics and nature of this specific occupation group - police officers are drastically different from most occupation groups. In addition, the results from this analysis
needs to be interpreted with care due to the small sample size. Lastly, this research was designed
as a preliminary examination of potential effects rather than to examine casual relationships
between the variables.
The main limitation of this research is the small size of the sample. The size of the
sample limits the types of statistical analysis that can be applied. Although insightful, the
statistical analyses - bivariate and multi-variate - of this research offered only initial and surface
knowledge on the factors and dynamics affecting the perception of risk among police officers.
Another limitation of this research is a lack of variety in the types of police departments
involved. The close proximity of both participating departments to a college could have potential

37

mediating effect on the correlation between gender and risk perception. It reduces the
applicability of the findings to departments in other types of locations.
Future research needs to address the issue of sample sizes. One approach to the issue is to
recruit police departments from different locations and of various capacities. This can potentially
increase the number of participant officers, particularly the number of female officers. In
addition, it increases the reliability of the research design because of the greater diversity in
participants. The increase in both numbers and diversity of participants can provide a more
complete picture on the relationship between gender and risk perception among police officers. It
will also allow us to apply more advanced statistical analysis to examine the complexity
involved. Qualitative methods, such as face-to-face interviews and grounded theory, should also
be utilized to provide more in-depth understandings on how police officers perceive risks in their
daily lifestyles and work environments.

Conclusion
Despite the long history of women's participation in policing, the acceptance of their
presence in this male-dominated occupation remains relatively low (Vega & Silverman, 1982;
Wells & Alt, 2005). The current study addresses a missing gap that can provide new insight in
understanding the dynamics between gender and police work. Risk perception has been
previously shown to affect an individual's behavior and decision-making framework (Baldassare
& Katz, 1992; Brody, 1984; Flin et al., 2007). Two types of risks are measured in this study:
general lifestyle risks and occupation-specific risks. The findings indicate the lack of gender
difference in the perception of both types of risks. In other words, female and male police
officers perceive the same level of risks with regard to scenarios officers can potentially face
while on-duty or in their daily routines.
38

Other factors previously found to affect risk perception, such as trust and worldview, are
also examined in this study. Although not the main focus, it provides some degree of replication.
Both types of risks were found to correlate with these two factors, but were generally weak. The
directions of the correlations were generally consistent with previous findings. Multivariate
analysis, however, found none of the variables of interest to be significant, correlated with
general lifestyle risks and occupation-specific risks. This indicates other factors having more
influence on one's perception of risks.
However, the findings of this study should be interpreted with care as the sample size was
relatively small. The findings may not be applicable to other occupations or in general. Future
researchers should definitely account for such issues. As well, future research should attempt to
understand the dynamics of risk perception among police officers with more qualitative
methodology. This may capture information that is not translated into survey data. Nonetheless,
it is an important initial step in encouraging new perspectives of evaluating police competency
and gender differences within policing.

39

Chapter 3

The importance of this thesis to me, especially in shaping my future academic and career
path, is undeniable. Upon receiving the admission decision for the doctoral program from
Michigan State University, I debated if I should continue with this thesis or switch over to
writing a policy paper. However, I was very pleased with my decision to go through with the
thesis track. Through the process, I have learned multiple valuable lessons. Some of the lessons
are related to specific technicalities involved in research conduct. Some other lessons are more
related to personal growth. In various ways, this thesis has become a practice and a foreshadow
of the dissertation process.
Due to the interdisciplinary nature of the topic, various obstacles were encountered at the
initial planning stage. The original intention was to search through databases for usable existing
datasets. However, due to the lack of convergence in the variables of interest (i.e. risk perception,
gender, and police officers), it was necessary to design and develop instruments and scales from
existing literature. I truly appreciate the helps and advices from all of my committee members
during this process. Their guidance assisted the development of scales that truly reflect and
capture what I wished to investigate. Without it, I would have easily lost sight of my research
during the stage of literature review. Constant reminders of my research focus proved to be
helpful.
The second challenge was in recruitment. A total of four police departments were
approached. However, one declined to participate (due to lack of resources), while another
department never replied to the request. It was frustrating because this meant a reduction in the
potential sample size. The size of the sample is very important because it restricts the types of

