(30-6950(";(;,$50=,9:0;?"(5,95(9+056(30-6950(";(;,$50=,9:0;?"(5,95(9+056
"$""*/63(9&692:"$""*/63(9&692:
3,*;9650*#/,:,: 961,*;:(5+0::,9;(;065: A*,6-9(+<(;,";<+0,:
% # '!%!% # '!%!
"! #%!"""! #%!""
(*8<,305,/90:;05,*655(<./?
(30-6950(";(;,$50=,9:0;?"(5,95(9+056
6336>;/0:(5+(++0;065(3>692:(;/;;7::*/63(9>692:30)*:<:),+<,;+
(9;6-;/,5+<:;90(3(5+9.(50@(;065(3 :?*/636.?64465:
!,*644,5+,+0;(;065!,*644,5+,+0;(;065
*655(<./?(*8<,305,/90:;05,% # '!%!"! #%
!""
3,*;9650*#/,:,: 961,*;:(5+0::,9;(;065:
/;;7::*/63(9>692:30)*:<:),+<,;+
#/0:#/,:0:0:)96<./;;6?6<-69-9,,(5+67,5(**,::)?;/,A*,6-9(+<(;,";<+0,:(;"$""*/63(9&692:;
/(:),,5(**,7;,+-6905*3<:065053,*;9650*#/,:,: 961,*;:(5+0::,9;(;065:)?(5(<;/690@,+(+4050:;9(;696-
"$""*/63(9&692:69469,05-694(;06573,(:,*65;(*;:*/63(9>692:*:<:),+<
DEVELOPMENT OF AN EMPLOYEE GREEN BEHAVIOR
DESCRIPTIVE NORMS SCALE
A Thesis
Presented to the
Faculty of
California State University,
San Bernardino
In Partial Fulfillment
of the Requirements for the Degree
Master of Science
in
Psychology:
Industrial-Organizational
by
Jacqueline Christine McConnaughy
June 2014
DEVELOPMENT OF AN EMPLOYEE GREEN BEHAVIOR
DESCRIPTIVE NORMS SCALE
A Thesis
Presented to the
Faculty of
California State University,
San Bernardino
by
Jacqueline Christine McConnaughy
June 2014
Approved by:
Dr. Mark D. Agars, Committee Chair, Psychology
Dr. Janelle Gilbert, Committee Member
Dr. Matt Riggs, Committee Member
© 2014 Jacqueline Christine McConnaughy
iii
ABSTRACT
With a growing interest in sustainability, organizations and researchers
have begun to examine pro-environmental behaviors in the workplace (i.e.
employee green behaviors). However, general understanding of employee green
behaviors is currently limited due to a lack of measurement tools. In this study, a
new scale was developed to measure employee green behavior descriptive
norms, which are a source of influence on employee green behaviors that
develops from observing others’ behaviors. Initial items and expected scale
structure for the Employee Green Behavior Descriptive Norms Scale were
developed based on the Green Five Taxonomy of employee green behaviors.
Items were refined through pilot test data and a retranslation task. Data on the
refined scale, the Ethical Leadership Questionnaire, and a Work-Family Culture
Scale were used to test scale structure and gather evidence of construct validity.
Study results supported the expected scale structure and construct validity of the
newly developed scale. A multi-item, validated scale contributes to organizational
assessment of employee green behavior descriptive norms and contributes to
the scientific literature on employee green behaviors.
iv
ACKNOWLEDGEMENTS
I would like to thank my thesis committee, Dr. Mark Agars, Dr. Janelle
Gilbert, and Dr. Matt Riggs, for their continued support throughout the thesis
process. Dr. Agars, your feedback and insight provided valuable guidance that
shaped my thesis, and our meetings kept me focused and moving forward. Dr.
Gilbert, my introduction is stronger for your suggestions. Dr. Riggs, our
discussion of the statistical analyses provided clarity and gave me confidence in
the results of this study. I appreciate the time and attention my thesis committee
provided to strengthen this study and help me complete this portion of my
Master’s degree.
I would also like to thank Alexander McKay for his helpful comments and
advice throughout the thesis process, and Dr. Donna Garcia for her feedback on
my thesis proposal. Finally, I would like to thank Associated Students
Incorporated and the Office of Student Research for their generous research
grants.
v
TABLE OF CONTENTS
ABSTRACT ................................................................................................. iii
ACKNOWLEDGEMENTS ............................................................................ iv
CHAPTER ONE: INTRODUCTION
Introduction ....................................................................................... 1
Types of Norms ................................................................................ 4
The Relationship Between Descriptive Norms and
Pro-Environmental Behaviors ........................................................... 8
The Focus Theory of Normative Conduct .............................. 8
The Outcomes of Normative Influence ........................ 10
The Interaction of Injunctive Norms and
Descriptive Norms ....................................................... 11
The Influence of Descriptive Norms is
Underdetected ............................................................. 14
How Do Descriptive (and Injunctive) Norms Work? ..... 16
Summary of Research on Theory of Normative
Conduct ....................................................................... 19
The Theory of Planned Behavior ............................................ 19
Summary of Research on the Theory of Normative Conduct
and the Theory of Planned Behavior ...................................... 21
Previous Measurement of Pro-Environmental Behavior
Descriptive Norms: In and Out of the Workplace .............................. 22
The Green Five Taxonomy of Employee Green Behaviors ............... 26
The Present Study ............................................................................ 31
CHAPTER TWO: DEVELOPMENT OF THE EMPLOYEE GREEN BEHAVIOR
DESCRIPTIVE NORMS SCALE
vi
Item Development ........................................................................... 32
Instrument Refinement .................................................................... 33
Pilot Test Procedure ............................................................... 33
Sample .................................................................................... 33
Subscale Reliability Analysis .................................................. 34
Item Retranslation Task ......................................................... 36
Final Revisions ....................................................................... 36
CHAPTER THREE: CONFIRMING THE FACTOR STRUCTURE AND
VALIDATING THE EMPLOYEE GREEN BEHAVIOR
DESCRIPTIVE NORMS SCALE
Construct Validity ............................................................................... 38
Ethical Leadership ................................................................... 39
Work-Family Culture ............................................................... 39
Method .............................................................................................. 40
Procedure ............................................................................... 40
Sample ................................................................................... 41
Survey Design ........................................................................ 43
Measures and Demographics ................................................ 43
Ethical Leadership ....................................................... 43
Work-Family Culture ..................................................... 44
Demographics ............................................................. 44
Confirmatory Factor Analysis Results ............................................... 45
Construct Validity Results ................................................................. 48
vii
Ethical Leadership ................................................................... 48
Work-Family Culture ............................................................... 49
CHAPTER FOUR: DISCUSSION
General Discussion ........................................................................... 50
Structural Evidence of the Employee Green Behavior
Descriptive Norms Scale ................................................................... 51
Evidence of Construct Validity .......................................................... 53
Implications ....................................................................................... 55
Future Research ............................................................................... 56
Limitations ......................................................................................... 58
Conclusion ........................................................................................ 61
APPENDIX A: INITIAL EMPLOYEE GREEN BEHAVIOR DESCRIPTIVE
NORMS SCALE .................................................................. 62
APPENDIX B: INITIAL EMPLOYEE GREEN BEHAVIOR DESCRIPTIVE
NORMS SCALE ITEMS ARRANGED BY
META-CATEGORY ............................................................. 65
APPENDIX C: RETRANSLATION TASK INSTRUCTIONS
AND CATEGORIES ............................................................ 68
APPENDIX D: REFINED EMPLOYEE GREEN BEHAVIOR DESCRIPTIVE
NORMS SCALE .................................................................. 72
APPENDIX E: ETHICAL LEADERSHIP QUESTIONNAIRE ........................ 75
APPENDIX F: WORK-FAMILY CULTURE SCALE ..................................... 77
APPENDIX G: FINAL, 27-ITEM EMPLOYEE GREEN BEHAVIOR
DESCRIPTIVE NORMS SCALE ......................................... 80
APPENDIX H: TABLES ............................................................................... 82
APPENDIX I: FIGURES .............................................................................. 89
viii
REFERENCES ............................................................................................ 93
1
CHAPTER ONE
INTRODUCTION
Introduction
Ones and Dilchert (2012a) state, “To be ecologically sustainable, we need
to promote, influence, and change employee behaviors such that they are
congruent with environmental sustainability goals of organizations” (p.112). They
call these environmentally-related employee behaviors employee green
behaviors (EGBs) and define them as “scalable actions and behaviors that
employees engage in that are linked with and contribute to or detract from
environmental sustainability” (Ones & Dilchert, 2012a, p. 87). These behaviors
can be performed as a requirement of the job or as optional organizational
citizenship behaviors. Sometimes, these behaviors can be counterproductive in
that they actually detract from the organization’s environmental performance,
rather than enhance it. As “scalable” actions, they can vary in terms of how
frequently or proficiently employees perform them, and this scalability allows
each employee’s contribution to be quantified.
To try and capture the range of EGBs present in the workplace, Ones and
Dilchert (2012a) developed a content-based, three-tier Green Five Taxonomy of
EGBs. The first tier consists of General Green Performance, whereas the second
tier is comprised of the five meta-categories of Working Sustainably, Avoiding
Harm, Conserving, Influencing Others, and Taking Initiative. Then the third tier
2
splits the five meta-categories into 16 categories. In addition to confirming the
structure of the Green Five Taxonomy on a new set of incidents from industries
within the United States, its generalizability was also supported through its
application to an international sample of incidents from European organizations
(Hill et al., 2011 as cited in Ones & Dilchert, 2012a). By identifying the content
domain of EGBs, the Green Five Taxonomy helps define the behavioral content
of future research on EGBs.
Because this area of research is so new, the definition and taxonomy are
currently the extent of psychology’s examination of EGBs. If this area of study is
going to grow, it will be critical to develop appropriate measurement tools. To
help forward this line of research, Ones and Dilchert (2012a) are currently in the
process of developing a measurement scale for EGBs using their Green Five
Taxonomy. However, there are other constructs whose examination could inform
our understanding of EGBs and that could benefit from improved measurement
scales as well.
Using the environmental psychology literature as a source of ideas, one
such construct is descriptive social norms. Social norms are a form of
communication among group members regarding whether behaviors are
appropriate, beneficial, and easy to perform (Bamberg & Möser, 2007). They can
be split into injunctive norms, which indicate what people should do, and
descriptive norms, which indicate what people actually do (Cialdini, Reno, &
Kallgren, 1990). Both types of norms have been found to influence pro-
3
environmental behaviors (PEBs; e.g. Cialdini et al., 1990; Reno, Cialdini, &
Kallgren, 1993), which are defined as “individual behaviors contributing to
environmental sustainability” (Mesmer-Magnus, Viswesvaran, & Wiernik, 2012, p.
160). However, research suggests that descriptive norms typically have a
stronger relationship with behavior than do injunctive norms (Manning, 2009;
Thøgersen, 2006). Though the reason for this finding is still unclear, there are
two possible explanations. First, injunctive norms are thought to require greater
cognitive processing before influencing behavior (Jacobson, Mortensen, &
Cialdini, 2011). Thus, descriptive norms are more influential because adherence
to descriptive norms requires less cognitive effort. Second, because descriptive
norms are heuristics for effective behavior, they are more likely to influence
behaviors in private settings (e.g. when individuals are alone in their office) than
would injunctive norms (Cialdini et al., 1990; Kallgren, Reno, & Cialdini, 2000).
Just as they influence PEBs, descriptive norms are also likely to have a
strong influence on EGBs. Descriptive norms influence behavior when behaviors
are performed in public (Lapinski & Rimal, 2005) and there are opportunities for
people to observe others and mimic their behaviors (Fornara, Carrus, Passafaro,
& Bonnes, 2011). This observational learning (Bandura, 1986) allows people to
pick up on effective behaviors and adapt to new and ambiguous environments
(Cialdini et al., 1990; Griskevicius, Goldstein, Mortensen, Cialdini, & Kenrick,
2006). As noted by Takeuchi, Yun, and Wong (2011), behaviors in the workplace
can be strongly influenced by social exchanges among co-workers. By working
4
and interacting with one’s supervisors and co-workers, employees within an
organization gain opportunities for such observational learning and for the
influence of descriptive norms to affect their behavior. Additionally, as stated by
Carrico and Riemer (2011), there is the potential for a stronger normative
influence at work because employees are a more “captive audience” than when
they are at home or out in public.
In order to examine the influence of descriptive norms on EGBs, a well-
developed and validated measure is needed that reflects the construct of EGB
descriptive norms. Currently, there are no measures for descriptive norms that
reflect the breadth of behaviors identified in the Green Five Taxonomy.
Additionally, an examination of the PEB literature provides a few examples of
scales that capture descriptive norms for PEBs, but they exhibit certain
psychometric limitations. In response, in this study I hope to add to the literature
by developing and validating an EGB Descriptive Norms Scale that reflects the
distinction between injunctive and descriptive norms, and captures the breadth of
EGBs found in the Green Five Taxonomy. In support of this scale, I will examine
(a) the distinct nature of descriptive norms, (b) the influence of descriptive norms
on PEBs, (c) the limitations of currently available descriptive norms scales for
PEBs, (d) the content of the Green Five Taxonomy, and (e) the preliminary
development of the nomological network for the EGB Descriptive Norms Scale.
Types of Norms
5
Norms are a shared understanding of what constitutes appropriate
behavior (Thøgersen, 2006). They shape and enforce behavior through the
perceived possibility of punishment for noncompliance and reward for adherence
(Schwartz & Howard, 1981). With the focus of this study on descriptive norms, it
is important to distinguish the descriptive norm construct from other normative
influences. Overall, norms can be distinguished by the source of enforcement
(social vs. personal norms), the source of behavioral influence (injunctive vs.
descriptive norms), and the source of the normative referent (subjective vs. local
norms).
Social norms and personal norms differ on their source of enforcement.
Social norms are externally enforced; they are shaped by the expectations of
others and are reinforced through perceptions of rewards and punishment
administered by these others (Cialdini & Trost, 1998; Schwartz & Howard, 1981).
In comparison, personal norms are self-created expectations about one’s own
behavior (Schwartz, 1977) that typically align with an individual’s internal values
and beliefs (Thøgersen, 2006). They are internally enforced in that the
associated rewards and punishments are self-administered (Schwartz, 1977).
Because the focus is on normative influences specific to the workplace, personal
norms will not be included in this study. However, the formation of personal
norms can be influenced by social norms as a result of the internalization of
social references about appropriate behavior (Bamberg & Möser, 2007).
Social norms can be further split into injunctive norms and descriptive
6
norms based on their different sources of motivation. Cialdini et al. (1990) define
injunctive norms as “rules or beliefs as to what constitutes morally approved and
disapproved conduct” (p. 1015). Behaviors that align with injunctive norms
constitute what people ought to do and what will be socially sanctioned by others.
It is the need for social approval from others that motivates people to act in
congruence with injunctive norms. In contrast, descriptive norms are perceptions
about what others typically do in a certain context (Cialdini et al., 1990). People
presume that because other people are doing it, it is likely an effective behavior
and adaptive to the situation in question. People are motivated to use the
information as a situational heuristic to help simplify behavioral decision-making.
Likely related to their use as a heuristic (Jacobson et al., 2011; Johnson & Eagly,
1989), descriptive norms have been found to have a stronger direct relationship
with behavior (Manning, 2009; Thøgersen, 2006). This prompted their being
chosen as the focus of this study.
Injunctive and descriptive norms have also been qualified as subjective or
local norms, which differ in the referent of normative influence. Local norms
(Fornara et al., 2011), also called provincial norms (Goldstein, Cialdini, &
Griskevicius, 2008), are created by the influence of affectively unimportant others
who have shared the same physical space. For example, Goldstein et al. (2008)
found that hotel guests could be influenced to reuse their towels if they knew that
previous guests had reused their towels in that same hotel room. These previous
guests had shared the same physical space, but were of no affective importance
7
to the current guests. In comparison, subjective norms are created by the
expectations of affectively important others, such as family members or friends
(Fornara et al., 2011). This label comes from the use of subjective norms in the
Theory of Planned Behavior (Ajzen, 1991), which were also based on the
expectations of affectively important others. Thus, you can have local descriptive
or local injunctive norms as well as subjective descriptive and subjective
injunctive norms influencing behavior.
Though identified and defined in the PEB literature, the local-subjective
distinction is still somewhat unclear and will not be explored in this study. Very
few PEB studies have explicitly compared local norms and subjective norms (e.g.
