Journal of Environmental Psychology 43 (2015) 13e23

Contents lists available at ScienceDirect

Journal of Environmental Psychology

journal homepage: www.elsevier.com/locate/jep

The environmental action scale: Development and psychometric

evaluation
Susan Alisat*, 1, Manuel Riemer 2
Wilfrid Laurier University, 75 University Ave. W., Waterloo, ON, Canada

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 13 November 2014
Received in revised form
4 May 2015
Accepted 12 May 2015
Available online 14 May 2015

The environmental action scale measures level of engagement in civic actions designed to have a collective impact on environmental issues. These actions are seen as distinct from personal practices
because environmental actions are more collective in nature and focus on system-level change. The scale
includes two sub-factors: one that is connected with what we are calling participatory actions and one
that we label leadership actions. Each of the actions in the scale is rated for frequency of engagement.
Following a rigorous six-step process, the scale was developed and proposed items were tested in a
diverse North American sample. The scale was rened into the nal 18-item scale which was tested on
two additional samples; one international sample comprised of students from six different countries, and
one consisting of known environmental activists. Analyses indicated that the nal scale showed good
reliability, and provided a valid measure of engagement in environmental actions.
2015 Elsevier Ltd. All rights reserved.

Keywords:
Environmental activism
Environmentalism
Scale
Pro-environmental behavior
Environmental action

1. Introduction
1.1. The environmental action scale: development and psychometric
evaluation
In response to growing concerns about serious environmental
threats such as global climate change, an increasing number of
psychologists are joining the ranks of those who are promoting
,
environmental sustainability (APA, 2010; Gifford, 2008; Harre
2011). So far, most efforts of psychologists have focused on mitigation by fostering changes in people's personal practice, such as
diverting waste through recycling and composting (Dittmer &
Riemer, 2013). Increasingly, authors such as Kenis and Mathijs
(2012), Ockwell, Whitmarsh, and O'Neill (2009), and RouserRenouf, Maibach, Leiserowitz, and Zhao (2014) point out,

* Corresponding author.
E-mail address: [email protected] (S. Alisat).
1
Susan Alisat has recently completed her PhD at Wilfrid Laurier University. Her
research interests include environmentalism and its relationship to the development of identity and generativity in adults.
2
Dr. Manuel Riemer is an Associate Professor of community psychology and
sustainability science at Wilfrid Laurier University and the director of the Centre for
Community Research, Learning and Action and the Community, Environment, and
Justice Research Group. His research focuses on environmental action, behavioral
sustainability, environmental justice and the application of psychology to global
climate change mitigation and resilience.
http://dx.doi.org/10.1016/j.jenvp.2015.05.006
0272-4944/ 2015 Elsevier Ltd. All rights reserved.

however, that there are signicant institutional and structural

barriers to changing these types of behavior. They therefore
conclude that grass-root organizing and citizen activism is the
most efcient method of achieving emission reductions (RouserRenouf et al., 2014, p.163). Currently, the prevalence of these
types of environmental actions is relatively low in the general
population (e.g., for the USA see Leiserowitz, Maibach, RoserRenouf, Feinberg, & Howe, 2012). In response, courses and programs are being developed to specically encourage environmental
civic actions, especially among young people (Hegarty, Thomas,
Kriewaldt, Holdsworth, & Bekessy, 2011; Riemer, Lynes, &
Hickman, 2013). Little empirical knowledge, however, is currently
available about what types of programs are most effective in
engaging people in environmental actions as compared to
encouraging behavior changes at a personal level (Kenis & Mathijs,
2012; Riemer et al., 2013). One barrier to the empirical investigation
of these types of program may be the lack of a psychometrically
sound measurement scale for assessing engagement in environmental actions. In response, the Environmental Action Scale (EAS)
was developed to provide a comprehensive, valid, and reliable
measure to facilitate the empirical evaluation of programs to foster
such actions and the development of theory related to engagement
in environmental action. This paper describes the conceptual
foundation, the development, and psychometric evaluation of the
EAS.