40

statistical analysis that can be utilized, and hence the results that can be obtained. Luckily, the
two participating departments provide a sufficient pool. Combined with the fact that the research
is preliminary, the issue was overcome.
Another challenge that was encountered was the length and return rate during data
collection. The length of time for data collection exceeded my expectations. The end date of the
period was pushed back for a few times due to insufficient numbers of surveys returned. At first,
I had a hard time accepting the extension because it delayed my schedule of completion.
Eventually, I realized that no matter how impeccable and perfect the plan, we can never account
for all the emergencies and sudden changes in life. I learned to be patient while handling the
situation professionally. Although the target number was much lower than proposed and
expected (i.e. 100 surveys), the results extracted from the collected data was adequate for the
purpose of this research.
The difficulties and challenges that were encountered from the initial planning stage to
the data collection stage are great overviews of the dissertation process. My chosen research
focus gave me the opportunity to experience a process that resembles the process of writing a
dissertation at a de-magnified level. It would serve as a guidance and reference point for me, as I
decide what and how I would approach my dissertation. In general, the journey of writing this
thesis makes the journey of writing a dissertation seems more realistic and manageable.
This thesis has been a great learning experience and opportunity for me. Despite the
various challenges and obstacles that emerged throughout the process, it was a pre-dissertation
challenge for me. Various abilities and skills were obtained and mentored when overcoming
these issues and completing the thesis. These abilities and skills would prove to be vital in the

41

process of dissertation, especially in the development of instruments and the process of data
collection.

42

APPENDIX

43

SURVEY

Gender and Risk Perception
There are several parts of the survey. At the start of each section, there are corresponding
instructions. The survey is printed on both sides of a page. If you are unsure about any items
please refer to the contact information found on the consent form. Thank you very much for your
participation!
SECTION 1
General Risks
How much risk would each of the following items pose to your general well-being? Please rate
on the scale of 1 (Low risk) to 10 (High risk).

SCALE ON RISKINESS
1

2

3

4

5
Some
Risk

Low
Risk

6

7

8

9

Some
Risk

10
High
Risk

AIDS





















Vaccines





















Medical X-rays





















Climate change





















Natural disasters





















Water pollution





















Bacteria in food





















Eating fatty food









































Cigarette smoking





















Drinking alcohol





















Stress





















Drinking sugarsweetened
beverages

44

Occupation-Specific Risks
Imagine each item below as a real scenario encounters during your shift. How much risk would
each pose to your personal safety? Please rate on the scale of 1 (Low risk) to 10 (High risk).

SCALE ON RISKINESS
1

2

3

4

Low
Risk

5

6

7

8

9

10

High
Risk

Some Risk Some Risk

Your use of
firearms















 



Suspect’s use of
firearms















 



Usage of a stun gun















 



Arresting a suspect
surrounded by
bystanders















 



Domestic violence
situations with
weapons















 



Domestic violence
situations without
weapons















 



Traffic stops















 



Vehicle accidents















 



Situations involving
hazardous/chemical
substances















 



Investigating a
methamphetamine
laboratory















 



Being assaulted by
suspects















 



Not being backed
up by partners















 



45

SECTION 2
For this section, please estimate the frequency that you have encountered or participated in each
item during the past six months.
General Risks
FREQUENCY OF ENCOUNTER/PARTICIPATION IN THE PAST
SIX MONTHS
Never

Rarely

Sometimes

Often

AIDS









Vaccines









Medical X-rays









Climate change









Natural pollution









Water pollution









Bacteria in food









Eating fatty food









Drinking sugarsweetened
beverages









Cigarette smoking









Drinking alcohol









Stress









46

Occupation-Specific Risks
FREQUENCY OF ENCOUNTER/PARTICIPATION IN THE
PAST SIX MONTHS
Never