Fornara et al., 2011; Goldstein et al., 2008). Goldstein et al. (2008) compared the
influence of local descriptive normative messages to the influence of social
identity normative messages on hotel towel reuse. Social identity was defined as
self-concept at the group level based on perceived membership, and could be
considered a more global descriptive normative influence than the local
message. What they found was that the local descriptive message promoted
greater reuse than did the social identity message. Additionally, Fornara et al.
(2011) examined the factor structure of local versus subjective descriptive and
injunctive norms and found four distinct, but correlated, constructs. Thus there is
some support for the local-subjective distinction. However, in non-laboratory
settings, others who are proximal and others who are affectively important can
easily be the same people (Fornara et al., 2011). This clouds the source of
8
normative influence and could make the distinction between local and subjective
norms somewhat arbitrary. Additionally, getting too specific about the location of
the normative referent would limit the generalizability of a scale and greatly
increase its length. As a result, the scale developed in this study may capture
both affectively important and affectively unimportant normative referents, but, in
doing so, should capture the overall descriptive norms experienced by
employees within their organizations.
The environmental psychology literature has identified and defined
descriptive norms as a distinct norm type, though the local-subjective concept is
still cloudy. Descriptive norms are externally enforced, and refer to what others
actually do in specific contexts. Using this definition, several studies have
examined the relationship between descriptive norms and PEBs. In particular,
the relationship has been explored in studies grounded in the Theory of
Normative Conduct (Cialdini et al., 1990) and the Theory of Planned Behavior
(Ajzen, 1991). Though PEBs and EGBs do not represent the same behavioral
domain, the relationship between PEBs and descriptive norms should inform the
understanding of the influence of descriptive norms on EGBs.
The Relationship Between Descriptive Norms
and Pro-Environmental Behaviors
The Focus Theory of Normative Conduct
Cialdini et al. (1990) proposed the Focus Theory of Normative Conduct to
help clarify mixed findings on the effects of social norms on behavior. They
9
identified two limitations in previous research. The first was an issue of definition.
The popular term “norm” could indicate either what was commonly done or what
was approved of by society (Cialdini et al., 1990; Kallgren et al., 2000; Reno et
al., 1993). To clarify, the researchers defined descriptive norms as what is
commonly done, and injunctive norms as what is approved of by society. Each
type of norm was a separate source of motivation that could influence behavior.
Thus, individual behavior could be influenced by norms of what is and norms of
what ought to be.
The second limitation identified by Cialdini and colleagues (1990, 1993)
was that because individuals were perfectly capable of internalizing contradictory
norms, norms could be used to explain any behavior. As noted by Cialdini
(2012), “accounts that can explain everything after the fact are probably too
vague or circular to explain anything” (p. 296). If people can act with or against a
norm, then how can we say it was influencing their behavior? To clarify this
circular argument, Cialdini et al. (1990) proposed that saliency was key to
whether a norm would hold sway in a certain situation. Therefore, individuals can
hold contradictory norms, but the norm most salient when a behavior is occurring
will be the norm influencing that behavior. Once proposed, early research tested
the theory using littering behaviors (Cialdini et al., 1990; Reno et al., 1993). Since
then, much of the research on the Focus Theory of Normative Conduct has been
tested through application to PEBs. This avenue of research has provided
valuable insight into the theory’s processes and influence.
10
The Outcomes of Normative Influence. By manipulating their saliency,
researchers have shown that descriptive norms can influence a variety of PEBs
in a range of environments. In a series of three studies, Cialdini et al. (1990)
found that descriptive littering norms could be manipulated through the presence
or absence of litter in parking lots, mailrooms, and amusements parks. People
would respond to the salient descriptive littering norm and litter less in a clean
environment and litter more in a littered environment. Reno et al. (1993)
replicated Cialdini et al.'s (1990) findings on littering behaviors; Kallgren et al.
(2000) extended these findings. They found that a salient descriptive norm could
influence behavior in private locations (e.g. alone in a stairwell). Schultz and
colleagues (1998, 2007) found that a descriptive normative message would
influence participants’ energy usage behaviors at home. In a study at the
Petrified Forest National Park, Cialdini (2003) noted that descriptive theft norms
conveyed through signs on park paths encouraged significantly more theft of
protected, petrified wood than did neutral messages on control signs. Goldstein
and colleagues (2007, 2008) studied the effects of descriptive norms and
environmental pleas on towel reuse at hotels. They found that descriptive norms
indicating that most people reuse their hotel towels prompted a 44 percent
increase in towel reuse. Additionally, this approach was more effective than an
environmental plea encouraging towel reuse in order to benefit the environment,
which prompted only a 35 percent increase in towel reuse.
In sum, descriptive norms have been found to influence a range of PEBs,
11
including littering, towel reuse, and energy conservation, in a range of
environments, including at home, in a hotel, at a national park, and in parking
lots. Their influence can be positive or negative depending on the content of the
norm (e.g. if stealing is indicated as the norm, then people will steal). Because
descriptive norms may not be the only social norm present in a given situation, it
would be beneficial to understand how norm alignment or misalignment would
affect behavioral outcomes.
The Interaction of Injunctive Norms and Descriptive Norms
. It may be
expected that what people actually do and what people should do are frequently
the same thing; however, there are instances when this alignment is not the
case. For example, people commonly use bottled water even though the bottle
can be damaging to the environment. To better understand social norms’
influence on PEBs, it is important to explore how behavioral outcomes differ
when injunctive and descriptive norms are aligned or misaligned.
If a situation is encouraging unwanted behavior through a descriptive
norm, then a salient injunctive norm should be used to help counteract its effects.
Cialdini (2003) explored the effects of norm salience on the theft of petrified wood
in the Petrified Forest National Park in Arizona. Signs highlighting a descriptive
theft norm (a.k.a. everyone steals wood), an injunctive anti-theft norm (a.k.a. you
shouldn’t steal wood so as to preserve the park), or a control, nonnormative
message were placed at three locations along park paths. Significantly more
wood was stolen from around the descriptive theft norm sign than from around
12
the control sign, and from around the control than from around the injunctive anti-
theft norm sign. This was a difference of 7.92 percent to 2.92 percent to 1.67
percent of people stealing wood, and indicated that a positive injunctive norm can
help reduce theft in an environment with negative descriptive norms. However, it
would have been more informative if they had created a balanced design by
including a descriptive no theft norm sign and an injunctive theft norm sign.
Even when a descriptive norm indicates that the desired behavior is
prevalent, it is advisable to pair it with a positive injunctive norm to prevent a
phenomenon entitled the boomerang effect (Schultz et al., 2007). The
boomerang effect occurs when an average amount of the desired behavior is
included in a descriptive norm. An example of such a message would be telling
people that others in their neighborhood utilize an average of 100 gallons of
water a day. By providing this behavioral anchor, people can compare and adjust
their behavior to align with the average. While studying recycling behaviors,
Schultz (1998) identified the boomerang effect after providing normative recycling
information to a California neighborhood. Those who had been recycling above
the normative level reduced their frequency of recycling, whereas those who had
been recycling below normative levels increased their frequency. However, if the
descriptive normative message is paired with an injunctive norm, then the
undesirable side of the realignment is mitigated. Schultz et al. (2007) found that
the boomerang effect was negated when households were provided an average
descriptive energy consumption norm along with a smiley face for those
13
consuming below the average, or a frowning face for those consuming above the
average. Individuals who were consuming more that the average still reduced
their energy consumption, but individuals who had been consuming below the
average continued to maintain their low levels of consumption. Additionally, these
effects held up even after participants were no longer receiving the normative
messages. So, providing an injunctive norm along with a descriptive norm
encouraged the desired behavior while negating the boomerang effect.
To prompt the most desired outcomes, it is best to have the two norms in
alignment. In a study about recycling, Cialdini (2003) created a set of three public
service announcements (PSAs) in each of which people were engaged in
recycling, spoke approvingly of recycling, and disapproved of a person who was
not recycling. Thus, these PSAs highlighted both positive descriptive recycling
norms as well as positive injunctive recycling norms. When examining the
tonnage of material recycled as a result of these PSAs, the experimental
communities who received the PSAs (i.e. all three messages) exhibited a 25.35
percent net advantage for material recycled compared to the control communities
who received no messages. In a follow up study, college students viewed and
rated the three PSAs on several relevant dimensions, including injunctive and
descriptive recycling norms, humor, and ad content. The results of this study
supported the proposition that, while not the only influence, injunctive and
descriptive recycling norms did significantly influence recycling outcomes as a
result of the PSAs. In regards to behavioral intentions, Smith et al. (2012) found
14
that intentions to conserve energy were highest when both the injunctive and
descriptive energy conservation norms were positive. Additionally, they found
that when both norms are salient but in conflict, having one norm support the
desired behavior while the other is unsupportive prompts a reduction in the
intention to engage in the desired behavior. So, having both descriptive and
injunctive norms in alignment has a highly positive influence on both intentions
and actual behaviors.
As distinct sources of motivation, the two types of norms have been found
to interact in interesting ways. Salient injunctive norms can reduce the influence
of undesirable descriptive norms, and vice versa. The strongest outcomes are
produced when both norms are salient and in agreement. Yet, research has
indicated that, although an effective source of influence, descriptive norms are
commonly unrecognized and underutilized to affect behavior.
The Influence of Descriptive Norms is Underdetected
. Research has found
that descriptive norms can influence behavior, both through witnessing the
behavior and through written messages. However, it has also been shown that
people tend to be unaware of and/or deny this influence (Cialdini, 2007;
Goldstein et al., 2007, 2008; Nolan, Schultz, Cialdini, Goldstein, & Griskevicius,
2008). Nolan et al. (2008) surveyed Californian residents concerning their energy
conservation behaviors, the importance of various reasons for these behaviors,
and their broad beliefs and their descriptive normative beliefs about energy
conservation. Though normative reasons were rated last below saving the
15
environment and benefiting society, they were the only reason for conservation
that significantly correlated with participants’ self-reported energy conservation
efforts. Additionally, in a follow up study that measured actual energy use,
normative messages were still rated as least motivational, but those who
received the normative messages consumed significantly less energy than those
who received the nonnormative messages (Nolan et al., 2008). Even though their
behavior was significantly influenced by these descriptive norms, people believed
that helping the environment was the reason for their behavior.
Furthermore, likely because they are unaware of descriptive norms’
influence, people in positions of power rarely use descriptive norms to encourage
PEBs. For example, hotels commonly encourage guests to reuse their towels via
environmental pleas or references to saved costs. Yet, Goldstein and colleagues
(2007, 2008) found that using a normative message increased hotel towel reuse
up to 44 percent on the first night, compared to only 30 percent when using an
environmental plea. On a wider scale, descriptive norms could also be used to
encourage the public to adhere to environmental regulation. Cialdini (2007)
outlines that many regulatory agencies commit the common error of emphasizing
the prevalence of bad behavior (a.k.a. negative descriptive norms). If they
emphasized the positive behavior instead, it would cost the same but encourage
better adherence to regulations.
Despite people’s unawareness of the influence of descriptive norms, these
norms still have a large impact on behavior. This impact makes descriptive
16
norms more useful than environmental pleas for encouraging desired PEBs.
Additionally, the ability to utilize them effectively should be increased by a better
understanding of how descriptive norms work.
How Do Descriptive (and Injunctive) Norms Work?
Research on the
processes behind the differential influence of descriptive and injunctive norms
has been limited. In response, there have been calls in the literature for such
research (Cialdini, 2012; Jacobson et al., 2011; Schultz et al., 2007), and recent
studies have produced some interesting findings.
First, the importance of norm saliency has been reiterated in numerous
studies (e.g. Cialdini et al., 1990; Kallgren et al., 2000; Reno et al., 1993). Two
common methods to induce norm saliency are modeling the desired behavior
and conveying normative messages. Role modeling was used in a series of three
experiments by Cialdini et al. (1990) to manipulate descriptive norm saliency of
littering behavior. They found that observers would respond to the salient norm
by either littering more in a littered environment, or littering less in a clean
environment. In a fifth experiment, Cialdini et al. (1990) switched to a series of
normative messages delivered on handbills to manipulate anti-littering injunctive
norms. Other studies that used written normative messages to manipulate norm
saliency have examined environment theft (Cialdini, 2003), and reuse of hotel
towels (Goldstein et al., 2008, 2007). Both found that normative messages could
influence norm saliency, and thus behavior.
Arousal and focusing techniques have also been used to manipulate
17
norm saliency, but have only been used in one study each. Kallgren et al. (2000)
found that arousal moderated the relationship between normative messages and
behavioral outcomes. Individuals who were more aroused responded more
strongly to the anti-littering messages they had been exposed to, whereas
participants who were not aroused did not respond to the normative message in
a systematic manner. Based on previous research (Berkowitz & Buck, 1967;
Hockey & Hamilton, 1970), it was proposed that the increased arousal prompted
participants to focus more intently on the dominant features of the situation,
which was the normative anti-littering message. Kallgren et al. (2000) also
utilized focusing techniques to either induce an inward, self-focus or an outward,
external focus, thus inducing saliency for either the external social norms, or
internal personal norms. When focus was placed on the social norms, rate of
littering was about the same whether the individual had strong personal anti-
littering norms or not. Conversely, when focus was inward, those with strong,
personal anti-littering norms littered less than those with weak, personal anti-
littering norms. Though less common than role modeling and normative
messages, arousal and focusing techniques have been supported as inducing
the saliency of norms.
Second, several studies indicate that injunctive norms operate through
greater cognitive processing than do descriptive norms. To start, high personal
involvement has been found to reduce the influence of descriptive norms, while
increasing the influence of injunctive norms (Gockeritz et al., 2010). This result
18
has been attributed to greater personal involvement leading to the use of
elaborative, central processing of the message instead of superficial, peripheral
processing (Johnson & Eagly, 1989). Building on this finding, Jacobson et al.
(2011) found that when self-regulatory capacity was depleted, individuals
increased their adherence to descriptive norms, while reducing their adherence
to injunctive norms. Because descriptive norms influence behavior using mental
shortcuts, decreased self-regulatory functioning should make it more difficult to
consider alternative options than to simply follow the norm. In comparison,
injunctive norms require individuals to compare their immediate, personal
interests against the standards of society. If depleted of their ability to self-
regulate, they will choose their own interests over complying with society’s
wishes and the injunctive norm. Additionally, some studies have identified
injunctive norms as influencing behavior through a cognitive assessment of the
normative message, including its persuasiveness and its level of vividness
congruency. Vividness-congruency is a measure of the alignment between the
message and the image it provokes (Cialdini, 2003; Oceja & Berenguer, 2009).
Overall, research has identified that injunctive norms influence behavior through
greater cognitive processing, whereas descriptive norms influence behavior
directly or through simpler processing.
By understanding the processes behind the influence, researchers will
better understand how and when they get the behavioral outcomes that they do.
Norms will have no systematic effect on behavior without norm saliency.
19
Additionally, research has indicated that injunctive norms require more cognitive
processing to take effect than do descriptive norms. Thus, both types of norms
have the power to influence behaviors, but descriptive norms are likely to
influence behavior directly.
Summary of Research on Theory of Normative Conduct
. In summation,
descriptive norms have been found to influence a variety of PEBs in a variety of
contexts. This process exhibits certain characteristics. First, the effect of a
descriptive norm can be enhanced or reduced depending on whether it is aligned
with relevant injunctive norms. Alignment of the two types of norms produces the
best results, and can prevent unintended outcomes such as the boomerang
effect. Second, people do not realize the effect that descriptive norms can have
on their actions. This unawareness results in the underutilization of descriptive
norms to encourage desired behavior. Third, a norm must be salient in order to
have an influence on behavior. Research has found that saliency can be induced
through role modeling, focusing techniques, normative messages, or arousal.
Fourth, descriptive norms require less cognitive processing to influence behavior,
allowing them to influence behavior more directly and to be used as heuristics.
Combined, these four factors suggest that positive EGB descriptive norms within
the workplace could be an effective and unnoticed way to shape behavior, and
support the decision to focus on descriptive norms in this study.
The Theory of Planned Behavior
The Theory of Planned Behavior (TPB) is a rational choice model wherein
20
behavioral intention, and thus behavior, is influenced by the three predictors of
subjective norms, attitudes, and perceived behavioral control (Ajzen, 1991).
Subjective norms are conceptualized in the model as the behavioral expectations
of relevant others (i.e. injunctive norms rather than descriptive norms). Attitudes
are conceptualized as evaluations about the intended behavior and its outcomes.
