BEYOND THE ORGANIZATIONAL BASES OF ETHICAL WORK CLIMATES:
A NEW THEORY AND MEASURE
Anke U. Arnaud
University of Central Florida
PO Box 161400
Orlando, FL 32816-1400
aarnaud@bus.ucf.edu
Marshall Schminke*
University of Central Florida
PO Box 161400
Orlando, FL 32816-1400
407-380-5819
mschminke@bus.ucf.edu
*Corresponding author
1
BEYOND THE ORGANIZATIONAL BASES OF ETHICAL WORK CLIMATES:
A NEW THEORY AND MEASURE
ABSTRACT
This research develops a new theory and measure of ethical work climate. It builds on Victor
and Cullen’s (1988) seminal work by extending the domain of ethical climate beyond moral
judgment, to include dimensions of moral sensitivity, moral motivation, and moral character.
Data from 652 employees across 101 organizations provided support for our four component
theory of ethical climate and for the reliability and validity of a new measure based upon it.
Results indicate that the components of ethical climate are distinct and relate to outcomes such as
ethical behavior, helping behavior, politics, interpersonal deviance, ethical program followthrough, and performance.
2
Recent and frequent reports of ethical transgressions such as embezzlement, insider trading, and
corporate fraud have rekindled researchers’ and practitioners’ interest in ethical work climate
(EWC), the organizational context in which ethical and unethical activities occur. The majority
of this research rests upon Victor and Cullen’s (1988) seminal ASQ paper “The organizational
bases of ethical work climates,” in which the authors developed and tested a theory and measure
of EWC. This framework has become the dominant foundation for ethical climate research,
providing the basis for nearly 75 percent of all studies of ethical climate.
However, the Victor and Cullen (1988) framework has been the target of a number of serious
critiques on both theoretical and empirical fronts (e.g., Wimbush, Shepard, and Markham, 1997a;
Vaicys, Barnett, and Brown, 1996; Arnaud and Schminke, 2006). In this study, we seek to
address several of these limitations by developing a new theory and measure of EWC. In
particular, we present a broader theoretical model of EWC, one that encompasses not only the
moral judgment component that provides the foundation for Victor and Cullen’s framework, but
includes other components of the ethical decision process as well: moral sensitivity, moral
motivation, and moral character. In addition, we develop, validate, and test a new measure of
EWC based on this theoretical model. By doing so, we seek to provide theoretical,
methodological, and practical contributions to research in the area through a comprehensive
theoretical framework and a more reliable and valid tool for assessing it.
CONCEPTUALIZING ETHICAL WORK CLIMATES
A work climate is defined as the shared perceptions of the content and strength of the prevalent
values, norms, attitudes, and behaviors of the members of a social system. These social systems
3
represent collectives of individuals who share a formal or informal structure such as a
department, organization, or network (Payne, 1990). Hence, we define EWC as the shared
perceptions regarding prevalent ethics-related values, norms, attitudes, and behaviors among
members of a social system.
Victor and Cullen’s Theory of Ethical Work Climates
In Victor and Cullen’s (1987, 1988) theory of EWC, ethical climates reflect employees’
collective perceptions of ethical events, ethical practices, and ethical procedures. These
collective perceptions are identified along two dimensions: the ethical criteria used for
organizational decision-making, and the loci of analysis used as a referent in the ethical decisionmaking process (see Figure 1).
--Insert Figure 1 about here-The ethical criteria dimension identifies the collective decision-making framework employed in
making moral judgments involving the social system. It is grounded in Kohlberg’s (1984) theory
of cognitive moral development. Kohlberg suggested that moral development is characterized
by three levels of moral reasoning: preconventional, conventional, and postconventional. At the
preconventional level, individuals define morality in terms of the personal consequences of their
actions. At the conventional level, they define morality in terms of peer influences, family, and
societal norms. At the postconventional level, individuals define morality in terms of personal
conscience and universalistic principles. Victor and Cullen (1988) argue that these three levels
of moral reasoning will be reflected in three distinct ethical decision criteria employed by
organization members. They term these egoism, benevolence, and principled criteria, which
reflect preconventional, conventional, and postconventional reasoning, respectively.
4
The loci of analysis dimension identifies the referent group to be considered when applying
ethical criteria to a moral problem involving the social system. It borrows from role theory
(Merton, 1957), which notes that individuals’ behaviors and attitudes are differentially
influenced according to reference groups (e.g., locals or cosmopolitans) and the roles individuals
assume within those groups. Victor and Cullen (1988) extend this traditional distinction between
locals and cosmopolitans to include the individual as well. They propose three possible loci of
analysis employed by organization members: individual, local, and cosmopolitan.
Crossing the three ethical criteria with the three loci of analysis results in nine theoretically
distinct climate types, illustrated in Figure 1. Although this framework has come to represent the
dominant paradigm in EWC research, in recent years it has been the target of a number of
critiques, involving both theoretical and methodological limitations.
Theoretical limitations. Numerous authors have offered theoretical critiques of Victor and
Cullen’s (1988) framework (e.g., Vaicys, Barnett, and Brown, 1996; Wimbush, Shepard, and
Markham, 1997a, b; Barnett and Schubert, 2002; Arnaud and Schminke, 2006). A variety of
issues have been raised, but we identify two as especially important. The first is the question of
whether the model’s two dimensions—ethical criteria and loci of analysis—represent distinct
and independent aspects of an ethical climate. The second is whether the model is
comprehensive enough to capture the true breadth of the ethical climate construct.
5
With regard to the first issue, we suggest that ethical criteria and loci of analysis do not represent
distinct dimensions of ethical climate. Victor and Cullen draw explicitly on Kohlberg’s (1984)
stages of moral development in identifying their ethical criteria of egoism, benevolence, and
principle. However, the loci of analysis dimension, with its focus on individual, local, and
cosmopolitan considerations, rests on a similar foundation. That is, it also reflects three levels of
reasoning reflective of Kohlberg’s levels of moral development: self interest, communal interest,
and a broader, more universalistic interest. Victor and Cullen explicitly acknowledge this
overlap in a footnote that states, “the loci of analysis are somewhat similar to Kohlberg’s (1984)
classification of individual moral reasoning” (1988: 106).
The theoretical challenges presented by such an overlap are considerable. For example, the
ethical criteria dimension of egoism is defined as an employee’s desire to maximize his or her
self-interest. Similarly, the loci of analysis dimension of an individual is defined as using the self
as the referent in ethical decision-making. Therefore, these two components are not independent,
in that employees who desire to maximize their self-interest naturally see themselves as the
referent in moral reasoning. Similar arguments hold across each of the three levels of ethical
criteria and loci of analysis. Thus we suggest that the Victor and Cullen (1988) framework
might be more accurately represented by a single dimension, one that captures the prevailing
model of moral judgment in a social system, based on Kohlberg’s (1984) three levels of moral
development.
6
The second significant theoretical challenge faced by the Victor and Cullen (1988) framework is
that of underspecification. That is, the question arises as to whether the model is comprehensive
enough to capture the true breadth of the ethical climate construct. We argue that it is not.
Victor and Cullen’s (1988) model assesses the prevailing norms of moral judgment—such as
those reflected in Kohlberg’s (1984) three stages of cognitive moral development—at work in an
organizational setting. However, research indicates that moral judgment represents only a single
component of a multifaceted process of ethical decision-making and behavior. For example,
Rest’s (1984, 1986) four component model suggests that four distinct processes must happen in
order for ethical behavior to occur: moral sensitivity, moral judgment, moral motivation, and
moral character. Thus, Victor and Cullen’s model appears to be underspecified with respect to
the other three components of ethical behavior. We suggest that to understand the impact of
ethical climate on ethical behavior, a more comprehensive model of ethical climate is needed,
one reflecting each of the four components.
Methodological limitations. These theoretical concerns provide the foundation for some of the
principle methodological and empirical challenges that have emerged in operationalizing Victor
and Cullen’s (1988) model. The Ethical Climate Questionnaire (ECQ), the instrument by which
Victor and Cullen’s framework is assessed (Victor and Cullen, 1987, 1988; Cullen, Victor, and
Bronson, 1993), has been plagued by at least three significant methodological concerns: an
unstable factor structure, the emergence of different climate types, and a modest ability to predict
ethical outcomes.
7
First, the overall factor structure of the instrument has been unstable. Researchers utilizing the
ECQ have discovered that different numbers of climate types emerge across studies. For
example, Victor and Cullen (1988) reported that only five of the nine theoretical climate types
emerged in their sample. Other researchers have reported factor structures indicating three
(Wimbush, Shepard, and Markham, 1997a), six (Vaicys, Barnett, and Brown, 1996), seven
(Cullen, Victor, and Bronson, 1993), eight (Fritzsche, 2000), or nine (Peterson, 2002b) emergent
climate types.
Second, in addition to different numbers of climates, different types of ethical climates have
emerged across studies as well. Some of these have differed only slightly from the original
theoretical and emergent climate types identified by Victor and Cullen (1988). However, others
differ in more substantial ways. For example, Elm and Nichols (1993) identified a broad climate
type they termed principled, which encompassed three of Victor and Cullen’s theoretical and
emergent climate types. Similarly, Wimbush, Shepard, and Markham (1997a) identified an
ethical service climate, and Treviño, Butterfield, and McCabe (1998) identified ethical climates
of employee focus and community focus. None of these map directly onto either Victor and
Cullen’s theoretical climate types, nor those that emerged from their 1988 study or their
subsequent work (e.g., Cullen, Victor, and Bronson, 1993). The reason for this instability may
lie in the fact that individual ECQ items have often not consistently clustered together according
to their underlying theoretical factors. For example, Schminke, Ambrose, and Neubaum (2005)
reviewed 31 studies utilizing the ECQ, and identified only 16 of 36 items that have consistently
loaded as expected across studies.