14

S. Alisat, M. Riemer / Journal of Environmental Psychology 43 (2015) 13e23

1.1.1. Conceptual and measurement considerations

Dono, Webb, and Richardson (2010) rightly point out that there
is some conceptual confusion in the literature regarding the
distinction of environmental behaviors, action and activism.
Often the term environmental (or pro-environmental) behavior is
used both as an umbrella term as well as a specic term for certain
kinds of behaviors. For the purpose of presenting a scale that is
intended to measure environmental action, it is therefore
important to clarify what its dening features are and how it is
differentiated from the other related concepts.
For the purpose of this paper we are dening environmental
actions as intentional and conscious civic behaviors that are
focused on systemic causes of environmental problems and the
promotion of environmental sustainability through collective efforts. These actions range from low-level participatory civic action,
such as informing oneself about environmental issues and participating in community events, to highly involved and political
leadership actions such as organizing a protest. Engaging in environmental actions is seen as distinct from changing personal
practice, which is focused on reducing the environmental impact of
individual private-sphere behaviors such as home energy use and
transportation to school or work. Dening features of environmental actions are that they are civic behaviors (as compared to
intentions or value) that a person consciously and intentional engages in to create a positive (mostly indirect) impact on the environment through collective and e to varying degrees e political
change. Engaging in these actions often requires specic types of
competencies (i.e., action competence). In the following sections
we will elaborate these different conceptual features of the denition by critically reviewing the relevant literature and considering measurement related questions as appropriate.
1.1.2. Environmental
Dening the rst part, that is environmental, seems relatively
straight forward. Stern (2000), for example, denes environmentally
signicant behavior (which we use here as an umbrella term as we
will discuss further below) as the extent to which it changes the
availability of materials or energy from the environment or alters
the structure and dynamics of ecosystems or the biosphere itself
(p.408). That is, these are things humans do that have a positive
impact on the natural environment, such as buying organic produce
(although this denition does not provide a specic direction of
impact, it is implied as positive in the remainder of Stern's article).
According to Stern (2000), these impacts can be direct, such as
purchasing a car that produces signicantly less carbon dioxide, or
indirect, such as raising awareness about environmental issues or
advocating for pro-environmental policy changes (with the
assumption that those indirect actions will facilitate an actual
impact on the physical world). It is noteworthy that in the context
of the sustainability discourse environmental issues are seen as
signicantly interconnected with other social issues such as social
justice (Riemer & Schweizer-Ries, 2012). Consequently, taking actions related to sustainability is often more broadly dened than
what we are focusing on with this current scale. Stern (2000) also
points out that some behaviors are intended to have a positive
impact on the environment but that impact does not actually
materialize. If the organic produce, for example, is imported from a
distant country and is heavily packaged, the net impact of buying
an organic produce compared to a conventional one could actually
be negative. As a remedy to this dilemma between intent and actual
environmental impact, Stern (2000) proposes another denition
for environmentally signicant behavior, that is, behavior that is
undertaken with the intention to change (normally, to benet) the
environment (p.408). While the rst denition given is important
when trying to assess the environmental impact of changes in