Perceived behavioral control represents perceptions of personal control over
performing the behavior. This model has been used to explain a wide range of
PEBs including mode of travel (Bamberg & Schmidt, 2003), water use,
purchasing energy-saving light bulbs, using unbleached paper (Harland, Staats,
& Wilke, 1999), and general pro-environmental behavior (Kaiser, Wolfing, &
Fuhrer, 1999).
Originally, TPB only included norms conceptualized as injunctive norms;
however, more recent research has begun to include descriptive norms as a way
to improve the variance explained by the model. Meta-analyses of the TPB have
found that the three traditional predictors account for 39 percent of the variance
in behavioral intention (Armitage & Conner, 2001; Rivis & Sheeran, 2003).
Although only a few studies have included descriptive norms as a part of the
model, a meta-analysis of these studies found that including descriptive norms
accounted for an additional 5 percent of variance in behavioral intention, after
controlling for the other predictors (Rivis & Sheeran, 2003). Additionally, the
correlation between descriptive and injunctive norms was only .38, lending
support to their discriminant validity and indicating that the effect of descriptive
21
norms was not due to their similarity to injunctive norms. Thus, including
descriptive norms improved the model and furthered the understanding of what
influences the enactment of PEBs.
Examining the relationships among the variables of the TPB has also
provided initial support for the processing differences identified in the literature on
the Theory of Normative Conduct. Manning (2009) found that although injunctive
norms were more strongly correlated with the other TPB predictors, descriptive
norms had a slightly stronger relationship with behavior. Additionally, a direct
path from descriptive norms to behavior significantly improved the TPB model. A
second finding was that the effect of descriptive norms on behavior did not
weaken as the time between cognition and behavior increased, whereas the
effect of injunctive norms on behavior did. These results seem to support the
idea that descriptive norms require less cognitive processing than do injunctive
norms, and operate as a heuristic for effective decision-making.
Though research on descriptive norms and the TPB is minimal, inclusion
of descriptive norms has been shown to improve the model and broaden our
understanding of what factors influence enactment of PEBs. Analysis of the
relationships among the variables in the model supports the idea that descriptive
and injunctive norms are distinct, and that descriptive norms require less
cognitive processing to influence behavior.
Summary of Research on the Theory of Normative Conduct and the Theory of
Planned Behavior
22
Research on the TPB and research on the Theory of Normative Conduct
are approached from different theoretical perspectives; however, the findings
from both in regards to descriptive norms seem to be complementary. Under
both theories, descriptive norms have been identified as a meaningful predictor
of behavior, related to but distinct from injunctive norms. The boundaries of the
influence of descriptive norms have been more fully explored under the Theory of
Normative Conduct, but both theories suggest that descriptive norms operate
directly on behavior and require less processing to influence behavior than do
injunctive norms. As a meaningful influence on behavior, being able to properly
measure descriptive norms would be beneficial for both organizations and
researchers. However, measurement of PEB descriptive norms has been
inconsistent within the psychological literature.
Previous Measurement of Pro-Environmental Behavior
Descriptive Norms: In and Out of the Workplace
Pro-environmental behaviors have been studied minimally in the
workplace (Ones & Dilchert, 2012b). A search of the scientific literature returned
three studies that had used a descriptive norms scale to examine the relationship
of PEB descriptive norms to other workplace phenomena. Evans, Russell,
Fielding, and Hill (2012) were interested in the effects of an energy conservation
intervention on organizational outcomes. They assessed the intervention’s
effects on energy-related descriptive norms through the single item, “Most staff
save energy in the workplace.” The item’s content was dictated by the needs of
23
the study, and it was not intended to be a comprehensive representation of
workplace PEB descriptive norms. Robertson and Barling (2013) were interested
in capturing the environmental descriptive norms of organizational leaders. Due
to a lack of existing measures, they developed their own 5-item scale composed
of items such as, “Do your friends and/or family endorse environmentally-friendly
programs?” Nag (2012) adapted items from Gärling, Fujii, Gärling, and
Jakobsson (2003) to assess perceptions of the general public’s engagement in
PEBs. Though both studies explored PEBs in the workplace, both assessed the
influence of descriptive norms created by referents outside the workplace (i.e.
friends and family, the general public). Thus, neither developed a multi-item
measure capturing workplace-specific descriptive norms, which should be the
descriptive norms with the most influence on behaviors in the workplace.
Outside of the workplace, there are a few PEB descriptive norm scales
used in research in private and other public settings (e.g. Nolan et al., 2008;
Smith et al., 2012). However, little attention has been paid to developing a multi-
item, validated measure capturing this construct. One reason is linked to theory,
in that any research grounded in the TPB model must adhere to two
measurement requirements. First, any measured behavior must be clearly
defined in terms of its Target, Action, Context, and Time (TACT; Ajzen, 2002).
Second, any measured predictor variable must comply with the compatibility
principle, which states that the measures of predictor variables must match the
behavioral measures in specificity of TACT (Ajzen & Fishbein, 1977; Ajzen,
24
2002). This specificity is intended to improve the reliability of the information
assessed by the items (Ajzen & Sexton, 1999; Ajzen, 2002). However, this
requirement tends to produce highly specific single or dual item measures (e.g.
Fornara et al., 2011). Such specific measures are inherently unreliable (Kaiser,
Schultz, & Scheuthle, 2007) because they are unlikely to capture the entire
domain of behavior. Trying to remedy this issue, while adhering to the two
measurement rules, would require an extensive scale to capture all behavioral
variations. This would be prohibitive and likely lead to participant fatigue.
A second reason for the lack of multi-item measures seems to be due to
the debate surrounding the aggregation of behavioral-based scales. Within the
research on PEBs, there is disagreement between researchers who propose that
PEBs can be measured through a general scale, and those who propose that
PEBs are different from and independent of one another (Kaiser, 1998). This
disagreement is in part due to the wide range of possible PEBs and the variety of
ways they could be aggregated. When it is assumed that PEBs do not
generalize, it results in the use of very specific, single-item scales (Kaiser, 1998).
Because descriptive norms are inherently tied to their referent behaviors, this
same issue then plagues descriptive norm scales. For example, Thøgersen
(2006) measured four different possible PEBs (e.g. buying organic milk,
composting kitchen waste) when assessing his norm taxonomy. The norm for
each behavior was assessed as a single item due to the inherent differences in
behaviors.
25
A third reason is that, as with the scale by Evans et al. (2012), short
descriptive norms scales are sometimes created to suit the exact needs and
context of a specific study. For example, Nolan et al. (2008) created a 3-item
descriptive norms scale that was specific to their study concerning energy
conservation at home. An example item is, “How often do you think residents of
your city try to conserve energy?” Frequently, these scales are used to check the
effectiveness of an experimental manipulation. Oceja and Berenguer (2009)
examined the normative influence of leaving bathroom lights on or off in a public
bathroom. To assess whether manipulating the lighting norm affected normative
perceptions, they used the single item, “To what extent do you think that most
people leave the lights on when exiting a public bathroom?” Similarly, Smith et al.
(2012) created a 3-item descriptive norms scale to assess an experimental
manipulation concerning energy conservation norms. In these studies, scale
content was dictated by the specific behaviors being examined. These situations,
just like the aggregation disagreement and using TPB, result in the use of single
or small groups of very specific scale items.
Compared to single-item, study specific descriptive norms scales, a multi-
item, validated EGB descriptive norms scale would provide many methodological
benefits. First, the use of multiple items to capture a phenomenon helps produce
more consistent and stable responses, which make the scale more precise and
reliable (Bowling, 2005). Second, the use of multiple items allows for item
aggregation, which helps increase the generalizability and replicability of study
26
results (Epstein, 1983). Third, the availability of an EGB Descriptive Norms Scale
would allow for comparison of EGB descriptive norms across studies and
organizational contexts (Bowling, 2005). Finally, scale validation provides support
to the assumption that the scale is measuring the intended construct (Schultz &
Whitney, 2005). All these factors contribute to a more consistent, reliable, and
comprehensive measurement of the phenomenon of interest.
Existing scales capturing PEB descriptive norms are limited due to
theoretical, conceptual, and study-specific reasons. Additionally, the behavioral
domain of EGBs is slightly different from PEBs. Thus, a multi-item, validated
EGB Descriptive Norms Scale would provide methodological benefits, and would
be useful for capturing the breadth of EGBs that could inform descriptive norms.
To identify what these behaviors are requires an examination of the Green Five
Taxonomy created by Ones and Dilchert (2012a).
The Green Five Taxonomy of Employee Green Behaviors
To understand the normative influence of descriptive norms on EGBs, it is
critical to identify what kinds of behaviors represent EGBs. Though a few studies
in the PEB literature have looked at PEBs enacted at work, these have typically
been narrowly focused on single behaviors, such as recycling (e.g. McDonald,
2011; Tudor, Barr, & Gilg, 2007). Although such behaviors may capture a
component of green behaviors at work, they are not representative of the broad
range of EGBs. Recently, Ones and Dilchert (2012a) addressed this limitation by
27
developing the Green Five Taxonomy.
The Green Five Taxonomy was developed using a critical incidents
methodology (Ones & Dilchert, 2012a). Critical incidents addressing behaviors at
work that either benefited or hurt the environment were collected from U.S.
employees working in a multitude of job positions, organizations, and industries.
Incidents were sorted to create behavioral categories, and these categories were
confirmed on an additional set of critical incidents. Categories were then tested
using critical incidents from employees in Europe to assess cross-cultural
relevance and generalizability (Hill et al., 2011 as cited in Ones & Dilchert,
2012a). This process produced sixteen behavioral categories that were
functionally distinct and internally homogenous.
Organizing the sixteen categories resulted in a three-tier taxonomy (Ones
& Dilchert, 2012a). The top tier is a general factor titled General Green
Performance. This tier was identified through the correlation of some of the
categories, supported by both supervisory and self-reports. The second tier
consists of five meta-categories: Working Sustainably, Avoiding Harm,
Conserving, Influencing Others, and Taking Initiative. The third and lowest tier
contains the original 16 categories. Each category belongs to a single meta-
category and anywhere from two to four of these categories are subsumed under
each meta-category. Each meta-category will be described in turn. See Figure 1
for a visual representation of the 2
nd
and 3
rd
tiers of the taxonomy.
The meta-category of Working Sustainably represents behaviors that help
28
work processes and products be more sustainable (Ones & Dilchert, 2012a). For
example, a supervisor could order a desk made from sustainably grown oak
trees or one made from endangered redwood trees. The four subsumed
categories include Choosing Responsible Alternatives, Changing How Work is
Done, Creating Sustainable Products and Processes, and Embracing Innovation
for Sustainability. Choosing Responsible Alternatives involves choosing the more
environmentally friendly option available, whereas Changing How Work is Done
involves changing work processes to become more sustainable. These two
categories reflect making modifications to existing products and processes. In
comparison, Embracing Innovation for Sustainability and Creating Sustainable
Products and Processes reflect creating and embracing new processes and
products. Incidents of Choosing Responsible Alternatives were the most common
behaviors for this meta-category. Psychologically, Working Sustainably
represents adaptability.
The Avoiding Harm meta-category is bipolar and contains three categories
(Ones & Dilchert, 2012a). Behaviors can either harm the earth and cause
increasing damage, or can enhance the earth, making its ecosystems healthier.
Psychologically, these behaviors are linked to altruism and responsibility on one
end, and lack of responsibility and self-control on the other. The primary category
under this meta-category is Polluting/Preventing Pollution, which captures
behaviors that pollute the environment or prevent pollution. The other two
categories of Monitoring Environmental Impact and Strengthening Ecosystems
29
support the primary category. Monitoring Environmental Impact represents
observing and assessing the environment to understand how work activities are
affecting it. Strengthening Ecosystems includes behaviors that help protect or
repair ecosystems from the effects of industry and business.
The Conserving meta-category represents behaviors related to helping
preserve resources and reduce waste (Ones & Dilchert, 2012a). A positive
example would be double-sided printing, whereas a negative example would be
leaving work computers on overnight. This meta-category, listed from highest
environmental impact to lowest, contains the four categories of Reducing Use,
Reusing, Repurposing, and Recycling. Reducing Use prevents the unnecessary
use of new materials. Reusing involves multiple uses of the same materials for
the same purpose, while Repurposing involves multiple uses of materials for new
purposes. Recycling allows for old materials to become new products, but
requires energy and additional resources to do so. Conserving EGBs comprised
about half of the total behavioral incidents, with Reducing Use and Recycling
being the most common behaviors within the Conserving meta-category.
Psychologically, Conserving represents thrift or frugality.
The meta-category of Influencing Others moves from what the individual
employee can accomplish to how individuals can influence each other to engage
in environmental behaviors. Psychologically, Influencing Others is associated
with spreading knowledge and helping others change their behaviors. Ones and
Dilchert (2012a) note that it is the only meta-category that is explicitly social and
30
that the influence can extend to other stakeholders in the company, such as the
local community. The two subsumed categories are Encouraging and Supporting
Others, which includes behaviors that bolster and encourage other’s EGBs, and
Educating and Training for Sustainability, which includes behaviors that help
others build their knowledge about environmentalism.
Taking Initiative, the last meta-category, captures behaviors that involve
stepping outside the box, taking a risk, and encouraging environmentally-related
change (Ones & Dilchert, 2012a). The focus is on how individuals encourage and
promote environmentally-friendly behaviors, so the behaviors being encouraged
might be included under the other meta-categories. For example, an employee
could help initiate a policy (a Taking Initiative behavior) that requires others in the
organization to buy sustainably produced printer paper (a Working Sustainably
behavior). This meta-category includes the three categories of Putting
Environmental Interests First, Initiating Programs and Policies, and Lobbying and
Activism. Putting Environmental Interests First captures behaviors that help the
environment at some personal cost to the individual. Initiating Programs and
Policies involves pushing for new programs and policies within the environmental
domain, whereas Lobbying and Activism capture behaviors that involve fighting
for environmental causes.
Due to the thorough critical incidents technique and process used to
develop the Green Five Taxonomy, there is strong evidence that the taxonomy
represents the breadth of possible EGBs (Ones & Dilchert, 2012a). Using the
31
taxonomy as the foundation of the EGB Descriptive Norms Scale lends support
to the scale capturing the full range of normative influence generated by these
behaviors. As a result, the scale should help identify which meta-categories are
being enacted, and how they are contributing to the overall strength of EGB
descriptive norms in an organization. Additional analysis examining the construct
validity of the EGB Descriptive Norms Scale will provide support for the
assumption that the new scale is capturing the intended norms.
The Present Study
In response to the lack of available measures and the recently created
Green Five Taxonomy, an EGB Descriptive Norms Scale was developed and
validity evidence was gathered. First, an initial item pool was developed based
upon the Green Five Taxonomy of EGBs (Ones & Dilchert, 2012a) and Cialdini et
al.'s (1990) definition of descriptive norms. Through pilot testing and an item
retranslation task (Smith & Kendall, 1963), data were collected to assess
subscale reliability and examine the content and clarity of initial scale items.
Using this data, items were revised, replaced, or removed. Second, new data
was collected on the refined scale and two additional constructs, a confirmatory
factor analysis was conducted to identify whether the EGB Descriptive Norms
Scale’s structure mimicked the structure of the Green Five Taxonomy, and
evidence for convergent and discriminant validity of the EBG Descriptive Norms
Scale was gathered.
32
CHAPTER TWO
DEVELOPMENT OF THE EMPLOYEE GREEN BEHAVIOR
DESCRIPTIVE NORMS SCALE
Item Development
An initial pool of 40 items was developed using One’s and Dilchert’s
(2012a) definition of EGBs, their Green Five Taxonomy, and Cialdini et al.'s
(1990) definition of descriptive norms. To capture the range of possible behaviors
influencing EGB descriptive norms, two to three items were written for each of
the 16 behavioral categories that comprise the taxonomy’s third and most
specific tier. Thus, the five meta-categories comprising the second tier of the
Green Five Taxonomy (i.e. the subscales of the EGB Descriptive Norms Scale)
were represented by six to ten items each. All items represented possible EGBs
that could be observed in an organizational environment; some of the items were
adapted from Nag (2012). Items were assessed for clarity and reading level by
two tenured professors and 12 undergraduate students enrolled at a mid-sized
public university in southern California. Responses were used to make items
easier to read and to reduce item ambiguity. Two versions of the initial survey
were created. Both versions contained the same 40 items. One version was
formatted as a 5-point, frequency response scale (1 = never, 5 = always). The
other was formatted as a 5-point, Likert-style scale (1 = strongly disagree, 5 =
strongly agree). Upon reverse coding negatively worded items, higher values
33
indicated stronger descriptive norms. The initial EGB Descriptive Norms Scale
can be found in Appendix A.