8
Finally, the research involving the ECQ has demonstrated only a modest ability to predict ethical
behavior. For example, Wimbush, Shepard, and Markham (1997a) examined the relationship
between ethical climate and four specific unethical acts and found an overall correlation of 0.12.
In particular, the mean correlations between climate and acts of stealing, lying, disobeying
company rules, and serving as an accomplice in unethical acts were 0.14, 0.12, 0.11, and 0.10,
respectively. Similarly, Peterson (2002a) found an average correlation between nine climate
types and nine unethical behaviors (exaggerating benefits of product or service, falsely calling in
sick, stealing, giving gifts for preferential treatment, divulging confidential information, lying,
falsifying reports, not reporting others’ violations, and padding expense account) of 0.22. In a
recent meta-analysis, Martin and Cullen (2006) report a mean correlation between ethical climate
and a broad category of unethical conduct and dysfunctional behavior of 0.15, an effect size they
identify as small.
Research indicates that the ECQ may be somewhat better at predicting unethical behavior under
two conditions: 1) when an aggregate measure of unethical activity is employed, and 2) when an
organization does not employ a code of ethics. For example, when Peterson’s (2002a) nine
unethical behaviors are combined into a single aggregate measure of ethical behavior, the mean
correlation between ethical climate and unethical behavior is 0.22 for organizations with codes of
ethics, and 0.35 for organizations without ethical codes. Similarly, Treviño, Butterfield, and
McCabe (1998) employed an aggregate measure of unethical behavior and found similar mean
correlations of 0.23 for organizations with codes, and 0.34 for organizations without. Because
most organizations have come to adopt formal ethical codes of some form (estimates range from
approximately 70 percent of firms overall, and up to 95 percent of Fortune 50 firms (Greengard,
9
1997; Weaver, Treviño, and Cochran, 1999)), it appears that the correlations of about 0.22-0.23
represent the most appropriate comparison. Thus, the ability of the ECQ to predict unethical
behavior is modest.
These results are not surprising. Rather, they are in line with studies at the individual level that
explore the relationship between individual moral judgments and individual ethical behavior.
Correlations at the individual level are similarly modest, commonly averaging around 0.20 (Rest
and Narvaez, 1994). Rest (1986) argues that because moral judgment represents but a single
component of the ethical decision-making process, its ability to predict ethical behavior on its
own should be limited.
In all, the Victor and Cullen (1988) framework has provided the foundation for an active and
growing stream of research on EWCs. However, if we are to continue making progress in our
understanding of ethical climate and its influence on ethical behavior in organizations, we see a
need to overcome these theoretical and methodological limitations. As a result, we propose a
new theory of EWC and develop a new measure designed to assess it.
A NEW THEORY OF ETHICAL WORK CLIMATES
Victor and Cullen’s (1988) model of EWC captures an important component of the ethical
decision-making process: moral judgment. However, for ethical behavior to occur, moral
judgments are but one of several necessary conditions. To identify the others, we turn to Rest’s
(1984, 1986) four component model of ethical decision-making. Rest’s model represents the
preeminent framework conceptualizing ethical decision-making in the literature today and has
10
served — either explicitly or implicitly — as the foundation for most contemporary models of
ethical decision-making and behavior. For example, Jones notes its attractiveness as a
foundation for models of organizational ethical decision-making due to its parsimony as well as
the fact that it “contains all of the key elements of moral decision-making and behavior” (1991:
379).
Rest’s Four Component Model
Rest (1984, 1986) argues that for individuals to engage in ethical acts, they must engage in four
basic psychological processes: moral sensitivity, moral judgment, moral motivation, and moral
character. Moral sensitivity involves recognizing that an ethical dilemma exists and evaluating
how one’s actions affect others. Moral judgment involves bringing one’s moral decision-making
framework (such as that reflected in Kohlberg’s (1984) concept of cognitive moral development)
to bear on the problem, to determine the ethical course of action. Moral motivation concerns the
degree to which ethical values dominate other potential values (e.g., power values or economic
values) in a particular situation. Finally, moral character relates to whether an individual
possesses the personal fortitude to follow-through on what he or she has determined to be the
correct ethical course of action.
A New Conceptualization of Ethical Work Climate
We argue here that each of these four components of the ethical decision-making process may be
present at the social system-level as well, reflected in four distinct facets of EWC. Of course, a
fundamental challenge for our theory is the question of whether it is appropriate to raise these
individual-level dimensions of ethical decision-making to the collective level of the social
11
system. Although ample precedent exists for doing so within the EWC literature (e.g., Victor
and Cullen’s (1988) framework does so with respect to moral judgment), research in multilevel
analysis offers additional support.
In distinguishing between types of multilevel phenomena, Kozlowski and Klein (2000)
distinguish between top-down and bottom-up processes. Our concern here is with bottom-up
processes, described by Kozlowski and Klein as those in which “phenomena in organizations
have their theoretical foundation in the cognition, affect, behavior, and characteristics of
individuals, which—through social interaction, exchange, and amplification—have emergent
properties that manifest at higher levels. In other words, many collective constructs represent the
aggregate influence of individuals” (2000: 15).
Kozlowski and Klein describe two types of bottom-up processes: composition and compilation.
Composition processes are those in which the lower-level construct is fundamentally isomorphic
with the higher-order construct. That is, the essence of the lower-level construct flows through
to the higher level of analysis relatively unchanged. Compilation processes, on the other hand,
may share a common domain or some common features, but are not identical. In assembling the
lower-level pieces, Kozlowski and Klein suggest, the higher-order construct becomes
“distinctively different” (2000: 16).
Kozlowski and Klein cite the organizational climate construct as an exemplar of a composition
model, noting that the lower-level (individual) and higher-order (organizational) factors
“reference the same construct, have the same meaning, and share the same nomological
12
network” (2000: 17). Citing examples of an organization’s climate for service, they observe that
organizational climate emerges from shared perceptions of individual organizational members
regarding service.
The same logic applies to ethical climates, which also represent the shared, aggregate
perceptions of employees with respect to the content and strength of the prevalent values, norms,
attitudes, and behaviors of the members of a social system. The same psychological components
that comprise individual ethical decision-making and subsequent action will also apply at the
social system-level. Following Rest (1986), we label these four components of ethical climates
collective moral sensitivity, collective moral judgment, collective moral motivation, and
collective moral character.
Consistent with Rest (1984, 1986), the collective moral sensitivity component of EWC involves
two subcomponents, moral awareness and empathetic concern. Moral awareness reflects the
prevalent mode within the social system of envisioning what alternative actions are possible.
Empathetic concern reflects the prevalent mode of evaluating the consequences of those actions
in terms of how they affect others and who would be affected by them. Thus, collective moral
sensitivity involves the prevalent norms that exist in a social system for both moral awareness
and empathetic concern.
The collective moral judgment component of EWC reflects the prevalent form of moral
reasoning utilized within the social system. More specifically, collective moral judgment is
defined as the norms of moral reasoning used to judge which course of action is morally right.
13
We have argued that Victor and Cullen’s (1988) theoretical framework might be more accurately
represented by a single dimension, one that captures this prevailing model of moral judgment in
a social system, based on Kohlberg’s (1984) three levels of moral development. Thus, modified
to a single dimension, we suggest that Victor and Cullen’s existing framework captures this
component well.
The collective moral motivation component of EWC involves assessing whether ethical concerns
dominate other concerns when determining actions. More specifically, collective moral
motivation involves the prevalent values of the social system, and whether moral values such as
honesty, fairness, or helping are generally prioritized over other values such as power, control, or
personal achievement.
Finally, the collective moral character component of EWC involves the norms for implementing
a planned course of action. Rest (1986) notes that this dimension is also characterized by
multiple subfactors such as norms of self-control and norms for assuming responsibility. He
identifies individuals with high levels of moral character, those able to follow through on
decisions regarding the correct course of ethical action, as exhibiting these traits. Self-control
reflects the ability to control one’s actions in committing to and implementing a planned course
of action. Assuming responsibility means that individuals accept responsibility for the welfare of
others, live up to moral commitments, and follow either personal or societal rules and dictates.
In all, we suggest that these four components comprise the foundation for the ethical climate of
organizations.
14
The Four Component Model and Ethics-Related Outcomes
One element of our theory of EWC is that the ethical climates of moral sensitivity, judgment,
motivation, and character will each relate to more positive ethics-related outcomes. We
characterize ethics-related outcomes as encompassing a broad constellation of activities with
potential ethical impact. We define such activities to include overtly unethical behavior like
theft or deceit, but also other negative activities like interpersonal deviance, political deviance,
and workplace aggression. We suggest that ethics-related outcomes include positive activities as
well, such as organizational citizenship behaviors, helping behaviors, and volunteering.
Evidence from the literature informs our thinking about how each component relates to ethicsrelated outcomes. Because three of the four components have not previously been
conceptualized and tested at the social system-level (Victor and Cullen’s (1988)
conceptualization of climate as collective moral judgment is the exception), the preponderance of
this evidence results from considering previous research on individuals, and how the components
of ethical decision-making relate to individual ethics-related outcomes. We discuss evidence
related to each component in turn.