Rarely

Sometimes

Often

Your use of firearms









Suspect’s use of
firearms









Use of a stun gun









Arresting a suspect
surrounded by
bystanders









Domestic violence
situations with weapons









Domestic violence
situations without
weapons









Traffic stops









Vehicle accidents









Situations involving
hazardous/chemical
substances









Investigating a
methamphetamine
laboratory









Being assaulted by
suspects









Not being backed up by
partners









47

SECTION 3
For this section, please rate your response to each statement.
Strongly
Disagree

Somewhat
Disagree

Neutral

Somewhat
Agree

Strongly
Agree

I feel that I have very little
control over the risks to my
general well-being.











I rarely count on good things
happening to me.











Those in power often
withhold information about
things that are harmful to us.











It's no use worrying about
public affairs; I can't do
anything about them anyway.











I hardly ever expect things to
go my way.











Experts are able to make
accurate estimates of the risks
from lifestyle and habits.











Many of the problems facing
the world today are associated
with technology.











Differences of opinion about
general risks can be resolved
by scientific data and
analysis.











In uncertain times, I usually
expect the best.











We can trust the experts and
engineers who build, operate
and regulate new technology.











48

Strongly
Disagree

Somewhat
Disagree

Neutral

Somewhat
Agree

Strongly
Agree

When a risk is small, it is
acceptable for society to
impose that risk on
individuals without their
consent.











I enjoy my friends a lot.











Experts are able to make
accurate estimates of the risks
involved in policing.











What this world needs is a
more equal distribution of
wealth.











People have gone too far in
pushing equal rights











In a fair system people with
more ability should earn
more.











When there is a really serious
health problem, the public
health officials will take care
of it. Until they alert me
about a specific problem, I
really don't have to worry.











It's important for me to keep
busy.











Government has no right to
regulate people's personal
risk-taking activities











If people in this country were
treated equally, we would
have fewer problems.











49

Strongly
Disagree

Somewhat
Disagree

Neutral

Somewhat
Agree

Strongly
Agree

Overall, I expect more good
things to happen to me than
bad.











Scientists are able to make
accurate estimates of the risks
from new technology.











I often feel discriminated
against.











The police should have the
right to listen to private phone
calls to investigate a crime.











Our government and industry
can be trusted with making
the proper decisions to
manage the risks from law
enforcement.











It's easy for me to relax.











Technological development is
destroying nature.











Scientists are able to make
accurate estimates of the risks
from lifestyle and habits.











If something can go wrong for
me, it will.











I feel that I have very little
control over the risks to my
occupation.











We can trust the experts and
officials who create and draft
policies on the criminal
justice system.











50

Strongly
Disagree

Somewhat
Disagree

Neutral

Somewhat
Agree

Strongly
Agree

I'm always optimistic about
my future.











People in positions of
authority tend to abuse their
power.











Experts are able to make
accurate estimates of the risks
from new technology.











I don't get upset too easily.











I am in favor of capital
punishment.











Our government and industry
can be trusted with making
the proper decisions to
manage the risks from
technology.











SECTION 4
1) What is your age?

2) What is your sex?
 Male
 Female
3) What is your race or ethnicity?
 Hispanic
 White/Caucasian
 Black/African American
 Asian and Pacific Islander
 Native American
 Other (Please specify: __________________________________________________)
4) What is the highest level of education you have attended or completed?
 High school
 Some college but no degree
 2-year college degree
 4-year college degree
 Master’s degree
 Doctoral degree
 Other (Please specify: __________________________________________________)
5) What is your current rank in the department?

6) How long have you served as a sworn officer?