Instrument Refinement
Pilot Test Procedure
A survey packet was created for each version of the initial scale. Each
contained an informed consent, the 40-item scale, demographic questions, work-
and industry-related questions, and a debriefing form. Surveys were distributed
to undergraduate students enrolled in four classes at a mid-sized public
university in southern California. Directions were to use either current work
experience or a past work experience to answer the 40-item scale. If using a past
work experience, participants were told to reference that same work experience
when answering the demographic, work- and industry-related questions. At the
discretion of the professor, participants received research credit or extra credit in
the course for their participation. No identifying information was collected making
the responses anonymous.
Sample
The total pilot test sample consisted of 274 responses (142 frequency
scale responses, 132 Likert scale responses). Of the 142 participants who
responded to the frequency scale version, 81.0 percent were female (n = 115)
and the average age was 25.34 years (SD = 6.75). The sample was 62.0 percent
Hispanic (n = 88), 21.8 percent White (n = 31), 3.5 percent African-American (n =
34
5), 2.8 percent Asian-American (n = 4), 0.7 percent Pacific Islander (n = 1), and
9.2 percent Other (n = 13). On average, participants had been at their
organization for 2.88 years (SD = 2.90), worked either from ten to 19 hours per
week (27.5%, n = 29) or from 20 to 29 hours per week (25.4%, n = 36), and was
a non-management/hourly worker (58.5%, n = 83). Industry type was not
requested in this version of the survey.
Of the 132 participants who responded to the Likert scale version, 81.1
percent were female (n = 107) and the average age was 24.86 years (SD =
6.70). The sample was 57.6 percent Hispanic (n = 76), 24.2 percent White (n =
32), 5.3 percent African-American (n = 7), 5.3 percent Asian-American (n = 7),
2.3 percent Pacific Islander (n = 3), and 5.3 percent Other (n = 7). An average
participant had been at his/her organization for 2.50 years (SD = 2.28), worked
24.33 hours per week (SD = 11.56), and was a non-management/hourly worker
(62.1%, n = 82). The most common industries were Sales and Related (19.7%, n
= 26) and Education/Training (18.2%, n = 24), followed by Office/Administration
(13.6%, n = 18), Food Preparation/Serving (12.1%, n =16), Healthcare (8.3%, n =
11), Transportation/Materials Moving (4.5%, n = 6), Production (0.8%, n = 1), and
Other (11.4%, n = 15).
Subscale Reliability Analysis
Basic data screening was performed on the pilot test data. Using a z-score
criterion set at p < .001, three univariate outliers were removed from the
frequency scale data and one univariate outlier was removed from the Likert
35
scale data. One multivariate outlier was removed from the Likert scale data
based on a Mahalanobis distance criteria set at p < .001. Several items in each
scale version were skewed based on a z-score criterion set at p < .001, but were
not transformed for the sake of interpretation. No variable was missing more than
5 percent data, indicating the data was missing completely at random. Missing
data were imputed using the expectation maximization algorithm, which is an
accepted estimation method for data missing completely at random (Tabachnick
& Fidell, 2013).
Reliability was assessed at the subscale/meta-category level of the EGB
Descriptive Norms/Green Five Taxonomy. See Table 1 for results of the reliability
analysis. With the Likert scale data, Cronbach’s alpha was as follows: Working
Sustainably (α = .82), Avoiding Harm (α = .73), Conserving (α = .77), Influencing
Others (α = .74), and Taking Initiative (α = .69). Reliability could be improved for
the Conserving, Influencing Others, and Taking Initiative subscales through the
removal of five items. Specifically, the data suggests removing Item 3 from the
Conserving subscale, Items 36 and 39 from the Influencing Others subscale, and
Items 8 and 13 from the Taking Initiative subscale.
Examination of the frequency scale data revealed much lower reliability
values for all subscales. Cronbach’s alpha for Working Sustainably was .71,
Avoiding Harm was .52, Conserving was .78, Influencing Others was .61, and
Taking Initiative was .69. Reliability could be improved for all subscales through
the removal of nine items. Specifically, the data suggests that reliability would be
36
improved by removing Item 32 from Working Sustainably, Items 19, 14, and 30
from Avoiding Harm, Item 3 from Conserving, Items 36 and 39 from Influencing
Others, and Items 8 and 13 from Taking Initiative. However, even after removing
these items, subscale reliability would still be lower than it was using the Likert
scale data.
Overall, results suggest that using the Likert scale may be a better choice
than using the frequency response scale. Using the Likert data, the five items
identified as improving reliability through their removal were examined for
possible revision. See Appendix B for initial scale items arranged by
subscale/meta-category.
Item Retranslation Task
Using the process outlined by Smith and Kendall (1963), six subject
matter experts independently retranslated the 40 items into the 16 behavioral
categories identified and defined by Ones and Dilchert (2012a). Results of the
retranslation task were reviewed for rater agreement. An acceptable hit rate was
set at four out of six raters categorizing the item correctly (i.e. 67% correct
categorization for each item). Eighteen items did not meet this threshold and,
using the categorization data, were examined for revision. Twelve of the
troublesome items were reverse-coded items. See Table 2 for a complete list of
items and their corresponding hit rates. See Appendix C for retranslation task
instructions and categories.
Final Revisions
37
Final revisions were made using item skewness data, the subscale
reliability data, and the results of the retranslation task. In the end, four items
were revised, one item was removed due to its highly varied retranslation task
results, and six items were replaced resulting in a final scale with 39 items. As
suggested by the subscale reliability analysis, a Likert response format was used
in the refined scale. The refined scale can be found in Appendix D; bolded items
were the revised or replaced items.
38
CHAPTER THREE
CONFIRMING THE FACTOR STRUCTURE AND VALIDATING
THE EMPLOYEE GREEN BEHAVIOR
DESCRIPTIVE NORMS SCALE
Pilot test and retranslation task data provided preliminary evidence for a
39-item scale comprised of five reliable subscales. A new, larger sample was
collected to test the expected scale structure through a confirmatory factor
analysis (CFA). Data was also collected on two additional constructs to assess
the convergent and discriminant validity of the EGB Descriptive Norms Scale.
Construct Validity
As explained by Shultz and Whitney (2005), evidence of construct validity
can be gathered by examining the relationships between a newly developed
scale and other constructs. Constructs that should relate and should not relate to
the new scale are identified through theory and the results of previous research.
The new scale is said to exhibit convergent validity when it relates to other
constructs to which it is expected to relate, and to exhibit discriminant validity
when it does not relate to other constructs to which it is not expected to relate. If
the new scale’s relationships to other constructs are supported by theory, it
provides evidence that the new scale represents the intended construct and
39
exhibits construct validity (Schultz & Whitney, 2005). The new scale’s
relationships with other constructs are what form its nomological network. To
begin exploring the nomological network of the EGB Descriptive Norms Scale
and gathering evidence of construct validity, two constructs were identified
through previous research.
Ethical Leadership
Ethical leadership is promoting normatively appropriate behavior through
role modeling, social interaction and communication, decision-making, and
behavioral reinforcement (Brown, Treviño, & Harrison, 2005). The promoted
behavior represents a range of ethical values including fairness, compassion,
honesty, and altruism (Yukl, Mahsud, Hassan, & Prussia, 2013). To comply with
these values, ethical leaders must exhibit an awareness of how their behaviors
affect immediate others, their organization, and society at large (Kalshoven, Den
Hartog, & De Hoogh, 2011). Thus, it is likely that ethical leaders would consider
the environmental impact of their behaviors and, through their actions, would
influence others to do the same. This tendency would contribute to the creation
of descriptive norms around EGBs. However, leaders are not the only source of
normative information within an organization, and environmental issues are not
the only issues that ethical leaders can choose to champion. Thus, the two
constructs are similar, but also distinct, and are expected to exhibit a moderate,
positive relationship.
Work-Family Culture
40
Thompson, Beauvais, and Lyness (1999) define work-family culture as
“the shared assumptions, beliefs, and values regarding the extent to which an
organization supports and values the integration of employees’ work and family
lives” (p. 394). They propose the construct to have three dimensions. Dimension
one, organizational time demands, refers to the amount of time an employee’s
organization expects them to spend at work. The second dimension, managerial
support for work-family balance, refers to the level of managerial support
employees experience for balancing their work and family lives. Career
consequences is the third dimension, and refers to employees’ perceiving
negative career outcomes if they use work-family benefits. Overall, work-family
culture is proposed to affect attitudes about the organization along with behaviors
and/or perceptions related to employees’ handling their work and family lives
(e.g. using work-family benefits; Mauno, Kinnunen, & Piitulainen, 2005). As such,
work-family culture should not affect or relate to organizational norms concerning
employee green behaviors, and the two constructs are expected to exhibit a
weak correlation.
Method
Procedure
Participants were recruited through a Qualtrics.com professional recruiting
panel after having met certain requirements. They were required to be English-
speaking adults over the age of 18 who were working a least part-time (i.e. 20+
41
hours/week), were not students, and had been at their current company for at
least one year prior to participation in this study. These requirements help ensure
that participants have spent enough time at their organization to understand its
norms related to EGBs and to be able to adequately answer questions about
these norms. Each participant was offered a small, monetary incentive for his or
her participation.
All participants completed the survey online through Qualtrics.com.
Participants read an informed consent, completed the included scales and
demographics form, and were debriefed and thanked for their participation.
Surveys were completed individually and participants were assured of the
confidentially of their responses and the anonymity of their participation.
Participants took a median of 15 minutes to complete the survey.
Sample
A power analysis indicated that a sample size of 61 participants was
needed to achieve a power level of .80 (Preacher & Coffman, 2006). This power
analysis was based on the root-mean-square error of approximation (RMSEA) fit
statistic and a test of the not-close hypothesis where null RMSEA = .05,
alternative RMSEA = .01, alpha = .05, and df = 697 (MacCallum, Browne, &
Sugawara, 1996). Though such a small sample would ensure adequate power, a
larger sample size was required to support the use of RMSEA as a measure of fit
and to use maximum likelihood for parameter estimation (MacCallum et al.,
1996). Referring to the scientific literature, a general rule of thumb for
42
determining sample size when conducting a CFA is to have a 10:1 ratio of
indicator variables to participants (Nunnally, 1967) or a 5:1 ratio of free
parameters to participants (Bentler & Chou, 1987). This approach would suggest
a sample size of 390 or 415 participants. However, model simulations and Monte
Carlo studies indicate that the relationship between parameters and sample size
is not linear (Westland, 2010), and Westland (2010) proposes a formula, n ≥ 50r
2
– 450r + 1100, to calculate sample size based on the ratio of indicator variables
to latent variables (r = p/k). This formula indicates that a sample size of 288
would be appropriate. Based on these three estimations, a sample size of 400
should suffice to conduct the needed CFA.
Four hundred surveys were completed. Survey completion included
passing the demographic requirements and responding correctly to four careless
responding items. After data screening, 367 usable cases remained. As can be
seen in Table 3, the final sample was 51.5 percent female (n = 189), an average
45 years old (SD = 11.8), and predominantly White (78.7%, n = 289), followed by
African-American (8.4%, n = 31), Hispanic/Latino (7.6%, n = 28), Asian-American
(3.3%, n = 12), Native American (1.1%, n = 4), and Bi-racial/Multi-racial (0.8%, n
= 3). The majority of participants held positions as non-management/hourly
employees (26.4%, n = 97), professionals (21.0%, n = 77), or middle
management (19.3%, n = 71). They worked an average 41.4 hours per week (SD
= 7.1), had been at their current company for an average 10.2 years (SD = 9.0)
and 4.1 months (SD = 3.15), and worked in industries such as
43
Office/administrative support (20.7%, n = 76), followed by Other (18%, n = 66),
Sales and related (15.0%, n = 55), Production (10.6%, n = 39), Healthcare
(10.4%, n = 38), Construction (4.9%, n = 18), Transportation/materials
moving/warehouse (4.9%, n = 18), and Food preparation/serving (4.4%, n = 16).
Average knowledge of other employees EGBs, maintenance-related work,
production-related work, and construction-related work at their companies was
“Some”. The majority of participants (56.4%, n = 207) were not required to
perform EGBs as a part of their job tasks.
Survey Design
The final survey contained the refined 39-item EGB Descriptive Norms
Scale, the Ethical Leadership Questionnaire, the Work-Family Culture Scale,
several demographic, work-, and industry-related questions, and several scales
not included in this study. The Ethical Leadership Questionnaire and Work-
Family Culture Scale were selected to provide evidence of convergent and
divergent validity respectively.
Measures and Demographics
Ethical Leadership
. Created by Yukl et al. (2013), the 15-item Ethical
Leadership Questionnaire captures four core components of ethical leadership:
communication of ethical standards, honesty and integrity, concern for others,
and fairness. All items were measured using a 6-point Likert scale (1 = strongly
disagree, 6 = strongly agree). An example item is: “My boss . . insists on doing
what is fair and ethical even when it is not easy.” Cronbach’s alpha from the
44
present study was .97. Ethical leadership is hypothesized to moderately,
positively correlate (r = .3 to .4) with the EGB Descriptive Norms Scale. The
Ethical Leadership Questionnaire can be found in Appendix E.
Work-Family Culture
. Work-family culture was measured using the 20-item
scale developed by Thompson et al. (1999). Cronbach’s alpha from the present
study was .90. This scale assesses three dimensions of work-family culture:
organizational time demands (e.g. “To get ahead at this organization, employees
are expected to work more than 50 hours a week, whether at the workplace or at
home”), managerial support for work-family needs (e.g. “Middle managers and
executives in this organization are sympathetic toward employees’ child care
responsibilities”), and career consequences associated with using work-family
benefits (e.g. “Many employees are resentful when men in this organization take
extended leave to care for newborn or adopted children”). Items were measured
using a 7-point, Likert scale (1 = strongly disagree, 7 = strongly agree) and
higher scores represented a more supportive work-family culture. Work-family
culture was not expected to have a relationship with EGB descriptive norms and
to exhibit a weak correlation (r = .1 to .2). The Work-Family Culture Scale can be
found in Appendix F.
Demographics
. Demographic information was collected including age,
gender, race/ethnicity, employment status, average hours worked/week, tenure,
work position, and industry. One item assessed whether employee green
behaviors are required as part of the participant’s job tasks. One item each
45
assessed how much knowledge the participants have about maintenance-related
work at their companies, production-related work at their companies,
construction-related work at their companies, and other employees’ green
behaviors within the organizations.
Confirmatory Factor Analysis Results
Before running the CFA, data (N = 400) were screened to identify careless
responding, missing data, univariate outliers, multivariate outliers, nonnormality,
and multicollinearity. Any participant who completed the survey in less than half
the median time was flagged for careless responding. Thus, 18 participants were
excluded for completing the survey in seven minutes or less. No item was
missing more than 5 percent data, indicating there was no pattern to the missing
data and the data was missing completely at random (Tabachnick & Fidell,
2013). Item data was imputed using the expectation maximization algorithm after
reverse-coding the reverse coded items. Scale scores were computed using the
imputed data. Using a z-score criterion of p < .001, univariate outliers were
evaluated at the scale level and ten participants were excluded from further
analyses. Multivariate outliers were also evaluated at the scale level and five
participants were removed based on a Mahalanobis distance criteria set at p <
.001. Based on a z-score criterion set at p < .001, several items in each scale
were negatively skewed indicating the need to use robust indices of fit
(Tabachnick & Fidell, 2013). Multicollinearity of EGB descriptive norm scale items
46
and subscales was examined through condition indices and variance proportions.
Belsely, Kuh, and Welsch (1980) propose that criteria for multicollinearity are a
conditioning index greater than 30 combined with at least two variables exhibiting
variance proportions greater than .50. Examining scale items revealed some
large condition indices (>30), but no two variance proportions greater than .50 for
any single condition index, indicating no issue with multicollinearity among scale
items. However, examining subscales revealed one large condition index (>30)
with two variance proportions greater than .50 and one condition index of 25 with
two variance proportions close to .50. This indicated potential multicollinearity
among the five factors the subscales are intended to represent.