Collective Moral Sensitivity. Recall that collective moral sensitivity is composed of two factors,
norms of moral awareness and norms of empathetic concern. Little empirical evidence exists
that demonstrates the effect of moral awareness on ethics-related behaviors. However, Cohen,
Pant, and Sharp (2001) found that students with higher levels of moral awareness had stronger
intention to behave ethically, leading us to expect that increased moral awareness would lead to
15
more ethical behavior. More research has linked norms of empathetic concern to ethics-related
behaviors. For example, empathetic individuals display more helping behaviors (Davis, 1983)
and more organizational citizenship behaviors such as altruism (Kidder, 2002). Conversely,
lower empathy has been linked to increased aggression (Hastings, Zahn-Waxler, and Robinson,
2002) and delinquent behavior (Hackenberg-Culotta, 2002). This leads us to expect that higher
levels of empathetic concern should lead to more ethical outcomes as well.
Collective Moral Judgment. Unlike the other three components, the literature provides some
evidence of the relationship between moral judgment and ethics-related outcomes at the social
system-level. We have argued that Victor and Cullen’s (1988) conceptualization of EWC
represents collective moral judgment. Thus, empirical work demonstrating a link between Victor
and Cullen’s conceptualization of EWC and unethical behavior is relevant here. As we have
noted, this relationship is modest. However, several studies report generally positive
relationships between ethical climates based on higher-level reasoning (e.g., rules and laws) and
ethical behavior, and generally negative relationships between climates based on lower-level
collective moral reasoning (e.g., self-interest and instrumental) and ethical behavior (Wimbush,
Shepard, and Markham, 1997a; Treviño, Butterfield, and McCabe, 1998; Peterson, 2002a;
Martin and Cullen, 2006).
Collective Moral Motivation. Evidence links individuals’ moral motivation to more ethical
behaviors. For example, Franc, Sakic, and Ivicic (2002) found that individuals identifying
ethics-related values like benevolence and universalism as guiding principles in their lives
engaged in more socially desirable behaviors, such as helping, than did individuals who
16
identified other values as guiding principles. Similarly, Gaerling (1999) found individual values
of universalism to be related to prosocial behavior. In addition, positive relationships have been
identified between benevolence and moral behavior, universalism and moral behavior, and
values that serve the collective interest and moral behavior (Schwartz and Bilsky, 1987;
Schwartz, 1992). These findings suggest that collective moral motivation characterized by
prevalent moral values such as benevolence and universalism should be positively linked to
ethics-related outcomes as well.
Collective Moral Character. Recall that collective moral character is also characterized by
subfactors, such as norms of self-control and norms for assuming responsibility. At the
individual level, some support exists for linking self-control to ethics-related outcomes. For
example, Gottfredson and Hirschi (1990) demonstrated low self-control was associated with
higher propensity to commit crimes, and Krebs (1969) noted individuals with stronger abilities to
self-regulate their actions cheated less. More evidence exists for the relationship between
assumption of responsibility and ethical outcomes. Schwartz (1973) found individuals with
stronger personal norms of accepting responsibility volunteered at higher rates, whereas
individuals high in responsibility denial tended to ignore standard norms and rationalized their
behavior by blaming depersonalized others, such as organizations. In addition, Schwartz and
David (1976) and Zuckerman, Siegelbaum, and Williams (1977) found individuals who had high
personal norms for assuming responsibility were more likely to engage in prosocial behaviors
such as helping others and volunteering. In all, these findings suggest that collective moral
character should be positively linked to ethics-related outcomes.
17
Overall, we theorize that ethical work climates will reflect each of the four components of the
ethical decision process: moral sensitivity (moral awareness and empathetic concern), moral
judgment, moral motivation, and moral character (self-control and assuming responsibility).
Further, we predict that each of these components will exert an impact on both positive and
negative ethics-related outcomes.
METHOD
To test our model, we developed an instrument—the Ethical Climate Index—that assesses the
four components of EWC in organizational settings. We did so in three steps. First, we
developed a preliminary version of the Ethical Climate Index. Second, we refined the instrument
to tighten its factor structure, increase its reliability, and enhance its parsimony. Finally, we
validated the instrument and tested its ability to predict ethics-related outcomes consistent with
our theory.
Step 1: Preliminary Scale Development
The first step of our study involved establishing a preliminary version of the Ethical Climate
Index. Where possible, we did so by modifying existing individual-level scales to the social
system (in this case, department) level. Where existing measures were not available, we created
new items to asses the ethical climate component.
Collective moral sensitivity includes two factors: norms of empathic concern (role taking) and
norms of moral awareness. Ten items for assessing empathy and role taking in the social system
were drawn from the empathetic concern and perspective taking dimensions of Davis’s (1980)
18
Interpersonal Reactivity Index. These items were modified to reflect the social system-level of
analysis. Psychometric properties of this scale have been satisfactory at the individual level,
with internal reliability (coefficient alpha) ranging from .72 to .78 (Davis, 1980, 1983).
Participants indicated on a five-point Likert-type scale the degree to which each item accurately
described their department (1=Describes my department very well, 5=Does not describe my
department at all). (Sample item: People in my department sympathize with someone who is
having difficulties in their job.)
We could identify no existing instrument based on Likert-type scales to assess moral awareness
at the individual or collective level.1 Therefore, we developed nine new items to assess norms of
moral awareness at the department level. Participants indicated on a five-point scale the degree
to which each item accurately described their department (1=Describes my department very well,
5=Does not describe my department at all). (Sample item: People around here are aware of
ethical issues.)
Collective moral judgment items were based on Victor and Cullen’s measure of ethical climate,
the ECQ. Following Schminke, Ambrose, and Neubaum (2005), we utilized 16 items from the
1993 version of the ECQ (Cullen, Victor, and Bronson, 1993) that have exhibited consistent
factor loadings. We used these items to assess the collective level of moral reasoning in the
department. Internal reliability coefficients for the ECQ have been satisfactory, ranging from .76
to .85 (Arnaud, Ambrose, and Schminke, 2002). Participants indicated on a five-point scale the
degree to which each item accurately described their department (1=Describes my department
1
Scenario-based measures exist, but were not consistent with the format of the remainder of our climate index.
19
very well, 5=Does not describe my department at all). (Sample item: People are very concerned
about what is best for themselves.)
Collective moral motivation items were adapted from Schwartz’s (1992) 57-item Value Survey.
This instrument groups values into ten types: power, achievement, hedonism, stimulation, selfdirection, universalism, benevolence, tradition, conformity, and security. The values of
universalism (which reflects understanding, tolerance and protection for the welfare of all people
and for nature), and benevolence (which reflects protecting the welfare of close others in
everyday interaction) are generally considered moral values and empirical evidence has linked
these value types to ethical and prosocial behavior (Gaerling, 1999; Myyry and Helkama, 2001;
Franc, Sakic, and Ivicic, 2002). Internal reliabilities of the individual-level scale range from .61
to .78 (Schwartz, 1992). Participants indicated on a five-point scale the extent to which each
value represents a guiding principle in the department (1=Opposed to my department’s values,
5=Consistent with my department’s values). (Sample values: social power, wealth, social
justice, honesty.)
Collective moral character includes norms of responsibility and norms of self-control (Rest,
1986). We adapted fourteen items from Schwartz’s Denial of Responsibility scale (Schwartz,
1973; Schwartz and David, 1976) to assess respondents’ perceptions of norms of responsibility
in their department. Internal reliabilities for the scale have ranged from .78 to .83 (Schwartz,
1973, Schwartz and David, 1976; Harrington, 1996). Participants indicated on a five-point scale
the degree to which each item accurately described their department (1=Describes my
20
department very well, 5=Does not describe my department at all). (Sample item: No matter how
much people around here are provoked, they are always responsible for whatever they do.)
Four items from the Core-Self Evaluations Scale (Judge, et al., 2003) were adapted to assess the
respondents’ perceptions of departmental standards for self-control. Internal reliabilities for the
scale have averaged .84 across various samples (Judge et al., 2003). Participants indicated on a
five-point scale the degree to which each item accurately described their department
(1=Describes my department very well, 5=Does not describe my department at all). (Sample
item: Generally people in my department feel in control over the outcomes when making
decisions that concern ethical issues.)
Before administering this initial version of the Ethical Climate Index to participants, we
employed a panel of 5 trained raters to conduct a sorting exercise. These raters were trained in
Rest’s (1984, 1986) Four-Component model of individual ethical decision-making and our new
theory of EWC based upon it. Raters received descriptions of each of the ethical climate
components. (The collective moral motivation scale was excluded from this sorting exercise
because items included single words (values) only.) Each rater received a randomized list of all
initial items of the Ethical Climate Index, and was asked to sort them according to the ethical
climate components. Items were retained only if correctly sorted by a majority of the raters. The
resulting Ethical Climate Index included 14 items for collective moral sensitivity (8 moral
awareness and 6 empathetic concern), 16 items for collective moral judgment, 11 items for
collective moral character (4 self control and 7 responsibility), and the 57 items for collective
moral motivation.
21
Preliminary scale assessment. The resulting scales were then administered to 240 individuals,
including part-time MBA students from a large state university as well as entrepreneurs and
organizational employees in the Southeastern United States. The 158 individuals who responded
represented a response rate of 72 percent. The sample was 58 percent male, with a mean age of
29 years, and mean organizational tenure of 3.2 years.