52

REFERENCES

53

REFERENCES

Austin, T. L., & Hummer, D. C. (1994). "Has a Decade Made a Difference?": Attitudes of Male
Criminal Justice Majors Toward Female Police Officers. Journal of Criminal Justice
Education , 5 (2), 229-239.
Baldassare, M., & Katz, C. (1992). The Personal Threat of Environmental Problems as Predictor
of Environmental Practices. Environment and Behavior , 24 (5), 602-616.
Barke, R. P., Jenkins-Smith, H., & Slovic, P. (1997). Risk Perceptions of Men and Women
Scientists. Social Science Quarterly , 78 (1), 167-176.
Bazley, T. D., Lersch, K. M., & Mieczkowski, T. (2007). Officer force versus suspect resistance:
A gendered analysis of patrol officers in an urban police department. Journal of Criminal
Justice , 35 (2), 183-192.
Brandl, S. G., & Stroshine, M. S. (2003). Toward an Understanding of the Physical Hazards of
Police Work. Police Quarterly , 6 (2), 172-191.
Brody, C. J. (1984). Differences by Sex in Support for Nuclear Power. Social Forces , 63 (1),
209-228.
Brown, J., & Heidensohn, F. (2000). Gender and Policing: Comparative Perspectives. New
York: St. Martin's Press.
Brown, R. A., & Frank, J. (2006). Race and Officer Decision Making: Examining Differences in
Arrest Outcomes between Black and White Officers. Justice Quarterly , 23 (1), 96-126.
Burke, R. J., & Mikkelsen, A. (2005). Gender Issues in Policing: Do They Matter? Women in
Management Review , 20 (2), 133-143.
Corrigan, R. S., Lester, D., & Loftus, T. (1980). Perception of Danger by Police Officers.
Perceptual and Motor Skills , 50 (1), 284.
Cutter, S. L., Tiefenbacher, J., & Solecki, W. D. (1992). En-gendered fears: femininity and
technological risk perception. Organization & Environment , 6 (1), 5-22.
Dake, K. (1992). Myths of Nature: Culture and the Social Construction of Risk. Journal of
Social Issues , 48 (4), 21-37.
Dake, K. (1991). Orienting Dispositions in the Perception of Risk: An Analysis of Contemporary
Worldviews and Cultural Biases. Journal of Cross-Cultural Psychology , 22 (1), 61-82.
DeJong, C. (2004). Gender Differences in Officer Attitude and Behavior: Providing Comfort to
Citizens. Women & Criminal Justice , 15 (3/4), 1-32.
54

DeJoy, D. M. (1989). The Optimism Bias and Traffic Accident Risk Perception. Accident
Analysis and Prevention , 21 (4), 333-340.
Delphy, C. (1993). Rethinking Sex and Gender. Women's Studies International Forum , 16 (1),
1-9.
Eiser, J. R., Miles, S., & Frewer, L. J. (2002). Trust, Perceived Risk, and Attitudes Toward Food
Technologies. 32 (11), 2423-2433.
Evans, J. R., & Mathur, A. (2005). The Value of Online Surveys. Internet Research , 15 (2), 195219.
Finucane, M. L., Slovic, P., Mertz, C. K., Flynn, J., & Satterfield, T. A. (2000). Gender, race,
and perceived risk: the 'white male' effect. Health, Risk & Society , 2 (2), 159-172.
Fischhoff, B., Watson, S., & Hope, C. (1984). Defining Risk. Policy Sciences , 17 (2), 123-139.
Flin, R., Pender, Z., Wujec, L., Grant, V., & Steward, E. (2007). Police Officers' Assessment of
Operational Situations. Policing , 30 (2), 310-323.
Flynn, J., Slovic, P., & Mertz, C. K. (1994). Gender, Race, and Perception of Environmental
Health Risks. Risk Analysis , 14 (6), 1101-1108.
Garcia, V. (2003). ''Difference'' in the Police Department : Women, Policing, and ''Doing
Gender''. Journal of Contemporary Criminal Justice , 19 (3), 330-344.
Gerber, G. L. (1996). Status in Same-Gender and Mixed-Gender Police Dyads: Effects on
Personality Attributions. Social Psychology Quarterly , 59 (4), 350-363.
Goldstein, H. (1963). Police Discretion: The Ideal versus the Real. Public Administration Review
, 23 (3), 140-148.
Greenberg, M. R., & Schneider, D. F. (1995). Gender Differences in Risk Perception: Effects
Differ in Stressed vs. Non-Stressed Environments. Risk Analysis , 15 (4), 503-511.
Greene, H. T., & del Carmen, A. (2002). Female Police Officers in Texas: Perceptions of
Colleagues and Stress. Policing: An International Journal of Police Strategies &
Management , 25 (2), 385-398.
Gustafson, P. E. (1998). Gender Differences in Risk Perception: Theoretical and Methodological
Perspectives. Risk Analysis , 18 (6), 805-811.
He, N., Zhao, J., & Archbold, C. A. (2002). Gender and Police Stress: The Convergent and
Divergent Impact of Work Environment, Work-Family Conflict, and Stress Coping
Mechanisms of Female and Male Police Officers. Policing , 25 (4), 687-708.
55