Because the content of the 39-item scale was based on the Green Five
Taxonomy, it was expected to mimic the structure of the Green Five Taxonomy
as well. Thus, General Green Performance was expected to predict the five
meta-categories, which were expected to predict the individual scales items. See
Figure 2 for the expected scale structure. To test this structure, a second-order
CFA was run using EQS 6.1 (Bentler, 2006). For the model to run, it was first
necessary to set the disturbances for the Working Sustainably and Taking
Initiative factors to .01 (i.e. constrain them just above zero). An examination of
the item path coefficients revealed items with negative path coefficients as well
as items with positive but insignificant path coefficients. Items 8, 13, 29, 31, and
38 were dropped because their negative path coefficients indicated that they did
not represent the intended construct. Items 17, 19, 24, and 33 exhibited positive
47
but insignificant path coefficients, and the multivariate Wald test suggested
dropping these paths to improve model fit. In response, these items were
dropped one by one and model fit improved with each path dropped. Though not
suggested by the Wald test, Items 3, 14, and 15 exhibited very high residuals
with other items (> .25) and were subsequently dropped as well. The twelve
dropped items were evenly spread among the five first-order factors, thus content
coverage of the construct was maintained. Due to the univariate nonnormality, as
well as the multivariate nonnormality suggested by Mardia’s coefficient (57.5),
the Satorra-Bentler scaled chi-square, robust comparative fit index (CFI), and
robust RMSEA were used to examine model fit (Tabachnick & Fidell, 2013).
Good fit is indicated by a non-significant Satorra-Bentler scaled chi-square, CFI
values greater than .95, and RMSEA values less than .05. However, the chi-
square is affected by large sample sizes. With a large sample, even minimal
differences between the sample and estimated population covariance matrices
can lead to a significant chi-square. To help remove this dependence on sample
size, the relative chi-square is a ratio of chi-square to degrees of freedom and a
3:1 ratio indicates acceptable model fit (Carmines & McIver, 1981). Under the
relative chi-square, the fit is acceptable at 705:321. Thus, the final model fit was
good, Satorra-Bentler scaled X
2
(321, N=367) = 705.3, p < .001, robust CFI =
.930, and robust RMSEA = .057. The final model, including standardized and raw
path coefficients, is shown in Figure 3. All standardized path coefficients are
greater than or equal to .400.
48
The path coefficients between the first-order factors and the second-order
factor reveal that the variance associated with the Working Sustainably, Avoiding
Harm, Influencing Others, and Taking Initiative factors is almost completely
subsumed by the General Green Performance factor (Taub, 2001). This
suggests that model fit might improve after removing these four factors and
allowing the items to load directly onto the General Green Performance factor.
However, when tested, this adjusted model exhibited basically the same fit,
Satorra-Bentler Scaled X
2
(323, N=367) = 723.7, p < .001, robust CFI = .927, and
robust RMSEA = .058, indicating that it did not significantly improve or degrade
the model to remove the four subsumed factors. Thus, it was decided to keep
them in the model for conceptual clarity.
Construct Validity Results
Evidence of construct validity was gathered by correlating the overall scale
score and the individual subscale scores of the EGB Descriptive Norms Scale
with the scale scores of other conceptually similar and conceptually distinct
constructs. Specifically, the relationships between EGB descriptive norms, work-
family culture, and ethical leadership were examined using bivariate correlations.
The reliability of the EGB Descriptive Norms Scale and its subscales was also
examined using Cronbach’s alpha. See Table 4 for these correlations and the
results of the reliability analyses.
Ethical Leadership
49
Ethical leadership was hypothesized to exhibit a moderate positive
correlation with the overall EGB Descriptive Norms Scale. The correlation was
moderate, r = .43, as expected. Similarly to work-family culture, no specific
relationships were hypothesized between ethical leadership and the five EGB
subscales, but they also exhibited moderate to low-moderate correlations with
ethical leadership. They ranged from r = .35 for the Working Sustainably and
Taking Initiative subscales up to r = .46 for the Conserving subscale.
Work-Family Culture
Work-family culture was hypothesized to have a low, positive relationship
with the overall EGB Descriptive Norms Scale. As expected, the correlation was
low and positive, r = .23. No specific relationships were hypothesized among the
EGB subscales and work-family culture. However, they were also weak to
moderate, ranging from r = .13 for the Taking Initiative subscale to .34 for the
Conserving subscale.
50
CHAPTER FOUR
DISCUSSION
General Discussion
Employees’ behaviors are critical to the success of an organization’s goals
and programs (Daily, Bishop, & Govindarajulu, 2009; Fugate, Stank, & Mentzer,
2009). Thus, understanding EGBs and the factors that influence their enactment
are critical to the success of an organization’s environmental goals and
programs. Identified in the environmental psychology literature, descriptive norms
have been found to influence PEBs in both public and private settings (e.g.
Cialdini et al., 1990; Kallgren et al., 2000). However, a lack of measurement tools
severely limited researchers’ ability to examine PEB descriptive norms in the
context of the workplace. With the recent development of the Green Five
Taxonomy (Ones & Dilchert, 2012a), the behavioral content of an EGB
descriptive norms scale was made available, and the development of an EGB
Descriptive Norms Scale was a meaningful next step in expanding the study of
EGBs.
The present study developed an EGB Descriptive Norms Scale based on
the structure of the Green Five Taxonomy. The results of the study supported the
expected scale structure of a second-order General Green Performance factor
and the five first-order factors of Working Sustainably, Avoiding Harm,
Conserving, Influencing Others, and Taking Initiative. However, there is evidence
51
that the five first-order factors are not as distinct as the Green Five Taxonomy
suggests. Additionally, the overall scale as well as the five subscales each
exhibits good reliability, and the discriminant and convergent validity of the
overall scale is supported by its correlations with work-family culture and ethical
leadership. Taken together, these results indicate that the 27-item EGB
Descriptive Norms Scale is a valid and reliable measure of EGB descriptive
norms. See Appendix G for the final 27-item scale.
Structural Evidence of the Employee Green Behavior
Descriptive Norms Scale
The results of the CFA support the proposed scale structure. Specifically,
multiple fit indices support overall General Green Performance being comprised
of the five factors of Working Sustainably, Avoiding Harm, Conserving,
Influencing Others, and Taking Initiative. However, the data indicates that the five
first-order factors may not be as distinct as the Green Five Taxonomy suggests.
The Working Sustainably, Avoiding Harm, Influencing Others, and Taking
Initiative factors exhibited such high standardized path coefficients (>.95) that
their variance was almost completely subsumed by the General Green
Performance factor (Taub, 2001). Only the Conserving factor captured some
unique variance. Yet, removing these four factors did not improve or degrade the
fit of the model, and thus were kept for conceptual clarity.
One possible explanation for the lack of distinct variance among the meta-
categories of EGBs is the presence of a halo effect. The halo effect is a
52
respondent’s inability to discriminate among concepts or attributes. This causes
the correlations among the affected items or scales to increase (Leuthesser,
Kohli, & Harich, 1995) or causes the variance among categories to be low
(Cooper, 1981). One way that the halo effect can be detected is by examining the
factor structure of a group of categories. If the structure is dominated by a
general factor that accounts for most of the variance in the categories (Cooper,
1981; Kafry, Jacobs, & Zedeck, 1979), as the structure in this study was, it
indicates the presence of halo. While there are several sources that cause the
halo effect (Cooper, 1981), it is likely that within this study the halo effect was
due to limited knowledge of others’ EGBs. On average, respondents indicated
that they had “Some” knowledge of others’ EGBs. Kozlowski, Kirsch, & Chao
(1986) found that when respondents have limited knowledge on a topic they rely
more heavily on their conceptual similarity schemata. This encourages
respondents to perceive co-occurrence among categories even when it does not
exist. Thus, instead of basing responses completely on their knowledge of EGBs,
respondents were likely influenced by the conceptual similarities of the items’
content and their beliefs about EGBs at work. These perceptions of similarity
then reduced the variance among the first-order factors.
The presence of a halo effect could affect the scale in several ways. First,
it may not be useful to examine the subscales of the EGB Descriptive Norms
Scale. If respondents lack knowledge of a meta-category of EGBs, then their
responses may not be specific enough to provide an accurate picture of norms
53
related to that meta-category. However, to help ensure that the breadth of norm
forming behaviors is captured, it is still meaningful to include items that tap into
each of the meta-categories of the Green Five Taxonomy. Second, there is the
possibility that the presence of normative behaviors is being overestimated. If
respondents are relying on their conceptual similarity schemata, then their
responses may “capture” normative behavior that does not actually occur in the
workplace. This could reduce the sensitivity of the scale to variations in EGBs.
However, if the scale is used to capture overall perceptions of EGB descriptive
norms in the workplace, then this is less of an issue. Third, it suggests that
certain meta-categories of EGBs are not as prevalent in the workplace and may
not be as relevant to EGB descriptive norms. When creating their Green Five
Taxonomy, Ones and Dilchert (2012a) noted that the majority of critical incidents
came from the Conserving meta-category of EGBs. As the only factor accounting
for unique variance, it is likely that these common behaviors are most frequently
seen and, thus, most likely to elicit knowledge-based responses.
Evidence of Construct Validity
The construct validity of the EGB Descriptive Norms Scale was supported
by interscale correlations. Evidence of convergent validity was provided by the
moderate, positive relationship between the overall EGB Descriptive Norms
Scale and the Ethical Leadership Questionnaire. This relationship supports the
idea that ethical leaders, through their concern for others and the consequences
54
of their actions, would likely support sustainability in their place of work
(Kalshoven et al., 2011). However, as only one of many social issues that an
ethical leader could choose to focus on, it may not be a high priority for all ethical
leaders (Yukl et al., 2013).
Similarly, the low, positive relationship between EGB descriptive norms
and work-family culture provides evidence of discriminant validity. Work-family
culture addresses aspects of the workplace that impact how employees are able
to handle their work and family lives (Thompson et al., 1999), not aspects of the
workplace that affect whether EGBs are enacted. The fact that the relationship is
positive is likely due to the fact that just as managerial support is a key
component of positive work-family culture, it is also a predictor of employee
willingness to enact environmental behaviors (Ramus, 2001). A supervisor who is
supportive in one area may also be supportive in the other.
Taken together, these results provide initial support for the construct
validity of the EGB Descriptive Norms Scale and its use as an appropriate
measurement tool. As expected, EGB descriptive norms were found to have a
moderate, positive relationship with ethical leadership and a weak, positive
relationship with work-family culture. Though no explicit predictions were made,
the correlations between each of the five subscales and the two outside
constructs were similar in magnitude to the correlations between the overall scale
and the two outside constructs.
55
Implications
A well-designed, validated EGB Descriptive Norms Scale has application
in both organizational and research settings. Organizations benefit when their
environmental goals and their employees’ behaviors are consistent (Ones &
Dilchert, 2012a). Descriptive norms are a representation of how commonly
others engage in the norm-forming behaviors. The EGB Descriptive Norms Scale
is capturing employee perceptions of how often others engage in EGBs. So, by
using the EGB Descriptive Norms Scale, organizations can get a sense for the
perceived prevalence of EGBs, and whether there are strong descriptive norms
to perform EGBs. Companies who find that they have weak EGB descriptive
norms will have pinpointed an issue they would need to resolve if they are trying
to implement company-wide environmental initiatives. Companies who find that
they have strong EGB descriptive norms could be more confident in the success
of their programs, or at least know that weak descriptive norms are not the cause
of program-related issues they might be experiencing.
In terms of research, EGBs are a new, applied topic and well-developed,
validated scales can help our understanding of this topic move forward.
Organizations are interested in reducing their environmental impact (Darnall,
Henriques, & Sadorsky, 2008), and descriptive norms have been shown to
influence environmental behaviors outside the workplace (Cialdini et al., 1990;
Schultz et al., 2007). However, research specifically on EGBs is quite new. By
utilizing the Green Five Taxonomy as its foundation, the EGB Descriptive Norms
56
Scale better captures the breadth of EGB descriptive norms present in the
workplace than do earlier PEB descriptive norms scales. Being able to measure
this construct will allow researchers to begin exploring the relationship between
EGB descriptive norms and meaningful organizational outcomes, such as
company reputation and long-term profitability. By developing strong EGB norms,
a company is more likely to be seen as an environmentally-friendly company,
and having a reputation as a socially responsible company has been linked to
attracting high quality job applicants (Greening & Turban, 2000). Similarly,
engaging in EGBs has been linked to increased profitability through reduced
energy use and waste production, process intensification, and development of
environmentally-friendly products (Bansal & Roth, 2000). Using a well-developed
scale to explore these relationships allows for consistency of measurement and
comparison across studies. This provides a clearer sense of the phenomenon
than using slightly different scales for each study, and it provides a point of
reference for additional scale refinement.
Future Research
It would be beneficial to explore the nomological network of EGB
descriptive norms beyond the two constructs included in this study. Additional
evidence of convergent and discriminant validity would clarify the construct
represented by the EGB Descriptive Norms Scale, and further support its use in
both research and applied settings. For example, it is likely that pro-
57
environmental concern would be positively related to EGB descriptive norms, and
would provide evidence of convergent validity. Research has shown that job
choice is influenced by person-organization fit, which is based on the alignment
of a job applicant’s values with the values of the organization (Cable & Judge,
1996; Judge & Bretz, 1992). Thus, it is likely that applicants with a high pro-
environmental orientation would choose to work for an organization with strong
EGB descriptive norms. Other possible constructs include aspects of the
workplace, such as abusive leadership, and/or individual differences that
influence job choice or job behaviors, such as the Big Five. Furthermore,
examining the relationship between the EGB Descriptive Norms Scale and other
available norms scales, descriptive or otherwise, would strengthen evidence of
discriminant validity.
Examination of scale generalizability would also be meaningful. In the
current study, the sample was considered generalizable because of the
demographic diversity (e.g. gender, age, ethnicity) as well as the diversity in
industry and job position. However, certain behaviors represented in the Green
Five Taxonomy are more likely to be witnessed or known about in certain
industries or job positions. For example, many of the behaviors included within
the Avoiding Harm subscale reference a big-picture understanding of what a
respondent’s company does. Thus, it would be more likely for an employee in a
managerial position or higher to know about such behaviors, and have a different
concept of EGB descriptive norms than an hourly or non-management worker.
58
Similarly, items across subscales could be differentially meaningful depending on
the industry. Adjusting workplace processes or product choice could potentially
be more salient in an industry where a lot of physical material is moved (e.g.
production or construction) compared to an industry where most of the work is
electronic (e.g. education or administration). Thus, examining the results of the
model when applied to specific job positions or industries would provide a better
understanding of EGB descriptive norms and how they manifest.
It would also be informative to use this new scale to expand the research
on EGBs by identifying antecedents and outcomes related to EGB descriptive
norms. Though included for the purpose of construct validity, the relationship
between EGB descriptive norms and ethical leadership is an interesting finding.
Exploring this relationship would clarify how leadership and EGBs and their
descriptive norms influence each another (Graves, Sarkis, & Zhu, 2013;
Robertson & Barling, 2013). Meaningful outcomes could include performance of
EGBs (Daily et al., 2009), cost-savings related to strong EGB descriptive norms,
and public perceptions of the organization (Rindova, Williamson, & Petkova,
2005). Furthermore, with the development of additional measurement tools, the
relationship between descriptive norms and injunctive norms could be explored in
the workplace just as it has been explored in private and other public contexts.
Limitations
One possible limitation is the existence of common method biases. The
59
source of these biases, common method variance, can influence study results by
either inflating or deflating observe relationships (Podsakoff, MacKenzie, Lee, &
Podsakoff, 2003). Within this study, certain methodological aspects likely helped
reduce the possibility of common method biases. First, scales with varying
anchors and endpoints likely helped with anchoring effects and response
consistency due to scale format (Podsakoff et al., 2003; Tourangeau, Rips, &
Rasinski, 2000). Second, the anonymity of responses and asking about the
behaviors of others within the EGB Descriptive Norms Scale should reduce
evaluation apprehension and encourage more honest responses (Podsakoff et
al., 2003). This reduces biases such as social desirability and leniency. Thus,
when collecting data, the convenience of a single method of data collection must
be balanced with consideration of reducing the effects of common method
biases.
A second limitation was measuring a group level phenomenon
(organizational norms) at the individual level. Interpreting aggregated individual
data at the group level can lead to the atomistic fallacy, wherein inferences about
group level variability are based on variability at a lower level of aggregation
(Diez Roux, 2002). As a result, use of this new scale is limited to perceptions of
EGB descriptive norms at the individual level, rather than measurement of EGB
descriptive norms at the group or organization level. Future research would
benefit from adjusting measurement of EGB descriptive norms to the group and
organization level, and examining its relationship to other constructs using multi-
60
level modeling.