Exploratory factor analyses (maximum likelihood extraction with oblique rotation) and reliability
analysis of the initial scales revealed reliable scales for two of the four components of ethical
climate. The fourteen collective moral sensitivity items factored into two interpretable subscales,
which we labeled norms of moral awareness (8 items, α = .88) and norms of empathetic concern
(3 items, α = .81). The sixteen collective moral judgment items factored into two interpretable
subscales, which we labeled focus on self (5 items, α = .91) and focus on others (5 items, α =
.89). Factor analytic results for the other two components were less clear. The 57 collective
moral motivation items did not produce interpretable factors consistent with Schwartz’s (1992)
dimensions. Likewise, the 11 collective moral character items did not factor as anticipated.
Only two of the norms of responsibility items loaded together, and the norms of self control
items did not yield an interpretable factor.
Step 2: Scale Refinement
Building on these results, we developed a revised version of the Ethical Climate Index. Based on
inter-item correlations and factor loadings, we deleted three items from the collective moral
sensitivity scale, retaining a more parsimonious five items (α = .87). In addition, we crafted four
22
new items to supplement the three existing norms of empathetic concern items. For collective
moral judgment, we were satisfied with the results of the initial testing, and therefore retained the
resulting ten item scale (five items for focus on self, five for focus on others).
For collective moral motivation, we recrafted our approach by limiting our focus to the value
dimensions of the Schwartz (1992) model that are related to ethics (benevolence and
universalism) and those that represent their opposites (power and achievement). We developed
18 items in which participants indicate on a five-point scale the degree to which each item
accurately describes their department (1=Describes my department very well, 5=Does not
describe my department at all). (Sample items: In my department social fairness is considered
more important than social power. In my department people strive to obtain power and control
even if it means compromising ethical values.)
For collective moral character we refocused our approach as well. Because our self control
subcomponent items did not yield an interpretable factor, we explored other individual
characteristics as reflective of moral character. The literature reveals that both self-efficacy and
locus of control have been directly linked to moral character at the individual-level (Rest, 1984,
1986). Thus, we developed 13 new items based on Bandura’s (1989) individual-level self
efficacy scale and Rotter’s (1966) individual-level locus of control scale. (Sample items:
Generally, people in my department feel in control over the outcomes when making decisions
that concern ethical issues. When necessary, people in my department take charge and do what
is morally right.) Only two items of our responsibility subcomponent loaded together.
Therefore, we adapted 9 additional items from the original Responsibility Denial Scale
23
(Schwartz, 1973). (Sample item: People in my department feel it is better to assume
responsibility for a mistake.) Participants indicate on a five-point scale the degree to which each
item accurately describes their department (1=Describes my department very well, 5=Does not
describe my department at all).
Revised scale assessment. These revised scales (excluding the moral judgment scale, which
was acceptably reliable and parsimonious after the first test) were then administered to 270 parttime MBA students from a large state university. The 261 returned surveys represented a
response rate of 96 percent. The sample was 53 percent male, had a mean age of 32 years, and
an average tenure of 4.5 years with their organization.
Exploratory factor analyses (maximum likelihood extraction with oblique rotation) and reliability
analysis revealed clear factor structures and reliable scales for each component of the revised
instrument. The 12 collective moral sensitivity items factored into two subscales as expected,
norms of moral awareness (5 items, α = .82) and norms of empathetic concern (7 items, α = .88).
The 18 collective moral motivation items yielded three interpretable factors: ethical values
dominant (universalism/benevolence over achievement/power) (3 items, α = .71); ethical values
subservient (achievement/power over universalism/benevolence) (8 items, α = .89), and benefit
for others versus self (6 items, α = .75). This third factor did not reflect Rest’s (1986)
conceptualization of moral motivation and thus was dropped. The first and second factors did
reflect Rest’s concept of moral motivation. However, the first factor had only three items and
was less reliable than the second, so it was dropped in favor of the second factor. Therefore, we
retained the ethical values subservient factor as our indicator of ethical values, reverse scoring it
24
to represent collective moral motivation. The 24 collective moral character items resulted in two
interpretable factors: norms of individual responsibility (6 items, α = .84) and norms of
comparative responsibility (3 items, α = .72). We determined that comparative responsibility
only indirectly reflects collective moral character as conceptualized by Rest. (Sample item: If
someone got away with stealing in my department, others would be more likely to steal, too.)
Thus, we determined that collective moral character was best reflected as a unitary construct,
assessed by norms of individual responsibility.
In all, the revised Ethical Climate Index contains 12 moral sensitivity items (5 norms of moral
awareness, 7 norms of empathetic concern), 10 moral judgment items (5 focus on self, 5 focus on
others), eight collective moral motivation items, and six collective moral character items. The
final instrument is provided in the appendix.
Step 3: Scale Validity and Theory Testing
In step three, we validated the Ethical Climate Index using Nunnally and Bernstein’s (1994)
approach to construct validation, and tested our predictions that the ethical climate components
1) will exist at the department level and 2) will be related to ethics-related outcomes.
Data collection and sample. With the assistance of a team of alumni and students, we identified
organizational departments of at least six members (including the supervisor) willing to
participate in a study of organizational climate. Members of this team served as contact persons
for each participating department. We received agreement from 113 departments across 101
organizations, which included both product- and service-oriented firms as well as for-profit and
25
not-for-profit organizations. The 652 individuals who responded to the survey represented a
response rate of 98 percent. Respondents’ mean age was 30 years and 51 percent of the sample
was male. Respondents averaged 4.1 years of tenure with their organizations and 3.1 years of
tenure with their department.
The contact person for each department received a package containing surveys to be completed
by five department members and the department supervisor, as well as specific, written
instructions regarding the completion of the surveys and procedures for returning them to the
researchers. (Participants had the option of returning the survey to the contact person or directly
to the researchers via mail.)
Survey measures. All surveys—both employee and supervisor—included demographic
questions (age, sex, education, organization and department tenure) and our Ethical Climate
Index. All Ethical Climate Index scales were subsequently coded such that higher values
reflected higher degrees of collective moral sensitivity, judgment, motivation, and character.
Additionally, all surveys included an abbreviated version of the Crowne and Marlowe (1960)
Social Desirability Scale (9 items; Ballard, 1992).
Employee surveys. Employee surveys also included multiple instruments to assess the validity
(both convergent and discriminant) of our instrument, as well as its ability to predict outcomes
consistent with our theory. To assess convergent validity, we included instruments to assess four
constructs we expected to relate to ethical climate. These included an abbreviated measure (4
items; Ehrhart, 2004) of procedural justice climate (Colquitt, 2001), a measure of general justice
26
climate (6 items; Ambrose and Schminke, 2006), and two alternative measures of climate that
reflect concern for the well-being of fellow employees, safety action climate and safety
expectation climate (5 items each; Zohar, 2000). To assess discriminant validity, we included
instruments to asses three constructs that we did not expect to be as strongly related to ethical
climate. These included measures of perceived functional dependence (the degree to which an
employee is directly dependent upon the work of other employees) (3 items; Morris and Steers,
1980), problem-solving demand (the degree of cognitive processing required to prevent or
recover errors in a job) (5 items; Wall, Jackson, and Mullarkey, 1995), and organizational
structure (organic versus mechanistic) (7 items; Khandwalla, 1977). To assess the ability of our
instrument to predict outcomes, we included a measure of an ethics-related outcome, general
political behavior (a 2 item subscale of Kacmar and Carlson’s (1997) Perception of Politics
Scale). (Four additional measures of ethics-related outcomes were provided by supervisors, as
described below.) Not all employees received all of these instruments. These scales were
randomly distributed among departments such that each scale was rated by at least 240
individuals across 40 departments.
Supervisor surveys. Surveys completed by supervisors included four additional measures of
ethics-related outcomes. These included measures of ethical behavior (10 items; Akaah, 1992),
interpersonal deviance (9 items; Bennett and Robinson, 2000), organizational citizenship helping
behavior (5 items; Moorman and Blakely, 1992), and ethics program follow-through, the extent
to which the department follows up on reports of ethical problems and its responses to ethical
lapses with appropriate discipline (4 items; Treviño and Weaver, 2001). For exploratory
purposes, we also included a measure of perceived department performance (7 items; Delaney
27
and Huselid, 1996). (Recent research supports the use of perceptual measures of performance
(Wall, et al., 2004). Such a measure is especially useful in our sample, which is comprised of
over 100 different departments, few of which perform similar work in similar contexts, making
objective performance comparisons difficult.) All supervisors received all of these instruments.
RESULTS
Aggregation analysis. We first assessed whether EWCs suggested by our theory exist at the
social system-level. That is, we evaluated whether shared perceptions of ethical climate types
exist at the department level. If so, aggregating individual climate perceptions to reflect
collective climate is appropriate. We evaluated rwg statistics to determine the degree of interrater
agreement with respect to the ethical climate components (George, 1990; George and James,
1993). Mean rwg values of .70 or greater provide evidence of acceptable agreement among
member responses on a scale (George, 1990; Janz, Colquitt, and Noe, 1997). We found mean rwg
values across the six ethical climate subscales of .73, suggesting that aggregation was
appropriate. In addition, rwg values for all other aggregated measures met the .70 standard as
well (procedural justice climate, .70; general justice climate, .80; safety action climate, .78;
safety expectation climate, .70; organizational structure, .76; climate for initiative, .74).