Heaton, R. (2010). We Could Be Criticized! Policing and Risk Aversion. Policing , 5 (1), 75-86.
Hoffman, P. B., & Hickey, E. R. (2005). Use of force by female police officers. Journal of
Criminal Justice , 33 (2), 145-151.
Hoffman, P. B., & Hickey, E. R. (2005). Use of Force by Female Police Officers. Journal of
Criminal Justice , 33 (2), 145-151.
Homant, R. J., Kennedy, D. B., & Hupp, R. T. (2000). Real and Perceived Danger in Police
Officer Assisted Suicide. Journal of Criminal Justice , 28 (1), 43-52.
Jermier, J. M., Gaines, J., & McIntosh, N. (1989). Reactions to Physically Dangerous Work: A
Conceptual and Empirical Analysis. Journal of Organizational Behavior , 10 (1), 15-33.
Kakar, S. (2002). Gender and Police Officers' Perceptions of Their Job Performance: An
Analysis of the Relationship between Gender and Perceptions of Job Performance.
Criminal Justice Policy Review , 13 (3), 238-256.
Kelman, I. (2003, Winter). Defining Risk. FloodRiskNet Newsletter (2), pp. 6-8.
Maguire, M., & Nolan, T. (2011). Faux hos: woman police attitudes about decoy sex work.
Police Practice and Research , 12 (3), 209–222.
Mayhew, C. (2001, February). Occupational Health and Safety Risks Faced by Police Officers.
Trends & Issues in Crime and Criminal Justice (196) , 1-6. Woden, Australia: Australian
Institute of Criminology.
Miller, S. L. (1999). Gender and Community Policing: Walking the Talk. Boston: Northeastern
University Press.
Morash, M., & Haarr, R. N. (1995). Gender, Workplace Problems, and Stress in Policing. Justice
Quarterly , 12 (1), 113-140.
Morash, M., Kwak, D.-H., & Haarr, R. (2006). Gender Differences in the Predictors of Police
Stress. Policing: An International Journal of Police Strategies & Management , 29 (3),
541-563.
Peters, E., & Slovic, P. (1996). The Role of Affect and Worldviews as Orienting Dispositions in
the Perception and Acceptance of Nuclear Power. Journal of Applied Social Psychology ,
26 (16), 1427-1453.
Poteyeva, M., & Sun, I. Y. (2009). Gender differences in police officers' attitudes: Assessing
current empirical evidence. Journal of Criminal Justice , 37 (5), 512-522.