A third limitation is the possibility of a halo effect influencing item
response. Since all three scales included in this study asked about the behaviors
of others in the workplace, a lack of knowledge about these behaviors likely lead
to the use of holistic impressions to inform item response (Cooper, 1981).
Though no such data was collected on ethical leadership or work-family culture,
four of the included demographic items assessed respondents’ knowledge of
item content included in the EGB Descriptive Norms Scale. On average,
participants indicated that they had some knowledge of the behavioral content.
While there is no clear cut-off for the minimum knowledge needed to infer typical
behavior, a lack of knowledge would make realistic detail unavailable and would
likely be replaced by the respondent’s conceptual similarity schemata and beliefs
about EGBs.
A fourth and last limitation is the possibility of volunteer bias in the study
sample. All participants were people who agreed to be a part of Qualtrics’
research panels and who volunteered to take the survey. Thus, there were two
opportunities for volunteer bias to affect the demographic, attitudinal, and
behavioral composition of the study sample. However, analysis of sample
demographics indicates that the sample is relatively diverse and, thus,
generalizable. Of course, application of the scale to additional samples would
help further support its generalizability. Attitudinal differences, due to
volunteering based on interest in the study topic, was likely mitigated through the
61
small incentive for participation. Additionally, the 2000 World Values Survey
found that 85 percent of U.S. respondents endorsed a perspective favoring a
more equal relationship with nature (as cited in Leiserowitz, Kates, & Parris,
2006). Thus, those interested in environmentalism should be highly represented
to mirror national trends. Furthermore, the behaviors assessed in the survey
were the behaviors of others in the workplace, not of the self-selected
respondents.
Conclusion
To better understand EGBs and the contextual factors that influence them,
an EGB Descriptive Norms Scale was developed based on the Green Five
Taxonomy of EGBs. The results of this study provide strong support for this
scale’s internal reliability and construct validity. Additionally, although
respondents’ limited knowledge of others EGBs may have created reduced
discrimination among categories of EGBs, scale structure was found to mimic the
structure of the Green Five Taxonomy. Using this new tool, organizations can
better understand the state of their EGB descriptive norms and researchers can
use it to expand our understanding of the EGBs. Overall, it will contribute to the
growth of EGBs as a new and applied area of interest.
62
APPENDIX A
INITIAL EMPLOYEE GREEN BEHAVIOR DESCRIPTIVE NORMS SCALE
63
INITIAL EMPLOYEE GREEN BEHAVIOR DESCRIPTIVE NORMS SCALE
Developed by Jacqueline C. McConnaughy.
Directions: Below are a series of statements about behaviors that can occur in
the workplace. Read the question prompt and, for each item, use the response
scale to indicate your experience of others in your workplace. Fill in the circle that
corresponds to your preferred response option.
Likert-Style Rating Scale: (1) Strongly Disagree; (6) Strongly Agree
Frequency Response Rating Scale: (1) Never; (6) Always
How often do others at your organization (e.g. co-workers, supervisors, leaders)
do the following while at work?
1. When there is a choice, choose products that are better for the
environment
2. Fix maintenance issues to prevent unintended pollution and waste of
resources
3. Utilize single-use, disposable products, such as paper towels (R)
4. Suggest ways for other employees to act in a more environmentally-
friendly manner
5. Propose a new environmentally-friendly program for the company
6. Maximize the life span of office equipment through repair and
maintenance
7. Design new, environmentally-friendly products
8. Voice concerns that acting pro-environmentally could hurt the company
(R)
9. Discuss environmentally-related topics with other employees
10. Push the company's leaders to take a stronger position on environmental
issues
11. Utilize new technologies that benefit the environment
12. Use supplies in new ways
13. Propose delaying an environmentally-related program for business
reasons (R)
14. Buy company supplies without thought for environmental impact (R)
15. Not prioritize actions that would benefit the environment (R)
16. Reduce water consumption by turning off faucets when not in use
17. Use inefficient work processes that waste natural resources (R)
18. Monitor the environmental impact of workplace processes
19. Improperly handle hazardous materials (R)
20. Provide environmentally-related literature to other employees
21. Give praise to other employees for their environmentally-friendly behavior
64
22. Reject a desirable project because it would be bad for the environment
23. Change work processes to reduce negative impacts on the environment
24. Neglect to clean up after an environmentally-harmful accident or event (R)
25. Save extra supplies or materials for a future project
26. Help implement new policies that reduce the company's impact on the
environment
27. Decrease energy consumption by turning off equipment when not in use
28. Choose a less convenient commute because it helps the environment
29. Recycle paper, plastic, metal cans, etc.
30. Knowingly cause unnecessary damage to the environment through work-
related decisions (R)
31. Develop new work processes that use fewer natural resources
32. Design a new product that contains harmful components (R)
33. Throw recyclable materials into trash cans (R)
34. Knowingly choose technologies that are more harmful to the environment
(R)
35. Through work, participate in projects that improve the local environment
36. Tease other employees for behaviors that benefit the environment (R)
37. Monitor workplace processes for potential sources of unintended pollution
38. Give materials a new use or purpose instead of throwing them away
39. Tell other employees that environmentally-friendly behaviors are
ineffective (R)
40. Reduce waste by reusing items such as water bottles, paper, plastic, etc.
65
APPENDIX B
INITIAL EMPLOYEE GREEN BEHAVIOR NORMS SCALE ITEMS
ARRANGED BY META-CATEGORY
66
INITIAL EMPLOYEE GREEN BEHAVIOR NORMS SCALE ITEMS
ARRANGED BY META-CATEGORY
Working Sustainably:
1. When there is a choice, choose products that are better for the environment
7. Design new, environmentally-friendly products
11. Utilize new technologies that benefit the environment
14. Buy company supplies without thought for environmental impact (R)
17. Use inefficient work processes that waste natural resources (R)
23. Change work processes to reduce negative impacts on the environment
31. Develop new work processes that use fewer natural resources
32. Design a new product that contains harmful components (R)
34. Knowingly choose technologies that are more harmful to the environment (R)
Avoiding Harm:
2. Fix maintenance issues to prevent unintended pollution and waste of
resources
18. Monitor the environmental impact of workplace processes
19. Improperly handle hazardous materials (R)
24. Neglect to clean up after an environmentally-harmful accident or event (R)
30. Knowingly cause unnecessary damage to the environment through work-
related decisions (R)
35. Through work, participate in projects that improve the local environment
37. Monitor workplace processes for potential sources of unintended pollution
Conserving:
3. Utilize single-use, disposable products, such as paper towels (R)
6. Maximize the life span of office equipment through repair and maintenance
12. Use supplies in new ways
16. Reduce water consumption by turning off faucets when not in use
25. Save extra supplies or materials for a future project
27. Decrease energy consumption by turning off equipment when not in use
29. Recycle paper, plastic, metal cans, etc.
33. Throw recyclable materials into trash cans (R)
38. Give materials a new use or purpose instead of throwing them away
40. Reduce waste by reusing items such as water bottles, paper, plastic, etc.
Influencing Others:
4. Suggest ways for other employees to act in a more environmentally-friendly
manner
9. Discuss environmentally-related topics with other employees
20. Provide environmentally-related literature to other employees
21. Give praise to other employees for their environmentally-friendly behavior
67
36. Tease other employees for behaviors that benefit the environment (R)
39. Tell other employees that environmentally-friendly behaviors are ineffective
(R)
Taking Initiative:
5. Propose a new environmentally-friendly program for the company
8. Voice concerns that acting pro-environmentally could hurt the company (R)
10. Push the company's leaders to take a stronger position on environmental
issues
13. Propose delaying an environmentally-related program for business reasons
(R)
15. Not prioritize actions that would benefit the environment (R)
22. Reject a desirable project because it would be bad for the environment
26. Help implement new policies that reduce the company's impact on the
environment
28. Choose a less convenient commute because it helps the environment
68
APPENDIX C
RETRANSLATION TASK INSTRUCTIONS AND CATEGORIES
69
RETRANSLATION TASK INSTRUCTIONS AND CATEGORIES
Background:
In the workplace, environmentally-friendly behaviors are considered
employee green behaviors (EGBs). Ones and Dilchert (2012a) define EGBs as
“scalable actions and behaviors that employees engage in that are linked with
and contribute to or detract from environmental sustainability” (p. 87). These
behaviors can be performed as part of an employee’s job duties, outside of an
employee’s job duties as organizational citizenship behaviors, or as
counterproductive work behaviors that actually detract from the organization’s
environmental performance. As “scalable” actions, they can vary in terms of how
frequently or proficiently employees perform them. This in turn allows each
employee’s contribution to the environmental performance of the organization to
be quantified. Ones and Dilchert (2012a) further developed a content-based,
three-tier Green Five Taxonomy of EGBs. The first tier consists of General Green
Performance, followed by the five meta-categories of Working Sustainably,
Avoiding Harm, Conserving, Influencing Others, and Taking Initiative in the
second tier. The third tier consists of a further breakdown into 16 categories.
Translation Task:
In the other word document attached to your email, you will find a list that
contains the 16 categories of the Green Five Taxonomy of EGBs along with their
definitions. They are organized by meta-category to help with understanding and
interpretation. They have also been tagged with an abbreviated code (e.g. WS1
or C3). Please read the category definitions to familiarize yourself with the
content of each category.
On the following pages you will find a list of all 40 items that are currently
included in the norms scale I am developing for my thesis. They are split into two
sets of 20 items. Each item represents an employee green behavior that could be
observed in the workplace. Next to each item you will see a box. Please read the
item and then place the abbreviated code of the category you think best fits the
item into the item’s box. If you don’t think the item fits into any category, please
put an NA in the box. I would also appreciate any feedback you might have about
unclear items, items that don’t fit neatly into a category, or anything else you
noticed while coding the items.
70
The 16 categories of employee green behaviors, organized by meta-category
Working Sustainably
: behaviors that help work processes and products be
more sustainable
WS1
Choosing Responsible Alternatives: behaviors wherein an employee
chooses the work product or process option that is more environmentally
friendly
WS2
Changing How Work is Done: behaviors wherein work processes are
changed to become more sustainable
WS3
Creating Sustainable Products and Processes: behaviors wherein new
products or processes are created that are more environmentally-friendly
WS4
Embracing Innovation for Sustainability: behaviors where in new, more
sustainable technology is adopted at work
Avoiding Harm
: behaviors that can either harm the earth and cause increasing
damage, or can enhance the earth, making its ecosystems healthier
AH1
Polluting/Preventing Pollution: behaviors that cause or prevent pollution
AH2
Monitoring Environmental Impact: behaviors wherein work activities are
monitored to assess and understand how they are affecting the
environment
AH3
Strengthening Ecosystems: behaviors that help protect or repair
ecosystems from the effects of industry and business
Conserving
: behaviors intended to help preserve resources and reduce waste
C1
Reducing Use: behaviors that prevent the unnecessary use of new
materials
C2
Reusing: behaviors wherein materials are used multiple times for the
same purpose
C3
Repurposing: behaviors wherein materials are used multiple times for new
purposes
C4
Recycling: behaviors wherein materials are recycled (aka end up at a
recycling center)
71
Influencing Others
: social behaviors used to influence others to engage in
environmental behaviors
IO1
Encouraging and Supporting Others: behaviors that bolster and
encourage other’s employee green behaviors
IO2
Educating and Training for Sustainability: behaviors that help others build
their knowledge about environmentalism
Taking Initiative
: behaviors that involve stepping outside the box, taking a risk,
and encouraging environmentally-related change
TI1
Putting Environmental Interests First: behaviors that help the environment
at some personal cost to the individual
TI2
Initiating Programs and Policies: pushing for new programs and policies at
work that would benefit the environment
TI3
Lobbying and Activism: behaviors that involve fighting for environmental
causes
72
APPENDIX D
REFINED EMPLOYEE GREEN BEHAVIOR DESCRIPTIVE NORMS SCALE
73
REFINED EMPLOYEE GREEN BEHAVIOR DESCRIPTIVE NORMS SCALE
Developed by Jacqueline C. McConnaughy.
*Bolded items were revised or replaced from the initial scale*
Directions: Below are a series of statements about behaviors that can occur in
the workplace. Read the question prompt and, for each item, use the response
scale to indicate your experience of others in your workplace. Fill in the circle that
corresponds to your preferred response option.
Rating Scale: (1) Strongly Disagree; (6) Strongly Agree
Others at my organization (e.g. co-workers, supervisors, leaders):
1. When there is a choice, choose products that are better for the
environment
2. Fix maintenance issues to prevent unintended pollution and waste of
resources
3. Throw disposable items away rather than reuse them (R)
4. Suggest ways for other employees to act in a more environmentally-
friendly manner
5. Propose a new environmentally-friendly program for the company
6. Save extra materials from one project to supply a different project
7. Design new, environmentally-friendly products
8. Discourage the environmentally-friendly behavior of other
employees (R)
9. Discuss environmentally-related topics with other employees
10. Push the company to stand behind an environmental cause
11. Utilize new technologies that benefit the environment
12. Use supplies in new ways
13. Propose a new company policy without addressing the
environmental impact of the policy (R)
14. Buy company supplies without thought for environmental impact (R)
15. Not prioritize actions that would benefit the environment (R)
16. Reduce water consumption by turning off faucets when not in use
17. Use inefficient work processes that waste natural resources (R)
18. Monitor the environmental impact of workplace processes
19. Improperly dispose of trash and waste materials (R)
20. Provide environmentally-related literature to other employees
21. Give praise to other employees for their environmentally-friendly behavior
22. Use extra time or energy to perform an environmentally-friendly
behavior over an environmentally-harmful behavior
23. Change work processes to reduce negative impacts on the environment
74
24. Neglect to clean up after an environmentally-harmful accident or event (R)
25. Help implement new policies that reduce the company's impact on the
environment
26. Decrease energy consumption by turning off equipment when not in use
27. Choose a less convenient commute because it helps the environment
28. Recycle paper, plastic, metal cans, etc.
29. Knowingly cause unnecessary damage to the environment through work-
related decisions (R)
30. Develop new work processes that use fewer natural resources
31. Design a new product that contains harmful components (R)
32. Throw recyclable materials into trash cans (R)
33. Knowingly choose technologies that are more harmful to the environment
(R)
34. Incorporate environmental protection into project ideas and
development
35. Lobby company leaders to donate to an environmental-friendly
nonprofit
36. Monitor workplace processes for potential sources of unintended pollution
37. Give materials a new use or purpose instead of throwing them away
38. Explain why environmentally-friendly behaviors may not be as
effective as people think (R)
39. Reduce waste by reusing items such as water bottles, paper, plastic, etc.
75
APPENDIX E
ETHICAL LEADERSHIP QUESTIONNAIRE
76
ETHICAL LEADERSHIP QUESTIONNAIRE
Yukl, G., Mahsud, R., Hassan, S., & Prussia, G. E. (2013). An improved measure
of ethical leadership. Journal of Leadership & Organizational Studies,
20(1), 38-48. doi: 10.1177/1548051811429352
Instructions: This questionnaire is designed to study the relevance of ethics to
effective leadership. The term “unit” refers to the team, department, division, or
company for which your boss is the formal leader, and the term “members” refers
to the people in the unit who report directly to your boss. Please indicate how
well each of the following statements describes your current boss by selecting
one of the following response choices. Write the number of the choice on the line
provided. Leave the item blank if you do not know the answer.
Rating Scale: (1) Strongly Disagree; (6) Strongly Agree
My boss:
1. _ Shows a strong concern for ethical and moral values.
2. _ Communicates clear ethical standards for members.
3. _ Sets an example of ethical behavior in his/her decisions and actions.
4. _ Is honest and can be trusted to tell the truth.
5. _ Keeps his/her actions consistent with his/her stated values (“walks the talk”).
6. _ Is fair and unbiased when assigning tasks to members.
7. _ Can be trusted to carry out promises and commitments.
8. _ Insists on doing what is fair and ethical even when it is not easy.
9. _ Acknowledges mistakes and takes responsibility for them.
10. _ Regards honesty and integrity as important personal values.
11. _ Sets an example of dedication and self-sacrifice for the organization.
12. _ Opposes the use of unethical practices to increase performance.
13. _ Is fair and objective when evaluating member performance and providing
rewards.
14. _ Puts the needs of others above his/her own self interest.
15. _ Holds members accountable for using ethical practices in their work.