Confirmatory factor analysis of the Ethical Climate Index. Our theory of EWC suggests that
EWC is composed of four components, and our initial scale development indicated that two of
these (collective moral sensitivity and collective moral judgment) contain two distinct
subcomponents. Thus, we conducted a confirmatory factor analysis to assess the accuracy of our
proposed six-factor model of EWC. Results suggest that the proposed six-factor model provides
28
a good fit for the data (X2 = 2115.76, df = 579; RMSEA=.07, CFI = .97, NFI = .96)(Hu and
Bentler, 1999; Rakov and Marcoulides, 2000). In addition, it provides a better fit than an
alternative four factor model that assumes no subcomponents (X2 = 3076.84, df = 588;
RMSEA=.10, CFI = .95, NFI = .95; X2 change = 961.08, df = 6, p<.01) or a one factor model
that assumes a single, general ethical climate (X2 = 6279.72, df = 594; RMSEA=.19, CFI = .88,
NFI = .87; X2 change = 4163.96, df = 15, p<.01).
Validity of the Ethical Climate Index. Convergent validity is the degree to which concepts that
should be related theoretically are interrelated in reality. Discriminant validity is the degree to
which concepts that should not be related theoretically are, in fact, not interrelated in reality.
Following Campbell and Fiske (1959), we assessed convergent and discriminant validity by
comparing the correlations of the Ethical Climate Index scales to measures of other constructs.
For these comparisons, we identified four constructs that we anticipated would be more strongly
related to ethical climate, in that they relate to fairness and concern for others (procedural justice
climate, general justice climate, safety action climate, and safety expectation climate). We also
identified three constructs with less obvious links to ethical issues (perceived functional
dependence, problem-solving demand, and organizational structure), which we anticipated would
be less strongly related to ethical climate. Thus, we expect higher correlations between our
ethical climate scales and the convergent validity constructs, and lower correlations between our
ethical climate scales and the discriminant validity constructs.
The mean correlation between the six ethical climate scales and our convergent validity scales of
procedural justice climate, general justice climate, safety action, and safety expectation were
29
strong, at .58, .33, .54, and .39, respectively. As expected, the mean correlation between the
ethical climate scales and our discriminant validity measures of perceived functional
dependence, problem-solving demand, and organizational structure were weaker, at .18, .09, and
.17, respectively. Thus, the scales of the Ethical Climate Index demonstrate good convergent
and discriminant validity.
Theory testing: The Ethical Climate Index and ethics-related outcomes. Our theory of
EWC suggests that each ethical climate component will be related to ethics-related outcomes,
including ethical behavior, interpersonal deviance, general political behavior, organizational
citizenship behavior, and ethics program follow-through. In addition, we performed an
exploratory analysis of the relationship between ethical climate and department performance.
We tested the predicted relationships using multiple regression analyses. The six ethical climate
components served as predictor variables for each of the six outcomes of interest. Because
ethics-related phenomena are subject to social desirability concerns (Chung and Monroe, 2003;
Bellizzi and Bristol, 2005), we controlled for social desirability effects in all models. (Results
are unchanged when this control is excluded). In addition, VIF statistics were below 10 for all
variables in all models, indicating that multicollinearity was not problematic (Ryan, 1997).
Because our primary interest here is in identifying the effects of the six climate types on our
array of ethics-related outcomes, we present results by climate type. We then briefly summarize
results of our exploratory analysis of the impact of ethical climates on department performance.
Descriptive statistics and correlations between variables appear in Table 1. Results of regression
analyses appear in Table 2.
30
--Insert Table 1 and Table 2 about here-Collective moral sensitivity. The first subcomponent of collective moral sensitivity, moral
awareness, was significantly related to all five ethics-related outcomes. As anticipated, it exerted
a positive effect on ethical behavior and ethics program follow-through, and was associated with
lower levels of deviance. Contrary to our expectations, it was related to higher levels of political
behavior and lower levels of helping. Standardized regression coefficients indicate its strongest
impact was on interpersonal deviance.
The second subcomponent of collective moral sensitivity, empathetic concern, was related to
three of the five ethics-related outcomes. As expected, it was related to lower levels of political
behavior. Contrary to expectations, it was associated with lower levels of ethical behavior and
ethics program follow-through. Regression coefficients indicate its strongest impact was on
decreasing political behavior.
Collective moral judgment. The first subcomponent of collective moral judgment, focus on
others, was related to three of the five ethics-related outcomes. As expected, it was related to
lower levels of political behavior. Unanticipated results include negative relationships with
helping and ethics program follow-through. Its strongest impact was on decreasing political
behavior.
The second subcomponent of collective moral judgment, focus on self, represents a dimension of
ethical climate that differs from the others. Consider the zero-order correlations reported in
Table 1. Although all of the other climate dimensions are positively correlated with one another,
31
each is negatively correlated with focus on self. This difference appears to be a reasonable one,
if we consider the meaning of a focus on self climate. Reflecting back on Kohlberg’s (1984)
theory of moral development on which the collective moral judgment component rests, the focus
on self subcomponent reflects moral judgment based on ethical egoism. That is, what is right is
what is best for me personally. The welfare or interests of others—either individuals or the
organization or those beyond the organization—are not in play when making ethical judgments.
Thus, we expect that such a climate would be at odds with a collective sense of an ethical
workplace that would precipitate ethical behavior, helping behavior, less political behavior, and
less interpersonal deviance.
Focus on self was related to two of the five ethics-related outcomes. As anticipated, it was
related to higher levels of political behavior. However, it was also unexpectedly related to
higher levels of helping behavior, which is where it exerted its strongest impact.
Collective moral motivation. Collective moral motivation was related to three of the five
ethics-related outcomes. All effects were as anticipated, as higher levels of collective moral
motivation were associated with higher levels of ethical behavior and helping, and lower levels
of deviance. Its strongest effect was on helping behavior.
Collective moral character. Collective moral character was related to four of the five ethics
related outcomes. Again, all effects were as anticipated, as higher levels of collective moral
character resulted in higher levels of ethical behavior, helping, and ethics program follow-
32
through, and lower levels of interpersonal deviance. Its strongest influence was on helping
behavior.
In all, each of the four components of ethical climate exerted significant effects on ethics-related
outcomes. For two of the climate types—collective moral motivation and collective moral
character—all of the effects were in the predicted direction. For the other two—collective moral
sensitivity and collective moral motivation—results were mixed. Each of these components
influenced at least one outcome in the anticipated direction, but each also revealed at least one
unanticipated significant relationship. We address these unanticipated results below.
Department performance. Although department performance does not represent an explicitly
ethics-related outcome, we included it in our analysis for exploratory purposes. Results indicate
that four of the climate components were related to perceptions of department performance.
Three of these (collective moral judgment-focus on self, collective moral motivation, and
collective moral character) were positively related to performance, whereas collective moral
sensitivity-norms of empathetic concern was negatively related to performance. The strongest
effect to emerge was for collective moral motivation, followed by collective moral sensitivitynorms of empathetic concern and collective moral character.
DISCUSSION
The goals of this paper were threefold. First, to present a new theory of EWC. Second, to
develop and validate an instrument by which it might be assessed. Finally, to test whether the
proposed climate types exist in organizational settings, and to determine whether they are
33
capable of predicting a variety of ethics-related outcomes. Overall, results were supportive of
these goals. We were able to develop a reliable and valid instrument that reflects our theorized
ethical climate components. Results indicate strong agreement among employees that such
climates exist, and that strong relationships exist between our climate components and an array
of ethics-related outcomes. These effects were not isolated among one or two climate
components or limited to one or two outcomes. Rather, each of the six climate dimensions
exerted significant effects on at least three of the six outcome variables, and each of the six
outcomes were affected by at least three different climate dimensions.
Anticipated and Unanticipated Findings
Most of the relationships that emerged in our study were as expected. That is, we anticipated
that more positive ethical climates would have a positive impact on ethics-related outcomes, and
in general this was the case. For example, higher levels of collective moral character and moral
motivation each had positive and significant effects on ethical behavior, helping behavior, and
ethics program follow-through, and lessened interpersonal deviance. For these two components
of ethical climate, the significant relationships between climate and outcomes were all in the
anticipated direction.
However, not all of the significant relationships emerged as expected for the other two climate
components, collective moral sensitivity and collective moral judgment. Consider the case of
norms of moral awareness, one of the subcomponents of collective moral sensitivity. As
expected, higher levels of moral awareness led to higher levels of ethical behavior, lower levels
of deviance, and greater ethics program follow-through. However, they also led to decreased
34
helping behavior, and increased political behavior. Similar anomalies emerged for the other
collective moral sensitivity subcomponent (empathetic concern) and for the two collective moral
judgment subcomponents (focus on others and focus on self).
To understand these counterintuitive results, it is useful to examine the univariate correlations
among the variables reported in Table 1. Having established above that the focus on self
subcomponent of moral judgment operates in a fundamentally different manner from the other
climate dimensions, we are in position to note an interesting point in Table 1. That is, unlike the
multivariate results in Table 2, all of the significant univariate relationships between the ethical
climate dimensions and ethics-related outcomes appear as anticipated. For example, the moral
awareness subcomponent of collective moral sensitivity is positively related to ethical behavior,
helping, ethics program follow-through, and performance, and negatively related to deviance and
political behavior. In other words, higher levels of moral awareness are associated with more
positive ethics-related outcomes. The same is true for each of the other ethical climate
dimensions. In every case in which the univariate relationships are significant, a positive
relationship exists between higher levels of climate and generally desirable outcomes (ethical
behavior, helping, follow-through, and performance), and a negative relationship exists between
higher levels of climate and undesirable outcomes (deviance and political behavior). (As
expected, because the focus on self subcomponent is negatively related to the other climate
dimensions, these relationships are reversed.)