human behavior eespecially in regard to differentiating between

low and high impact behaviors e the latter is more relevant in
regard to understanding what individual and contextual factors
drive people to engage in environmentally focused behaviors. For
conceptual clarity, we will use Stern's term environmentally signicant behaviors to refer to behaviors that t the rst type of
impact-related denition while environmental behavior is used as
an umbrella term to refer to any behavior intended by the actor to
have a positive impact on the environment. Environmental actions
are then a specic type of environmental behaviors. In most cases,
the actual environmental impact of these actions is indirect and
difcult to assess.
1.1.3. Actions versus changing personal practice
Providing conceptual clarity regarding the term action is more
challenging. From a psychological perspective it is difcult to
differentiate between behavior and action as one is often used to
dene the other. A common denition for behavior, for example, is:
the actions by which an organism adjusts to its environment
(American Psychological Association, n.d.). The main intention with
this denition is to describe something that a person (theoretically)
can be observed as doing such as using the voice to communicate or
driving a car. This differentiates behaviors and actions from concepts such as intentions, motivation, attitudes, values and emotions, which can only be inferred (e.g., through communication).
There are many different types of behaviors that are subsumed
under this general category such as instinctual motor movements,
communication, habits, and complex decisions. Some key distinctions are related to the level of consciousness, the degree of
intentionality, the complexity, the amount of effort it takes, and the
timing. When authors refer to environmental behaviors they
typically mean decision-making, habits and other routinized behaviors that are related to a person's personal practice, such as
house-hold energy use, producing and diverting household waste,
purchasing of products and services with environmental impact,
and use of transportation. These behaviors related to personal
practices are also referred to as pro-environmental behaviors (e.g.,
Kollmuss & Agyeman, 2002),environmental-friendly behaviors
(Tindall, Davies, & Mauboules, 2003) or private-sphere behaviors
(e.g., Stern, Dietz, Abel, Guagnano, & Kalot, 1999). We will refer to
these as personal practices as the term pro-environmental
behavior can be misleading and sometimes has included other
types of behaviors such as giving money to an environmental organization (e.g., Dono, et al., 2010).
What is confusing in the literature is whether a term such as
pro-environmental behavior is referring to the act of changing a
personal practice (e.g., making a plan to bike to work, buying a bike
and relevant gear, and guring out the bike route) or to the target
practice itself (e.g., biking to work). Kollmuss and Agyeman (2002),
for example, understand pro-environmental behaviour as
behavior that consciously seeks to minimize the negative impact
of one's actions on the natural and built world (Kollmuss &
Agyeman, 2002, p.240), which Dono et al. (2010) rephrased as
reducing the negative impact of one's actions (p.178) which could
suggest an intended change to an existing practice. In most cases,
scales intended to measure these types of behaviors ask about the
target practice such as How often do you bike to work? While it
may be pertinent to further explore the difference between the
change in personal practice and the personal practice itself, for the
purpose of this current discussion and the development of the EAS,
we consider environmental actions as distinct from either as will
become more clear below.
Jensen and Schnack (1997), propose to use the term environmental action to refer to things that a person does that are
intentional, or consciously undertaken with reference to

S. Alisat, M. Riemer / Journal of Environmental Psychology 43 (2015) 13e23

environmentally driven motives and reasons. This description in

itself is not sufcient to differentiate environmental actions from
consciously and with environmentally-focused motives pursuing
personal practices with a relatively low negative environmental
impact. A dening characteristic of taking environmental action e
as we understand it e as compared to focusing on personal practices, is the emphasis on changes at the societal rather than individual level. We agree with Jensen and Schnack (1997) who point
out that that environmental problems are structurally anchored in
society and our ways of living. For this reason, it is necessary to nd
solutions to these problems through changes at both the societal
and individual level (p.164). Environmental actions are often
described as collective and system-oriented in nature (Dono et al.,
2010). Jensen and Schnack (1997), argue that environmental action
is closely related to democracy as a form of participation and is
targeted on the root causes of a problem. Similarly, Riemer, Lynes,
and Riemer (2013) have described environmental action as the
activities of engaged citizens. Clayton (2003) suggests that environmentalists by denition tend to focus on the larger community
e the species or the ecosystem (p.55). These authors contrast this
orientation from the individualization of the problem, which assumes that change will happen through individuals acting as individuals. The anticipated outcomes of environmental action are e
to different degrees e collective outcomes (Lubell, 2002). Because
of the collective nature of these actions, participation in these is
connected to the belief that collective action will lead to success in
improving a common good and that one's participation will
contribute to that (Lubell, 2002; Rouser-Renouf et al., 2014). The
collective nature also explains why social connections to other
people engaged in environmental action have been found to be an
important predictor of engagement in these actions (Tindall,
s, 2003). Jensen and Schnack (1997) argue that
Davies, & Mauboule
engagement in environmental action also requires a specic set of
skills and knowledge (e.g., the ability to envision an alternative to
the current system), which they termed action competence. The
belief that one lacks such competence can be an important barrier
to become engaged in action (Rouser-Renouf et al., 2014).
1.1.4. Environmental actions as civic actions
There is a lack of clarity in the literature which types of environmental actions should be grouped together under a common
category. A common distinction in the literature has been made
between activist behaviors and what we have dened as personal
practices (Dono et al., 2010). Based on their review of the literature
guin, Pelletier, and Hunsley (1998)
on environmental activism, Se
suggest that activist behaviors include things such as: taking action on a particular environmental problem or conservation issue,
identifying strongly with a social group, signing a petition or giving
money to a group, organizing a campaign, attempting to change the
attitudes of policy makers, citizens, and those who threaten the
environment, being a member of an environmental organization,
and engaging in environmental protection behaviors. These authors as well as Dono et al., (2010) provide convincing evidence
from the literature that these types of behaviors are empirically
distinct from changing personal practices for environmental reasons. For example, they tend to correlate with different types of
predictors. While personal practices are often found to be related to
individual characteristics such as demographics, personal beliefs,
and attitudes as well as social norms, activist behaviors seem to
have only weak connections to these predictors and are more likely
linked to contextual factors (e.g., local environmental problems)
and social connections (e.g., frequent contact with people who are
also involved in environmental activism).
There is less conceptual clarity, however, in regard to the characteristics that make activist type behaviors distinct from a focus on