56

Rabe-Hemp, C. E. (2008). Female officers and the ethic of care: Does officer gender impact
police behaviors? Journal of Criminal Justice , 36, 426-434.
Rabe-Hemp, C. E., & Schuck, A. M. (2007). Violence Against Police Officers: Are Female
Officers at Greater Risk? Police Quarterly , 10 (4), 411-428.
Riksheim, E. C., & Chermak, S. M. (1993). Causes of Police Behavior Revisited. Journal of
Criminal Justice , 21 (4), 353-382.
Rivers, L., Arvai, J., & Slovic, P. (2010). Beyond a Simple Case of Black and White: Searching
for the White Male Effect in the African-American Community. Risk Analysis , 30 (1),
65-77.
Rydberg, J., & Terrill, W. (2010). The Effect of Higher Education on Police Behavior. Police
Quarterly , 13 (1), 92-120.
Savage, I. (1993). Demographic Influences on Risk Perceptions. Risk Analysis , 13 (4), 413-420.
Sax, L. J., Gilmartin, S. K., & Bryant, A. N. (2003). Assessing Response Rates and Nonresponse
Bias in Web and Paper Surveys. Research in Higher Education , 44 (4), 409-432.
Schuck, A. M., & Rabe-Hemp, C. (2005). Women Police: The Use of Force by and Against
Female Officers. Women & Criminal Justice , 16 (4), 91-117.
Sherman, L. W. (1980). Causes of Police Behavior: The Current State of Quantitative Research.
Journal of Research in Crime and Delinquency , 17 (1), 69-100.
Siegrist, M., & Cvetkovich, G. (2000). Perception of Hazards: The Role of Social Trust and
Knowledge. Risk Analysis , 20 (5), 713-719.
Sjoberg, L. (2003). The Different Dynamics of Personal and General Risk. Risk Management , 5
(3), 19-34.
Skolnick, J. H. (2010). A Sketch of the Policeman's Working Personality. In S. K. Rice, & M. D.
White (Eds.), Race, Ethnicity, and Policing: New and Essential Readings (pp. 15-31).
New York: New York University Press.
Slovic, P. (1987). Perception of Risk. Science , 236 (4799), 280-285.
Slovic, P. (1999). Trust, Emotion, Sex, Politics, and Science: Surveying the Risk-Assessment
Battlefield. Risk Analysis , 19 (4), 689-701.
Slovic, P., Malmfors, T., Mertz, C., Neil, N., & Purchase, I. F. (1997). Evaluating chemical risks:
results of a survey of the British Toxicology Society. Human & Experimental Toxicology
, 16 (6), 289-304.
Stoller, R. J. (1974). Sex and Gender. New York: J. Aronson.
57

Sun, I. Y. (2007). Policing domestic violence: Does officer gender matter? Journal of Criminal
Justice , 35 (6), 581-595.
Uchida, C. D., Brooks, L. W., & Kopers, C. S. (1987). Danger to Police During Domestic
Encounters: Assaults on Baltimore County Police, 1984-86. Criminal Justice Policy
Review , 2 (4), 357-371.
Unger, R. K. (1979). Toward a Redefinition of Sex and Gender. American Psychologist , 34 (11),
1085-1094.
Van der Velde, F. W., Hooykaas, C., & Van der Pligt, J. (1992). Risk Perception and Behavior:
Pessimism, Realism, and Optimism about AIDS-Related Health Behavior. Psychology
and Health , 6 (1-2), 23-38.
Van Liere, K. D., & Dunlap, R. E. (1980). The Social Bases of Environmental Concern: A
Review of Hypotheses, Explanations and Empirical Evidence. Public Opinion Quarterly ,
44 (2), 181-197.
Vega, M., & Silverman, I. J. (1982). Female Police Officers as Viewed by Their Male
Counterparts. Police Studies , 5 (1), 31-39.
Vicente, P., & Reis, E. (2010). Using Questionnaire Design to Fight Nonresponse Bias in Web
Surveys. Social Science Computer Review , 28 (2), 251-167.
Walker, S. (1993). Taming the System: The Control of Discretion in Criminal Justice, 19501990. New York: Oxford University Press.
Weinstein, N. D. (1987). Unrealistic Optimism About Susceptibility to Health Problems:
Conclusions from a Community-Wide Sample. Journal of Behavioral Medicine , 10 (5),
481-500.
Wells, S. K., & Alt, B. L. (2005). Police Women: Life with the Badge. Connecticut : Praeger .
Wilson, J. Q. (1968). Varieties of Police Behavior: The Management of Law and Order in Eight
Communities. Cambridge: Harvard University Press.

58