This sheds a somewhat different light on the multivariate results reported in Table 2. It suggests
that the unanticipated relationships that emerge there are not due to unexpected patterns in the
35
data at the most basic level. Indeed, at the univariate level, climate and outcomes behave just as
we would anticipate. Rather, the unanticipated results appear only after controlling for all of the
other ethical climate dimensions. This suggests some relatively complex interactions may be
present between the climate dimensions.
It is possible to provide plausible explanations for some of the unanticipated relationships in
Table 2. For example, contrary to our expectations, norms of empathetic concern were inversely
related to ethical behavior. This suggests that employees who work in environments that
promote sympathy, understanding, and concern for others in the department are actually more
prone to engage in unethical behavior. One possible explanation for this counterintuitive finding
may be that in certain settings, higher levels of empathy may lead employees to identify more
events or interactions as having an unethical constitution. That is, if I continually place myself in
others’ shoes, I might be more likely to recognize potentially unethical actions imposed upon
them. Consistent with its role as a component of moral sensitivity, therefore, higher levels of
empathy might be related to higher recognition and reporting of unethical behavior, rather than
to an actual increase in its objective level of occurrence.
However, we have already noted that at the univariate level, the relationship does not unfold that
way; the correlation between norms of empathetic concern and ethical behavior is positive and
significant (0.22, p<.01). Absent other information about other climate levels, more empathy is
associated with more ethical behavior. Thus, the real explanation must be more complex than a
simple empathy-leads-to-recognition argument. The same is true for crafting explanations for
any of the counterintuitive relationships revealed in Table 2. Taken together, the results reported
36
in Table 1 and Table 2 suggest that although ethical climate and outcomes might behave as
expected in simple settings, more complex relationships emerge in more complex settings in
which multiple climate dimensions are in operation at the same time.
The standardized regression coefficients in Table 2 offer some evidence that such relationships
may play a somewhat marginal role, however. These coefficients reflect the relative impact of
the various ethical climate dimensions on an outcome. It is notable that for each outcome, the
climate variable exerting the strongest effect does so in the anticipated direction. Additionally,
in most (but not all) cases, the difference in relative effect sizes between the anticipated effects
and the unanticipated effects is fairly large. This suggests that at least the primary impacts of
ethical climate on outcomes are much as we would anticipate.
However, the lesser, unexpected effects are present, and they suggest issues worthy of further
attention. Additional work—both theoretical and empirical—will be required to create and test
more fine-grained models that might be capable of clarifying the meaning of these unexpected
results. For example, individual-level research suggests that constructs such as moral sensitivity
and moral reasoning may interact (Raboteg-Saric, 1997). Therefore, we suggest that future
research investigate possible interaction effects among the various ethical climate dimensions.
Further, we know that organizational context, broadly defined, represents an important
moderator in organizational research (Pfeffer and Davis-Blake, 1990). Organizational climate in
general, and ethical climate in particular, comprise part of that organizational context. Thus, we
suggest subsequent work might benefit from embracing ethical climate as a potential moderating
influence on other organizational relationships.
37
Relative Impact of Climate Dimensions
The six dimensions of ethical climate did not exert equally strong effects across all outcomes.
Collective moral sensitivity-norms of moral awareness and collective moral character played a
significant role in five outcomes, collective moral sensitivity-norms of empathetic concern and
collective moral motivation in four, and collective moral judgment-focus on others and collective
moral judgment-focus on self in just three. Further, not only do the two collective moral
judgment measures have an effect on only three of the outcomes, standardized regression
coefficients indicate that even when significant, the relative impact of these dimensions was
among the lowest of the six EWC factors, with a mean value of .19 versus a mean across the
other dimensions of .29. This suggests that the Victor and Cullen (1988) framework, which is
based on collective moral judgment, attempts to assess ethical climate from what may be one of
the weaker components of the ethical decision process.
Theoretical and Practical Implications
This research has both theoretical and practical implications. From a theoretical perspective, our
model of EWC opens new doors for researchers interested in understanding the impact of the
ethical context of organizations. Rest’s (1984, 1986) four component model of individual ethical
decision-making changed the way we think about how ethical and unethical actions emerge. In
doing so, it provided new and varied options for researchers wishing to understand the ethical
decision-making process. In particular, it allowed us to ask new questions about what sort of
failures might lead to unethical acts. For instance, it urged us to consider that unethical acts may
not simply be the result of unethical judgments. Rather, they may result from a lack of moral
38
sensitivity, moral motivation, or moral character. Each represents a very different problem, with
a very different solution. The same holds true for our conceptualization of EWC. The four
components of EWC suggest four distinct avenues for understanding the collective ethical
context in organizations. With this model and instrument researchers will have the opportunity
to assess each of them and how they relate to each other and to important outcomes.
This research has practical implications as well. In recognizing that more than one component of
ethical climate exists, and that each exerts unique effects on outcomes, it offers practitioners the
opportunity to diagnose their specific ethical context with an eye toward understanding which
components are strong and which are lacking. Understanding the differential effects of the EWC
dimensions on organizational outcomes is especially important in order to create effective ethics
training and development programs. For example, training efforts to enhance employee
sensitivity toward ethical issues may show limited effectiveness in bringing about more ethical
behavior in an organization with an already-strong climate of moral sensitivity but weak
collective moral motivation or character. In addition, our results indicate that the Ethical
Climate Index is effective in predicting a variety of ethics-related behaviors. However, our
exploratory analysis also suggests it may have the capacity to explain links between ethical
climate and performance. We are not aware of any study to date that has been able to
demonstrate a clear link between EWC and department or organizational performance. The
practical appeal of being able to do so is considerable.
Limitations
39
All studies have limitations and of course, this one is no exception. First, all data were collected
by survey. Although we utilized two separate sources (employees and supervisors) for our
predictor and outcome variables, and respondents represented a wide array of organizational and
demographic backgrounds, common method variance still exists as a potential concern. Second,
our outcome variables do not constitute an exhaustive list of ethics-related outcomes in
organizations. Although we utilized multiple outcome measures, future research could explore
additional outcomes such as shrinkage rate records, sexual harassment complaints, and actual
ethical violation reports.
A third limitation is that results were based on self-reports. Evidence generally supports the
validity of self-reports (Spector, 1992), and Ones, Viswesvaran, and Schmindt’s (1993) metaanalysis of integrity measures suggests that self-report measures may provide benefits in this
type of research, in that they tend to result in higher estimates of validity than external measures.
Ones, Viswesvaran, and Schmindt explain that many deviant behaviors go undetected, which
may therefore limit the validity of external measures. They note there is substantial evidence
that the correlation between self-report admission of deviant behavior and actual behavior is
substantial. In this study, we assured participants anonymity and did not ask them to disclose
unethical conduct they may have personally committed, but rather to share their perceptions of
the environment in which they work. However, we still need to be alert to the fact that selfreports may be vulnerable to self-enhancement biases.
A final limitation is that we assessed only direct or main effects of the influence of climate on
outcomes. Inasmuch as our work represents a first step in the development of a new theory and
40
measure of EWC, we believe it provides an important first step in understanding the links
between ethical climate and outcomes. However, we anticipate that subsequent research will be
able to address the more complex relationships that underlie these effects.
CONCLUSION
The central goal of this research was to provide a new theory and measure of EWC, one capable
of speaking to the weaknesses that have emerged in Victor and Cullen’s (1988) seminal offering.
Theoretically, our model of EWC addresses a more comprehensive range of the ethical climate
of organizations than that of Victor and Cullen. As such, we suggest it provides a more sound
theoretical footing for increasing our understanding of the ethical context of organizations.
Methodologically, our measure of EWC has been subjected to a more rigid construction,
development, and validation protocol than previous EWC instruments. Empirically, the result is
a measurement tool that has the capacity to provide researchers with a more precise
understanding of the ethical climate of organizations, and represents a more powerful tool for
predicting ethics-related outcomes. Above, we noted previous research suggesting correlations
of between climate and ethical actions ranging from 0.14 (Wimbush, Shepard, and Markham,
1997a) to 0.23 (Treviño, Butterfield, and McCabe, 1998) for organizations with codes of ethics
(which represents the vast majority of organizations). In our study, mean correlations between
ethical climate and measures of ethical behavior, deviance, political behavior, and helping
behavior were .30, .48, .39, and .25, respectively, all considerably higher than those utilizing
other measures. Although our results present some new questions, in all we believe this research
represents an important step in improving our understanding of ethical work climates.
41
REFERENCES
Akaah, I. P.
1992 “Social inclusion as a marketing ethics correlate.” Journal of Business Ethics, 11: 599608.
Ambrose, M. L., and M. Schminke
2006 “An examination of overall justice.” Paper presented at the Society of Industrial and
Organizational Psychologists Meeting, Dallas, TX.
Arnaud, A., and M. Schminke
In press “Ethical work climate: A weather report and forecast.” In S.W. Gilliland, D.D. Steiner,
and D.P. Skarlicki (eds.), Research in Social Issues in Management: Managing Social and
Ethical Issues in Organizations, 5: Greenwich, CT: IAP.
Arnaud, A., M. L. Ambrose, and M. Schminke
2002 “Individual moral development and ethical climate: The influence of P-O fit on job
attitudes.” Paper presented at the Academy of Management Meeting, Denver, CO.
Ballard, R.
1992 “Short forms of the Marlowe-Crowne Social Desirability Scale.” Psychological Reports,
71: 155-1160.
Bandura, A.
1989 The Multidimensional Self-Efficacy Scales. Unpublished Test, Stanford University
Stanford, CA.
Barnett, T., and E. Schubert
2002 “Perceptions of the ethical work climates and covenantal relationships.” Journal of
Business Ethics, 36: 279-290.