77
APPENDIX F
WORK-FAMILY CULTURE SCALE
78
WORK-FAMILY CULTURE SCALE
Thompson, C. A., Beauvais, L. L., & Lyness, K. S. (1999). When work-family
benefits are not enough: The influence of work-family culture on benefit
utilization, organizational attachment, and work-family conflict. Journal of
Vocational Behavior, 54, 392-415.
Instructions: Following are several statements about how organizations handle
the work and family needs of employees in less formal ways. Please read each
item carefully and then indicate the extent to which you agree or disagree with
each statement as it pertains to working in your organization.
Rating Scale: (1) Strongly Disagree; (5) Strongly Agree
1. In my work organization employees can easily balance their work and family
lives.
2. In the event of a conflict, managers are understanding when employees
have to put their family first.
3. In my work organization it is generally okay to talk about one’s family at
work.
4. Employees are often expected to take work home at night and/or on
weekends. (R)
5. Higher management in my work organization encourages supervisors to be
sensitive to employees’ family and personal concerns.
6. Employees are regularly expected to put their jobs before their families. (R)
7. To turn down a promotion or transfer for family-related reasons will seriously
hurt one’s career progress in my work organization. (R)
8. In general, managers in my work organization are quite accommodating of
family-related needs.
9. Many employees are resentful when women in my work organization take
extended leaves to care for newborn or adopted children. (R)
10. To get ahead at my work organization, employees are expected to work
more than 50 hours a week, whether at the workplace or at home. (R)
11. To be viewed favorably by top management, employees in my work
organization must constantly put their jobs ahead of their families or personal
lives. (R)
12. In my work organization employees who participate in available work–family
programs (e.g., job sharing, part-time work) are viewed as less serious about
their careers than those who do not participate in these programs. (R)
13. Many employees are resentful when men in my work organization take
79
extended leaves to care for newborn or adopted children. (R)
14. In my work organization it is very hard to leave during the workday to take
care of personal or family matters. (R)
15. My work organization encourages employees to set limits on where work
stops and home life begins
16. Middle managers and executives in my work organization are sympathetic
toward employees’ child care responsibilities.
17. My work organization is supportive of employees who want to switch to less
demanding jobs for family reasons.
18. Middle managers and executives in my work organization are sympathetic
toward employees’ elder care responsibilities.
19. In my work organization employees who use flextime are less likely to
advance their careers than those who do not use flextime. (R)
20. In my work organization employees are encouraged to strike a balance
between their work and family lives.
80
APPENDIX G
FINAL, 27-ITEM EMPLOYEE GREEN BEHAVIOR
DESCRIPTIVE NORMS SCALE
81
FINAL, 27-ITEM EMPLOYEE GREEN BEHAVIOR
DESCRIPTIVE NORMS SCALE
Developed by Jacqueline C. McConnaughy.
Directions: Below are a series of statements about behaviors that can occur in
the workplace. Read the question prompt and, for each item, use the response
scale to indicate your experience of others in your workplace. Fill in the circle that
corresponds to your preferred response option.
Rating Scale: (1) Strongly Disagree; (6) Strongly Agree
Others at my organization (e.g. co-workers, supervisors, leaders):
1. When there is a choice, choose products that are better for the environment
2. Fix maintenance issues to prevent unintended pollution and waste of
resources
3. Suggest ways for other employees to act in a more environmentally-friendly
manner
4. Propose a new environmentally-friendly program for the company
5. Save extra materials from one project to supply a different project
6. Design new, environmentally-friendly products
7. Discuss environmentally-related topics with other employees
8. Push the company to stand behind an environmental cause
9. Utilize new technologies that benefit the environment
10. Use supplies in new ways
11. Reduce water consumption by turning off faucets when not in use
12. Monitor the environmental impact of workplace processes
13. Provide environmentally-related literature to other employees
14. Give praise to other employees for their environmentally-friendly behavior
15. Use extra time or energy to perform an environmentally-friendly behavior
over an environmentally-harmful behavior
16. Change work processes to reduce negative impacts on the environment
17. Help implement new policies that reduce the company's impact on the
environment
18. Decrease energy consumption by turning off equipment when not in use
19. Choose a less convenient commute because it helps the environment
20. Recycle paper, plastic, metal cans, etc.
21. Develop new work processes that use fewer natural resources
22. Throw recyclable materials into trash cans (R)
23. Incorporate environmental protection into project ideas and development
24. Lobby company leaders to donate to an environmental-friendly nonprofit
25. Monitor workplace processes for potential sources of unintended pollution
82
26. Give materials a new use or purpose instead of throwing them away
27. Reduce waste by reusing items such as water bottles, paper, plastic, etc.
APPENDIX H
TABLES
83
Table 1
Pilot Data Reliability Analysis for the Subscales of the EGB Descriptive
Norms Scale
Cronbach's α
Subscales Likert Scale Frequency Scale
Working Sustainably .82 (.82) .71 (.72)
Avoiding Harm .73 (.73) .52 (.76)
Conserving .77 (.80) .78 (.82)
Influencing Others .74 (.86) .61 (.83)
Taking Initiative .69 (.83) .47 (.76)
Note. Cronbach's alpha includes all subscale items. Values in parentheses
indicate the improved Cronbach's alpha after removing problem items.
84
Table 2
Initial Scale Items with Hit Rates from Retranslation Task
Item
Hit Rate
1. When there is a choice, choose products that are better for the
environment
67%
2. Fix maintenance issues to prevent unintended pollution and
waste of resources
67%
3. Utilize single-use, disposable products, such as paper towels
33%
4. Suggest ways for other employees to act in a more
environmentally-friendly manner
83%
5. Propose a new environmentally-friendly program for the company
67%
6. Maximize the life span of office equipment through repair and
maintenance
0%
7. Design new, environmentally-friendly products
83%
8. Voice concerns that acting pro-environmentally could hurt the
company
17%
9. Discuss environmentally-related topics with other employees
67%
10. Push the company's leaders to take a stronger position on
environmental issues
17%
11. Utilize new technologies that benefit the environment
83%
12. Use supplies in new ways
83%
13. Propose delaying an environmentally-related program for
business reasons
17%
14. Buy company supplies without thought for environmental impact
17%
15. Not prioritize actions that would benefit the environment
17%
16. Reduce water consumption by turning off faucets when not in
use
100%
17. Use inefficient work processes that waste natural resources
17%
18. Monitor the environmental impact of workplace processes
100%
19. Improperly handle hazardous materials
67%
20. Provide environmentally-related literature to other employees
83%
21. Give praise to other employees for their environmentally-friendly
behavior
83%
22. Reject a desirable project because it would be bad for the
environment
33%
23. Change work processes to reduce negative impacts on the
environment
67%
24. Neglect to clean up after an environmentally-harmful accident or
event
0%
85
25. Save extra supplies or materials for a future project
33%
26. Help implement new policies that reduce the company's impact
on the environment
100%
27. Decrease energy consumption by turning off equipment when
not in use
83%
28. Choose a less convenient commute because it helps the
environment
50%
29. Recycle paper, plastic, metal cans, etc.
100%
30. Knowingly cause unnecessary damage to the environment
through work-related decisions
17%
31. Develop new work processes that use fewer natural resources
83%
32. Design a new product that contains harmful components
17%
33. Throw recyclable materials into trash cans
50%
34. Knowingly choose technologies that are more harmful to the
environment
33%
35. Through work, participate in projects that improve the local
environment
17%
36. Tease other employees for behaviors that benefit the
environment
67%
37. Monitor workplace processes for potential sources of
unintended pollution
100%
38. Give materials a new use or purpose instead of throwing them
away
83%
39. Tell other employees that environmentally-friendly behaviors are
ineffective
33%
40. Reduce waste by reusing items such as water bottles, paper,
plastic, etc.
83%
Note. Hit rate threshold set at 67% (4 out of 6 subject matter experts correctly
categorizing the item).
86
Table 3
Final Sample Demographic Characteristics
Characteristic
n % X (SD)
Range
Gender
Female
189 51.5% - -
Male
178 48.5% - -
Age
- -
45.0
(11.8)
20-72
Race/Ethnicity
African-American
31* 8.4%* - -
Asian-American
12* 3.3%* - -
Bi-racial/Multi-racial
3** 0.8%* - -
Hispanic/Latino
28* 7.6%* - -
Native American
4** 1.1%* - -
White
289 78.7% - -
Job Position
Non-
management/Hourly
97* 26.4% - -
Non-
management/Salaried
39* 10.6% - -
Entry-level manager
27* 7.4%* - -
Middle management
71* 19.3% - -
Top level executive
21* 5.7%* - -
Professional
77* 19.3% - -
Self-employed
31* 8.4%* - -
Other
4** 1.1%* - -
Industry
Office/Admin. support
76* 10.7% - -
Transportation/Warehouse
18* 4.9%* - -
Sales and related
55* 15.0% - -
Food prep./Serving
16* 4.4%* - -
Healthcare
38* 10.4% - -
Production
39* 10.6% - -
Education/Training
41* 11.2% - -
Construction
18* 4.9%* - -
Other
66* 18.0% - -
Avg. hours worked/week
- - 42.4 (7.1) 20-84
Tenure
87
Years
- - 10.2 (9.0) 1-57
Months
- - 4.1 (3.2)* 0-11
Job Requires EGBs
Yes
131 35.7% - -
No
207 56.4% - -
Don't know
29* 7.9%0 - -
Knowledge: EGBs
- - 3.2 (1.1) 1-5
Knowledge: maintenance
- - 3.1 (1.1) 1-5
Knowledge: production
- - 3.3 (1.2) 1-5
Knowledge: construction
- - 2.8 (1.3) 1-5
Note. Demographic knowledge items are measured on a Likert scale
from 1 (none at all) to 5 (a great deal).
88
Table 4
Scale Score Bivariate Correlations and Internal Reliability Analyses
Scale
EGB Desc.
Norms
WS
Subscale
AH
Subscale
C
Subscale
IO
Subscale
TI
Subscale WFC EL
α
EGB Desc. Norms
- .96
WS Subscale
.94** - .88
AH Subscale
.93** .86** - .86
C Subscale
.84** .69** .72** - .86
IO Subscale
.91** .85** .83** .64** - .89
TI Subscale
.94** .90** .87** .65** .87** - .90
WFC
.23** .16** .20** .34** .17** .13** - .90
EL
.43** .35** .41** .46** .38** .35** .61** - .97
Note. WS = Working Sustainably. AH = Avoiding Harm. C = Conserving. IO = Influencing Others.
TI = Taking Initiative. WFC = Work-Family Culture. EL = Ethical Leadership
** p < .01
89
APPENDIX I
FIGURES
90
FIGURES
Figure 1. The 2
nd
and 3
rd
tiers of the Green Five Taxonomy of EGBs. Adapted
from Ones and Dilchert (2012a).
• Preventing
Pollution
• Monitoring
environmental
impact
• Strengthening
ecosystems
• Reducing use
• Reusing
• Repurposing
• Recycling
• Encouraging
and supporting
others
• Educating and
training for
sustainability
• Putting
environmental
interests first
• Initiating
programs and
policies
• Lobbying and
activism
• Changing how
work is done
• Choosing
responsible
alternatives
• Creating
sustainable
products and
processes
• Embracing
innovation for
sustainability
W
o
rki
n
g
Sustainably
Av
o
i
d
i
n
g
Harm
Conserving
I
n
f
l
u
e
n
ci
n
g
Others
T
a
ki
n
g
Initiative
91
Figure 2. Expected structure of the EGB Descriptive Norms Scale based on the
structure of the Green Five Taxonomy (Ones & Dilchert, 2012a).
a
Reverse-coded items
Item
1
Item
14
a
Item
11
Item
7
Item
17
a
Item
12
Item
3
a
Item
37
Item
3
2
a
Item
28
Item
26
Item
16
Item
4
Item
8
a
Item
9
Item
20
Item
13
a
Item
15
a
Item
35
Item
27
Item
25
Item
22
Taking Initiative Conserving
Influencing Others Avoiding Harm
Working
Sustainably
G
e
n
e
r
a
l
G
r
e
e
n
P
e
r
f
o
r
m
a
n
c
e
E
E
E
E
E
E
E
E
E
D
D
D
D
D
Item
21
Item
38
a
Item
6
Item
39
Item
10
Item
5
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
Item
23
Item
30
Item
3
1
a
Item
33
a
Item
24
a
Item
19
a
Item
18
Item
2
Item
36
Item
34
Item
29
a
92
Figure 3. The final structure of the EGB Descriptive Norms Scale including
standardized and raw path coefficients; raw path coefficients are in parentheses.
a
Reverse-coded items
Item
1
Item
4
Item
23
Item
11
Item
7
Item
4
Item
2
Item
30
Item
18
Item
34
Item
36
Item
16
Item
12
Item
4
Item
6
Item
39
Item
37
Item
32
a
Item
28
Item
26
Item
4
Item
9
Item
20
Item
21
Item
5
Item
10
Item
35
Item
27
Item
25
Item
22
Taking Initiative
Conserving Influencing Others Avoiding Harm
Working
Sustainably
General Green Performance
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
D
D
D
D
D
.
6
9
(.
7
2
)
.
7
5
(.
9
8
)
.
8
5
(1
.
0
0
)
.72 (.83)
.
8
1
(.
8
9
)
.
8
2
(1
.
0
4
)
.82 (1.00)
.
8
1
(.
91
)
.
6
6
(.
7
1
)
.
8
2
(.
9
3
)
.70 (.92)
.
6
4
(.
8
4
)
.
7
3
(.
9
3
)
.
6
9
(.
9
5
)
.79 (1.00)
.
8
1
(.
9
2
)
.
6
9
(.
91
)
.40 (.68)
.75
(1
.
0
0
)
.72 (.99)
.
8
2
(1
.
0
1
)
.84 (1.05)
.
8
4
(1
.
0
4
)
.
7
8
(1
.
0
0
)
.76 (1.01)
.
8
0
(.
93
)
.64 (.83)
.
7
8
(.
6
0
)
.
9
9
(.
8
9
)
.
9
9
(.
9
5
)
.
9
6
(.
9
2
)
.
9
9
(.
8
9
)
93
REFERENCES
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and
Human Decision Processes, 50, 179–211.
Ajzen, I. (2002). Constructing a TPB questionnaire: Conceptual and
methodological considerations. Retrieved February 04, 2014, from
http://chuang.epage.au.edu.tw/ezfiles/168/1168/attach/20/pta_41176_768
8352_57138.pdf
Ajzen, I., & Fishbein, M. (1977). Attitude-behavior relations: A theoretical analysis
and review of empirical research. Psychological Bulletin, 84(5), 888–918.
doi:10.1037//0033-2909.84.5.888
Ajzen, I., & Sexton, J. (1999). Depth of processing, belief congruence, and
attitude-behavior correspondence. In S. Chaiken & Y. Trope (Eds.), Dual-
process theories in social psychology (pp. 117–140). New York, NY:
Guilford.
Armitage, C. J., & Conner, M. (2001). Efficacy of the theory of planned
behaviour: A meta-analytic review. The British Journal of Social
Psychology, 40, 471–499.
Bamberg, S., & Möser, G. (2007). Twenty years after Hines, Hungerford, and
Tomera: A new meta-analysis of psycho-social determinants of pro-
environmental behaviour. Journal of Environmental Psychology, 27(1),
14–25. doi:10.1016/j.jenvp.2006.12.002
Bamberg, S., & Schmidt, P. (2003). Incentives, morality, or habit?: Predicting
94
students’ car use for university routes with the models of Ajzen, Schwartz,
and Triandis. Environment and Behavior, 35(2), 264–285.
doi:10.1177/0013916502250134
Bandura, A. (1986). Social foundations of thought and action: A social cognitive
theory. Englewood Cliffs, NJ: Prentice Hall.
Bansal, P., & Roth, K. (2000). Why companies go green: A model of ecological
responsiveness. Academy of Management Journal, 43(4), 717–736.
Belsley, D. A., Kuh, E., & Welsch, R. E. (1980). Regression diagnostics:
Identifying influential data and source of collinearity. New York, NY: Wiley.
Bentler, P. M. (2006). EQS 6 structural equations program manual. Encino, CA:
Multivariate Software, Inc.
Bentler, P. M., & Chou, C.-P. (1987). Practical issues in structural modeling.
Sociological Methods & Research, 16(1), 78–117.
doi:10.1177/0049124187016001004
Berkowitz, L., & Buck, R. W. (1967). Impulsive aggression: Reactivity to
aggressive cues under emotion arousal. Journal of Personality, 35(3),
415–424.