Bellizzi, J.A., and T. Bristol
2005 “Supervising the unethical selling behavior of top sales performers: Assessing the impact
of social desirability bias.” Journal of Business Ethics, 57: 377-388.
Bennett, R.J., and S.L. Robinson
2000 “Development of a measure of workplace deviance.” Journal of Applied Psychology, 85:
349-360.
Campbell, D., and D. Fiske
1959 “Convergent and discriminant validation by the multitrait-multimethod matrix.”
Psychological Bulletin, 56: 81-105.
Chung, J., and G.S. Monroe
2003 “Exploring social desirability bias.” Journal of Business Ethics, 44: 291-302.
42
Cohen, J. R., L.W. Pant, and D.J. Sharp
2001 “An examination of differences in ethical decision-making between Canadian business
students and accounting professionals.” Journal of Business Ethics, 30: 319-336.
Colquitt, J.A.
2001 “On the dimensionality of organizational justice: A construct validation of a measure.”
Journal of Applied Psychology, 86: 386-400.
Crowne, D.P., and D. Marlowe
1960 “A new scale of social desirability independent of psychopathology.” Journal
of Counseling Psychology, 24: 349-354.
Cullen, J. B., B. Victor, and J. W. Bronson
1993 “The Ethical Climate Questionnaire: An assessment of its development and validity.”
Psychological Reports, 73: 667-674.
Davis, M. H.
1980 “A multidimensional approach to individual differences in empathy.” JSAS Catalog of
Selected Documents in Psychology, 10: 85.
1983 “Measuring individual differences in empathy: Evidence for a multidimensional
approach.” Journal of Personality and Social Psychology, 44: 113-126.
Delaney, J.T., and M.A. Huselid
1996 “The impact of human resource management practices on perceptions of organizational
performance.” Academy of Management Journal, 39: 949-969.
Ehrhart, M.G.
2004 “Leadership and procedural justice climate as antecedents of unit-level organizational
citizenship behavior.” Personnel Psychology, 57: 61-94.
Elm, D.R., and M.L. Nichols
1993 “An investigation of the moral reasoning of managers.” Journal of Business Ethics, 12:
817-833.
Franc, R., V. Sakic, and I. Ivicic
2002 “Values and value orientation of adolescents: Hierarchy and correlation with attitudes
and behaviours.” Rdustvena Istrazivanja, 11: 215-238.
Fritzsche, D. J.
2000 “Ethical climates and the ethical dimension of decision-making.” Journal of Business
Ethics, 24: 125-140.
Gaerling, T.
1999 “Value priorities, social value orientation and cooperation in social dilemmas.” British
Journal of Social Psychology, 38: 397-408.
43
George, J.M.
1990 “Personality, affect, and behavior in groups.” Journal of Applied Psychology, 75: 107116.
George, J.M., and I.R. James
1993 “Personality, affect, and behavior in groups revisited: Comment on aggregation, levels of
analysis, and a recent application of within and between analysis.” Journal of Applied
Psychology, 78: 798-804.
Gottfredson, M.R., and T. Hirschi
1990 General Theory of Crime. Palo Alto, CA: Stanford University Press.
Greengard, S.
1997 “50% of your employees are lying, cheating and stealing.” Workforce, October: 44-53.
Hackenberg-Culotta, S.L.
2002 “Empathy development and its relationship with aggression and delinquent behavior in
adolescents.” Dissertation Abstracts International: Section B: The Sciences and
Engineering, 63: 1551.
Harrington, S. J.
1996 “The effect of codes of ethics and personal denial of responsibility on computer abuse
judgments and intentions.” MIS Quarterly, September: 257-278.
Hastings, P.D., C. Zahn-Waxler, and J. Robinson
2002 “The development of concern for others in children with behavioral problems.”
Developmental Psychology, 36: 531-546.
Hu, L., and P. Bentler
1999 “Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria
versus new alternatives.” Structural Equation Modeling, 6: 1-56.
Janz, B.D., J.A. Colquitt, and R.A. Noe
1997 “Knowledge worker team effectiveness: The role of autonomy, interdependence, team
development, and contextual support variables.” Personnel Psychology, 50: 877-904.
Jones, T. M.
1991 “Ethical decision-making by individuals in organizations: An issue-contingent model.”
Academy of Management Review, 18: 366-395.
Judge, T. A., A. Erez, J.E. Bono, and C.J. Thoresen
2003 “The Core Self-Evaluation Scale: Development of a measure.” Personnel Psychology, 56:
303-331.
Kacmar, K.M., and D.S. Carlson
44
1997 “Further validation of the perceptions of politics scale (POPS): A multiple sample
investigation.” Journal of Management, 23: 627-658.
Khandwalla, P. N.
1977 The Design of Organizations. San Diego, C.A: Harcourt Brace Jovanovich.
Kidder, D. L.
2002 “The influence of gender on the performance of organizational citizenship behavior.”
Journal of Management, 28:629-648.
Kohlberg, L.
1984 The Psychology of Moral Development. San Francisco, CA: Harper and Row.
Kozlowski, S. W. J., and K. J. Klein.
2000 “A multilevel approach to theory and research in organizations: Contextual, temporal,
and emergent processes.” In K. J. Klein and S. W. J. Kozlowski (eds.), Multilevel
Theory, Research, and Methods in Organizations: Foundations, Extensions, and New
Directions: 3–90. San Francisco, CA: Jossey-Bass.
Krebs, R. L.
1969 “Teacher perceptions of children’s moral behavior.” Psychology in the Schools, 6: 394395.
Merton, R. K.
1957 Social Theory and Social Structure. New York: Free Press.
Martin, K. D., and J. B. Cullen
In press “Continuities and extensions of ethical climate theory: A meta-analytic review.” Journal
of Business Ethics.
Moorman, R. H., and G. L. Blakely
1992 “A preliminary report on a new measure of organizational citizenship behavior.” Paper
presented at the Annual Meeting of the Southern Management Association, Valdosta:
GA.
Morris, J.H., and R. M. Steers
1980 “Structural influences on organizational commitment.” Journal of Vocational Behavior,
17: 50-57.
Myyry, L., and K. Helkama
2001 “University students' value priorities and emotional empathy.” Educational Psychology, 1:
25-40.
Nunnally, J., and I. Bernstein
1994 Psychometric Theory, 2nd ed. New York, NJ: McGraw-Hill.
45
Ones, D.S., C. Viswesvaran, and F. L. Schmindt
1993 “Meta-analysis of integrity test validities. Findings and implications for personnel
selection and theories of job performance.” Journal of Applied Psychology, 78: 679-703.
Payne, R. L.
1990 “Method in our madness: A reply to Jackofsky and Slocum.” Journal of Organizational
Behavior, 11: 77-80.
Peterson, D. K.
2002a “Deviant workplace behavior and the organization's ethical climate.” Journal of Business
and Psychology, 17: 47-61.
2002b “The relationship between unethical behavior and the dimensions of the ethical climate
questionnaire.” Journal of Business Ethics, 41: 313-326.
Pfeffer, J., and A. Davis-Blake
1990 “Unions and job satisfaction.” Work and Occupations, 17: 259-283.
Raboteg-Saric, Z.
1997 “The role of empathy and moral reasoning in adolescents’ prosocial behavior.” Drustvena
Istrazivanja, 6: 493-512.
Rakov,T., and G.A. Marcoulides
2000 A First Course in Structural Equation Modeling. Mahwah, NJ: LEA.
Rest, J. R.
1984 The Major Components of Morality. New York: Wiley.
1986 Moral Development: Advances in Research and Theory. New York: Praeger.
Rest, J. R., and D. Narvaez
1994 Moral Development in the Professions: Psychology and Applied Ethics. Hillsdale, NJ:
Lawrence Erlbaum Associates.
Rotter, J. B.
1966 “Generalized expectancies for internal versus external control of reinforcement.”
Psychological Monographs, 80: No. 609.
Ryan, T. P.
1997 Modern Regression Methods. New York, NJ: John Wiley and Sons.
Schminke, M., M.L. Ambrose, and D.O. Neubaum
2005 “The effect of leader moral development on ethical climate and employee attitudes.”
Organizational Behavior and Human Decision Processes, 97: 135-151.
Schwartz, S. H.
46
1973 “Normative explanations of helping behavior: A critique, proposal and empirical test.”
Journal of Experimental Social Psychology, 9: 349-364.
1992 “Universals in the content and structure of values: Theory and empirical tests in 20
countries.” In M. Zanna (ed.), Advances in Experimental Social Psychology, 25: 1-65.
New York: Academic Press.
Schwartz, S.H., and A.B. David
1976 “Responsibility and helping in an emergency: Effects of blame, ability and denial of
responsibility.” Sociometry, 39: 406-415.
Schwartz, S. H., and W. Bilsky
1987 “Toward a universal psychological structure of human values.” Journal of Personality and
Social Psychology, 53: 550-562.
Spector, P.
1992 “A consideration of the validity and meaning of self-report measures of job conditions.”
In C.L. Cooper and I. T. Robertson (eds.), International Review of Industrial and
Organizational Psychology: 123-151. Chichester, England: Wiley.
Treviño, L.K., K.D. Butterfield, and D.L. McCabe
1998 “The ethical context in organizations: Influences on employee attitudes and behaviors.”
Business Ethics Quarterly, 8: 447-476.
Treviño, L.K., and G.R. Weaver
2001 “Organizational justice and ethics program “follow-through”: Influences on employees’
harmful and helpful behavior.” Business Ethics Quarterly, 11: 651-671.
Vaicys, C., T. Barnett, and G. Brown
1996 “An analysis of the factor structure of the ethical climate questionnaire.” Psychological
Reports, 79: 115-200.
Victor, B., and J.B. Cullen
1987 “A theory and measure of ethical climate in organizations.” Research in Corporate Social
Performance and Policy, 9: 51-71.
1988 “The organizational bases of ethical work climates.” Administrative Science Quarterly,
33: 101-125.
Wall, T.D., P.R. Jackson, and S. Mullarkey
1995 “Further evidence on some new measures of job control, cognitive demand and
production responsibility.” Journal of Organizational Behavior, 16:431-455.
Wall, T.D., J. Michie, M. Patterson, S.J. Wood, M. Sheehan, C.W. Clegg, and M. West
2004 “On the validity of subjective measures of company performance.” Personnel
Psychology, 57, 95-218.
Weaver, G. R., L.K. Treviño, and P.L. Cochran
47
1999 “Corporate ethics practices in the mid-1990’s: An empirical study of the Fortune 1000.”
Journal of Business Ethics, 18: 283-294.
Wimbush, J. C., J.M. Shepard, and S.E. Markham
1997a “An empirical examination of the multi-dimensionality of ethical climate in
organizations.” Journal of Business Ethics, 16: 67-77.
1997b “An empirical examination of the relationship between ethical climate and ethical
behavior from multiple levels of analysis.” Journal of Business Ethics, 16: 1705-1716.
Zuckerman, M., H. Siegelbaum, and R. Williams
1977 “Predicting helping behavior: Willingness and ascription of responsibility.” Journal of
Applied Social Psychology, 7: 295-299.
Zohar, D.
2000 “A group-level model of Climate for Safety: Testing the effect of group climate on
microaccident in manufacturing jobs.” Journal of Applied Psychology, 85: 120-136.
48
Figure 1: Theoretical Ethical Work Climates Types
Loci of Analysis
Ethical
Criteria
Individual
Local
Cosmopolitan
Egoism
Self-Interest
Company Profit
Efficiency
Benevolence
Friendship
Team Interest
Social Responsibility
Principle
Personal
Morality
Company Rules
Laws and
and Procedures
Professional Codes
49
Table 1: Descriptive statistics and correlationsa
Mean
St. d.
1
1. Age
30.37
10.94
--
2. Sex
.54
.71
-.03
--
3. Edu
3.34
1.10
.17**
.02
--
4. TO
49.42
68.04
.59**
-.04
.13**
--
5. TD
37.52
54.84
.51**
-.06
.01
.81**
--
6. MSA
3.52
.44
.11**
-.12**
.02
.08*
.10*
(.76)
7. MSC
3.45
.43
.05
-.09*
-.04
-.03
.04
.73**
(.79)
8. MJO
3.35
.46
-.01
-.06
-.12**
-.05
.00
.57**
.58**
(.82)
9. MJS
2.88
.61
-.10*
.06
.05
-.03
-.08*
-.40**
-.52**
-.28**
(.89)
10. MM
3.72
.67
.10*
-.11**
-.06
.06
.12**
.62**
.66**
.45**
-.77**
(.93)
11. MC
3.46
.47
.08*
-.06
-.05
-.03
.04
.62**
.65**
.64**
-.45**
.52**
(.87)
12. SD
1.64
.26
.25**
-.12**
.09*
.14**
.09*
.30**
.20**
.15**
-.20**
.28**
.26**
(.70)
13. EB
3.76
.78
.13**
-.07
.11**
.10*
.11**
.31**
.22**
.18**
-.35**
.40**
.34**
.21**
(.91)
14. ID
1.92
.75
-.18**
.09*
-.07
-.10*
-.12**
-.47**
-.38**
-.29**
.32**
-.41**
-.45**
-.26**
-.56**
15. GPB
2.43
.68
-.19**
.02
-.01
-.02
.06
-.42**
-.60**
-.50**
.39**
-.45**
-.49**
-.15*
.02
.17**
(.74)
16. OCB
3.75
.77
.08*
-.08*
.03
.02
.02
.21**
.22**
.19**
-.19**
.30**
.37**
.19**
.20**
-.24**
-.13*
(.88)
17. EPF
3.97
.77
.09*
-.07
.08*
.03
-.04
.15**
.04
-.02
-.03
.06
.17**
.11**
.06
-.06
.08
.39**
(.95)
18. PER
3.86
.68
.02
-.04
.07
.06
.03
.13**
.03
.09*
-.12**
.21**
.20**
.12**
.37**
-.12**
.08
.43**
.33**
a
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
(.89)
(.89)
Scale reliabilities on the diagonal
TO=Tenure in Organization, TD=Tenure in Department, MSA=Collective moral sensitivity: Norms of moral awareness, MSC=Collective moral sensitivity:
Norms of empathetic concern, MJO=Collective moral judgment: Focus on others, MJS= Collective moral judgment Focus on self, MM=Collective Moral
Motivation, MC=Collective Moral Character, SD=Social Desirability, EB=Ethical Behavior, ID=Interpersonal Deviance, GPB=General Political Behavior,
OCB=Organizational Citizenship Behavior Helping, EPF=Ethical Performance Follow-Through, PER= Performance
50
Table 2: OLS regression resultsa
Dependent variable:
Ethical
Behavior
Interpers.
Deviance
Political
Behavior
Collective moral sensitivity:
Norms of moral awareness
.16**
-.27***
.15*
Collective moral sensitivity:
Norms of empathetic concern
-.28***
.08
Collective moral judgment:
Focus on others
-.07
Collective moral judgment:
Focus on self
E. P. Follow
Through
Department
Performance
-.16**
.18**
.02
-.48***
-.07
-.18**
-.36***
.06
-.22**
-.10*
-.22***
-.04
-.08
-.03
.17*
.27***
.11
.18**
Collective Moral Motivation
.31***
-.18**
-.01
.44***
.06
.42***
Collective Moral Character
.25***
-.27***
-.03
.47***
.32***
.32***
.05
-.09*
.01
.08*
.05
.03
R2
.23***
.29***
.42***
.20***
.09***
.12***
F
26.06
36.02
36.09
22.67
8.28
12.50
N
628
628
628
628
622
622
Social Desirability
OCB
Helping
a
Standardized regression coefficients reported
*p<.05; **p<.01; ***p<.001
51
Appendix: The Ethical Climate Index
Collective moral sensitivity – Norms of moral awareness
1.
2.
3.
4.
5.
People around here are aware of ethical issues.
People in my department recognize a moral dilemma right away.
If a rule or law is broken, people around here are quick to notice.
People in my department are very sensitive to ethical problems.
People around here do not pay attention to ethical issues.
Collective moral sensitivity – Norms of empathetic concern
6. People in my department sympathize with someone who is having difficulties in their job.
7. For the most part, when people around here see that someone is treated unfairly, they feel pity for that
person.
8. People around here feel bad for someone who is being taken advantage of.
9. Sometimes people in my department do not feel very sorry for others who are having problems.
10. Others’ misfortunes do not usually disturb people in my department a great deal.
11. When people in my department see someone being treated unfairly, they sometimes don’t feel much
pity for them.
12. In my department people feel sorry for someone who is having problems.
Collective moral judgment – Focus on self
1.
2.
3.
4.
5.
People around here protect their own interest above other considerations.
People in my department are very concerned about what is best for them personally.
People around here are mostly out for themselves.
People in my department think of their own welfare first when faced with a difficult decision.
In my department people’s primary concern is their own personal benefit.
Collective moral judgment – Focus on others
6.
7.
8.
9.
10.
In my department it is expected that you will always do what is right for society.
People around here have a strong sense of responsibility to society and humanity.
What is best for everyone in the department is the major consideration.
The most important concern is the good of all the people in the department.
People in my department are actively concerned about their peers’ interests.
Collective moral motivation
1.
2.
3.
4.
5.
6.
In my department people are willing to break the rules in order to advance in the company.
Around here, power is more important than honesty.
In my department authority is considered more important than fairness.
Around here, achievement is valued more than commitment and loyalty.
In my department personal success is more important than helping others.
In my department people strive to obtain power and control even if it means compromising ethical
values.
7. Around here, people are willing to tell a lie if it means advancing in the company.
8. In order to control scarce resources, people in my department are willing to compromise their ethical
values somewhat.
52
Appendix: The Ethical Climate Index (cont’d)
Collective moral character
1. People around here are confident that they can do the right thing when faced with moral dilemmas.
2. People I work with would feel they had to help a peer even if that person were not a very helpful
person.
3. People in my department feel it is better to assume responsibility for a mistake.
4. No matter how much people around here are provoked, they are always responsible for whatever they
do.
5. Generally people in my department feel in control over the outcomes when making decisions that
concern ethical issues.
6. When necessary, people in my department take charge and do what is morally right.
Short form: Including only the three highest factor loadings from each dimension results in the following
18-item short form of the Ethical Climate Index (with resulting alphas and correlation with original
scale):
α
r
___
___
Collective moral sensitivity (moral awareness):
Collective moral sensitivity (empathetic concern):
Collective moral judgment (focus on self):
Collective moral judgment (focus on others):
Collective moral motivation:
Collective moral character:
Items 1, 2, 4
Items 6, 7, 8
Items 3, 4, 5
Items 7, 8, 9
Items 1, 2, 8
Items 2, 3, 4
53
.80
.83
.89
.81
.90
.82
.93
.85
.95
.96
.94
.93