Bowling, A. (2005). Just one question: If one question works, why ask several?
Journal of Epidemiology and Community Health, 59(5), 342–345.
Brown, M. E., Treviño, L. K., & Harrison, D. A. (2005). Ethical leadership: A social
learning perspective for construct development and testing. Organizational
Behavior and Human Decision Processes, 97(2), 117–134.
95
doi:10.1016/j.obhdp.2005.03.002
Cable, D. M., & Judge, T. A. (1996). Person-organization fit, job choice decisions,
and organizational entry. Organizational Behavior and Human Decision
Processes, 67(3), 294–311.
Carmines, E. G., & McIver, J. P. (1981). Analyzing models with unobserved
variables: Analysis of covariance structures. In G. Bohrnstedt & E.
Borgatta (Eds.), Social measurement: Current issues (pp. 65-115).
Beverly Hills, CA: Sage Publications, Inc.
Carrico, A. R., & Riemer, M. (2011). Motivating energy conservation in the
workplace: An evaluation of the use of group-level feedback and peer
education. Journal of Environmental Psychology, 31(1), 1–13.
doi:10.1016/j.jenvp.2010.11.004
Cialdini, R. B. (2003). Crafting normative messages to protect the environment.
Current Directions in Psychological Science, 12(4), 105–109.
doi:10.1111/1467-8721.01242
Cialdini, R. B. (2007). Descriptive social norms as underappreciated sources of
social control. Psychometrika, 72(2), 263–268. doi:10.1007/s11336-006-
1560-6
Cialdini, R. B. (2012). The focus theory of normative conduct. In P. Van Lange,
A. Kruglanski, & E. Higgins (Eds.), Handbook of theories of social
psychology (Vol. 2, pp. 295–312). Thousand Oaks, CA: Sage Publications
Ltd.
96
Cialdini, R. B., Reno, R. R., & Kallgren, C. A. (1990). A focus theory of normative
conduct: Recycling the concept of norms to reduce littering in public
places. Journal of Personality and Social Psychology, 58(6), 1015–1026.
doi:10.1037//0022-3514.58.6.1015
Cialdini, R. B., & Trost, M. R. (1998). Social influence: Social norms, conformity,
and compliance. In D. Gilbert, S. Fiske, & L. Gardner (Eds.), The
handbook of social psychology (4th ed., pp. 151–192). New York, NY:
McGraw-Hill.
Cooper, W. H. (1981). Ubiquitous halo. Psychological Bulletin, 90(2), 218–244.
doi:10.1037//0033-2909.90.2.218
Daily, B. F., Bishop, J. W., & Govindarajulu, N. (2009). A conceptual model for
organizational citizenship behavior directed toward the environment.
Business & Society, 48(2), 243–256. doi:10.1177/0007650308315439
Darnall, N., Henriques, I., & Sadorsky, P. (2008). Do environmental management
systems improve business performance in an international setting?
Journal of International Management, 14(4), 364–376.
doi:10.1016/j.intman.2007.09.006
Diez Roux, A. V. (2002). A glossary for multilevel analysis. Journal of
Epidemiology & Community Health, 56(8), 588–594.
doi:10.1136/jech.56.8.588
Epstein, S. (1983). Aggregation and beyond: Some basic issues on the
prediction of behavior. Journal of Personality, 52(3), 360–392.
97
Evans, A., Russell, S., Fielding, K., & Hill, C. (2012). Turn it off: Encouraging
environmentally-friendly behaviours in the workplace. Asia Pacific Work in
Progress, 8, 1–18. Retrieved from
http://www98.griffith.edu.au/dspace/handle/10072/52389
Fornara, F., Carrus, G., Passafaro, P., & Bonnes, M. (2011). Distinguishing the
sources of normative influence on proenvironmental behaviors: The role of
local norms in household waste recycling. Group Processes & Intergroup
Relations, 14(5), 623–635. doi:10.1177/1368430211408149
Fugate, B. S., Stank, T. P., & Mentzer, J. T. (2009). Linking improved knowledge
management to operational and organizational performance. Journal of
Operations Management, 27(3), 247–264. doi:10.1016/j.jom.2008.09.003
Gärling, T., Fujii, S., Gärling, A., & Jakobsson, C. (2003). Moderating effects of
social value orientation on determinants of proenvironmental behavior
intention. Journal of Environmental Psychology, 23(1), 1–9.
doi:10.1016/S0272-4944(02)00081-6
Gockeritz, S., Schultz, P. W., Rendon, T., Cialdini, R. B., Goldstein, N. J., &
Griskevicius, V. (2010). Descriptive normative beliefs and conservation
behavior: The moderating roles of personal involvement and injunctive
normative beliefs. European Journal of Social Psychology, 40, 514–523.
doi:10.1002/ejsp.643
Goldstein, N. J., Cialdini, R. B., & Griskevicius, V. (2008). A room with a
viewpoint: Using social norms to motivate environmental conservation in
98
hotels. Journal of Consumer Research, 35(3), 472–482.
doi:10.1086/586910
Goldstein, N. J., Griskevicius, V., & Cialdini, R. B. (2007). Invoking social norms:
A social psychology perspective on improving hotels’ linen-reuse
programs. Cornell Hotel and Restaurant Administration Quarterly, 48(2),
145–150. doi:10.1177/0010880407299542
Graves, L. M., Sarkis, J., & Zhu, Q. (2013). How transformational leadership and
employee motivation combine to predict employee proenvironmental
behaviors in China. Journal of Environmental Psychology, 35, 81–91.
doi:10.1016/j.jenvp.2013.05.002
Greening, D. W., & Turban, D. B. (2000). Corporate social performance as a
competitive advantage in attracting a quality workforce. Business &
Society, 39(3), 254–280.
Griskevicius, V., Goldstein, N. J., Mortensen, C. R., Cialdini, R. B., & Kenrick, D.
T. (2006). Going along versus going alone: When fundamental motives
facilitate strategic (non)conformity. Journal of Personality and Social
Psychology, 91(2), 281–294. doi:10.1037/0022-3514.91.2.281
Harland, P., Staats, H., & Wilke, H. A. M. (1999). Explaining proenvironmental
intention and behavior by personal norms and the theory of planned
behavior. Journal of Applied Social Psychology, 29(12), 2505–2528.
Hockey, G. R. J., & Hamilton, P. (1970). Arousal and information selection in
short-term memory. Nature, 226, 866-867.
99
Jacobson, R. P., Mortensen, C. R., & Cialdini, R. B. (2011). Bodies obliged and
unbound: Differentiated response tendencies for injunctive and descriptive
social norms. Journal of Personality and Social Psychology, 100(3), 433–
448. doi:10.1037/a0021470
Johnson, B. T., & Eagly, A. H. (1989). The effects of involvement on persuasion:
A meta-analysis. Psychological Bulletin, 106, 290–314.
Judge, T. A., & Bretz, R. D. (1992). The effects of work values on job choice
decisions. Journal of Applied Psychology, 77(3), 261–271.
Kafry, D., Jacobs, R., & Zedeck, S. (1979). Discriminability in multidimensional
performance evaluations. Applied Psychological Measurement, 3(2), 187–
192.
Kaiser, F. G. (1998). A general measure of ecological behavior. Journal of
Applied Social Psychology, 5, 395–422.
Kaiser, F. G., Schultz, P. W., & Scheuthle, H. (2007). The theory of planned
behavior without compatibility? Beyond method bias and past trivial
associations. Journal of Applied Social Psychology, 37(7), 1522–1544.
doi:10.1111/j.1559-1816.2007.00225.x
Kaiser, F. G., Wolfing, S., & Fuhrer, U. (1999). Environmental attitude and
ecological behaviour. Journal of Environmental Psychology, 19, 1–19.
Kallgren, C. A., Reno, R. R., & Cialdini, R. B. (2000). A focus theory of normative
conduct: When norms do and do not affect behavior. Personality and
Social Psychology Bulletin, 26(8), 1002–1012.
100
doi:10.1177/01461672002610009
Kalshoven, K., Den Hartog, D. N., & De Hoogh, A. H. B. (2011). Ethical
leadership at work questionnaire (ELW): Development and validation of a
multidimensional measure. The Leadership Quarterly, 22(1), 51–69.
doi:10.1016/j.leaqua.2010.12.007
Kozlowski, S. W., Kirsch, M. P., & Chao, G. T. (1986). Job knowledge, ratee
familiarity, conceptual similarity and halo error: An exploration. Journal of
Applied Psychology, 71(1), 45–49. doi:10.1037//0021-9010.71.1.45
Lapinski, M. K., & Rimal, R. N. (2005). An explication of social norms.
Communication Theory, 15(2), 127–147. doi:10.1111/j.1468-
2885.2005.tb00329.x
Leiserowitz, A. A., Kates, R. W., & Parris, T. M. (2006). Sustainability values,
attitudes, and behaviors: A review of multinational and global trends.
Annual Review of Environment and Resources, 31, 413–444.
doi:10.1146/annurev.energy.31.102505.133552
Leuthesser, L., Kohli, C. S., & Harich, K. R. (1995). Brand equity: The halo effect
measure. European Journal of Marketing, 29(4), 57–66.
MacCallum, R. C., Browne, M. W., & Sugawara, H. M. (1996). Power analysis
and determination of sample size for covariance structure modeling.
Psychological Methods, 2(1), 130–149. doi:10.1037//1082-989X.1.2.130
Manning, M. (2009). The effects of subjective norms on behaviour in the theory
of planned behaviour: A meta-analysis. The British Journal of Social
101
Psychology, 48, 649–705. doi:10.1348/014466608X393136
Mauno, S., Kinnunen, U., & Piitulainen, S. (2005). Work–family culture in four
organizations in Finland. Community, Work & Family, 8(2), 115–140.
doi:10.1080/13668800500049563
McDonald, S. (2011). Green behaviour: Differences in recycling behavior
between the home and the workplace. In S. Bartlett (Ed.), Going green;
The psychology of sustainability in the workplace (pp. 59-64). Leicester,
UK: The British Psychological Society.
Mesmer-Magnus, J., Viswesvaran, C., & Wiernik, B. M. (2012). The role of
commitment in bridging the gap between organizational sustainability and
environmental sustainability.
In S. Jackson, D. Ones, & S. Dilchert (Eds.),
Managing human resources for environmental sustainability (pp. 155-186).
Retrieved from
http://site.ebrary.com/lib/csusb/docDetail.action?docID=10575604
Nag, M. (2012). Pro-environmental behaviors in the workplace: Is concern for the
environment enough? (Unpublished doctoral dissertation). University of
Maryland, College Park, MD.
Nolan, J. M., Schultz, P. W., Cialdini, R. B., Goldstein, N. J., & Griskevicius, V.
(2008). Normative social influence is underdetected. Personality & Social
Psychology Bulletin, 34(7), 913–923. doi:10.1177/0146167208316691
Nunnally, J. C. (1967). Psychometric theory. New York: McGraw Hill.
Oceja, L., & Berenguer, J. (2009). Putting text in context: The conflict between
102
pro-ecological messages and anti-ecological descriptive norms. The
Spanish Journal of Psychology, 12(2), 657–666.
Ones, D. S., & Dilchert, S. (2012a). Employee green behaviors. In S. Jackson, D.
Ones, & S. Dilchert (Eds.), Managing human resources for environmental
sustainability (pp. 85-116). Retrieved from
http://site.ebrary.com/lib/csusb/docDetail.action?docID=10575604
Ones, D. S., & Dilchert, S. (2012b). Environmental sustainability at work: A call to
action. Industrial and Organizational Psychology, 5, 444–466.
Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., & Podsakoff, N. P. (2003).
Common method biases in behavioral research: a critical review of the
literature and recommended remedies. The Journal of Applied
Psychology, 88(5), 879–903. doi:10.1037/0021-9010.88.5.879
Preacher, K. J., & Coffman, D. L. (2006, May). Computing power and minimum
sample size for RMSEA [Computer software]. Available from
http://quantpsy.org/.
Ramus, C. A. (2001). Organizational support for employees: Encouraging
creative ideas for environmental sustainability. California Management
Review, 43(3), 85–106.
Reno, R. R., Cialdini, R. B., & Kallgren, C. A. (1993). The transsituational
influence of social norms. Journal of Personality and Social Psychology,
64(1), 104–112. doi:10.1037//0022-3514.64.1.104
Rindova, V. P., Williamson, I. O., & Petkova, A. P. (2005). Being good or being
103
known: An empirical examination of the dimensions, antecedents, and
consequences of organizational reputation. Academy of Management
Journal, 48(6), 1033–1049.
Rivis, A., & Sheeran, P. (2003). Descriptive norms as an additional predictor in
the theory of planned behaviour: A meta-analysis. Current Psychology:
Development, Learning, Personality, Social, 22(3), 218–233.
doi:10.1007/s12144-003-1018-2
Robertson, J. L., & Barling, J. (2013). Greening organizations through leaders’
influence on employees’ pro-environmental behaviors. Journal of
Organizational Behavior, 34, 176–194. doi:10.1002/job
Schultz, P. W. (1998). Changing behavior with normative feedback interventions:
A field experiment on curbside recycling. Basic and Applied Social
Psychology, 21(1), 25–36.
Schultz, P. W., Nolan, J. M., Cialdini, R. B., Goldstein, N. J., & Griskevicius, V.
(2007). The constructive, destructive, and reconstructive power of social
norms. Psychological Science, 18(5), 429–434. doi:10.1111/j.1467-
9280.2007.01917.x
Schwartz, S. H. (1977). Normative influences on altruism. In L. Berkowitz (Ed.).
Advances in experimental social psychology (Vol. 10, pp. 221–279). New
York: Academic Press.
Schwartz, S. H., & Howard, J. A. (1981). A normative decision-making model of
altruism. In J. Rushton & R. Correntiono (Eds.), Altruism and helping
104
behavior (pp. 83-98). Hillsdale, NJ: L. Erlbaum Associates.
Shultz, K. S., & Whitney, D. J. (2005). Measurement theory in action: Case
studies and exercises. Thousand Oaks, CA: Sage Publications, Inc.
Smith, J. R., Louis, W. R., Terry, D. J., Greenaway, K. H., Clarke, M. R., &
Cheng, X. (2012). Congruent or conflicted? The impact of injunctive and
descriptive norms on environmental intentions. Journal of Environmental
Psychology, 32(4), 353–361. doi:10.1016/j.jenvp.2012.06.001
Smith, P. C., & Kendall, L. M. (1963). Retranslation of expectations: An approach
to the construction of unambiguous anchors for rating scales. Journal of
Applied Psychology, 47(2), 149–155.
Tabachnick, B. A., & Fidell, L. S. (2013). Using multivariate statistics (6
th
ed.).
Boston: Pearson Education, Inc.
Takeuchi, R., Yun, S., & Wong, K. F. E. (2011). Social influence of a coworker: A
test of the effect of employee and coworker exchange ideologies on
employees’ exchange qualities. Organizational Behavior and Human
Decision Processes, 115(2), 226–237. doi:10.1016/j.obhdp.2011.02.004
Taub, G. E. (2001). A confirmatory analysis of the Wechsler adult intelligence
scale-third edition: Is the verbal/performance discrepancy justified?
Practical Assessment, Research & Evaluation, 7(22).
Thøgersen, J. (2006). Norms for environmentally responsible behaviour: An
extended taxonomy. Journal of Environmental Psychology, 26(4), 247–
261. doi:10.1016/j.jenvp.2006.09.004
105
Thompson, C. A., Beauvais, L. L., & Lyness, K. S. (1999). When work–family
benefits are not enough: The influence of work–family culture on benefit
utilization, organizational attachment, and work-family conflict. Journal of
Vocational Behavior, 54(3), 392–415. doi:10.1006/jvbe.1998.1681
Tourangeau, R., Rips, L. J., & Rasinski, K. (2000). The psychology of survey
response. Cambridge, England: Cambridge University Press.
Tudor, T., Barr, S., & Gilg, A. (2007). A tale of two locational settings: Is there a
link between pro-environmental behaviour at work and at home? Local
Environment, 12(4), 409–421. doi:10.1080/13549830701412513
Westland, J. C. (2010). Lower bounds on sample size in structural equation
modeling. Electronic Commerce Research and Applications, 9(6), 476–
487. doi:10.1016/j.elerap.2010.07.003
Yukl, G., Mahsud, R., Hassan, S., & Prussia, G. E. (2013). An improved measure
of ethical leadership. Journal of Leadership & Organizational Studies,
20(1), 38–48. doi:10.1177/1548051811429352