15

guin et al. (1998) propose that the distinction

personal practices. Se
is related to the difculty and effort required to carry activist behaviors out, which would explain why only the very committed
environmental activists engage in these behaviors. There are two
problems with this conceptualization. First, there are some nonactivist behaviors that can be difcult and require a lot of
commitment, such as installing a solar panel system in a house,
while there are activist behaviors that require relatively little effort,
such as participating in a demonstration. Second, from a measurement point, this differentiation in regard to the degree of difculty would still be considered a oneedimensional scale rather
than warrant distinct scales.
In Stern's (Stern, 2000; Stern et al., 1999) typology of environmentally signicant behaviors he links environmental behaviors in
general to environmentalism as a social movement and differentiates highly committed and engaged activist behaviors from less
intense behaviors in support of the environmental movement. In
the latter he includes low-commitment active citizenship (e.g.,
contributing to environmental organization; signing a petition) and
support of policies that may require some kind of personal sacrice
for the benet of the environment (e.g., carbon tax; Stern et al.,
1999) and private-sphere behaviors. In the more recent article
(Stern, 2000) he refers to the citizenship behaviors and policy
support as non-activist behaviors in the public sphere and also
adds another general category of other environmentally signicant
behaviors. As the name suggest, this latter category is not clearly
dened and subsumes other relevant behaviors that may not t
within the other types, such as an engineer developing a product
with a low environmental impact. Because this type is theoretically
not clearly dened, has not been operationalized, and may include
behaviors that are specic to certain professions, we did not
consider this category further for the purpose of this scale. As
discussed earlier, there is good conceptual and empirical justication to consider private sphere behaviors (i.e., personal practices) as
distinct from public-sphere behaviors. The question, however, is
whether activist and non-activist public-sphere behaviors indeed
represent two different latent dimensions of two distinct scales or
whether they are better operationalized as two ends of one general
scale of environmental action.
Before we approach an answer to that question it is important to
note that for the purpose of this current scale development we will
not consider the policy support type included in the typology of
Stern (2000) for two reasons: First, the way it has been operationalized (e.g., I would be willing to pay much higher taxes in
order to protect the environment; Stern et al., 1999, p.96) indicates
that these are not actual behaviors but behavioral intentions for
behaviors that ultimately the person has relatively little control
over. Second, for the person to actually support the policy through
behavior they would have to engage in either participatory civic
action (e.g., voting for a politician who advocated for the policy) or
activism (e.g., organizing a petition) to support the policy, which is
represented by the other two categories in Stern (2000) typology.
This reduces the above question to whether activist behaviors and
citizenship behaviors warrant two distinct scales or are different
aspects of the same environmental action scale.
There is some empirical evidence to support that activist and
citizenship behaviors are part of the same dimension. As potential
evidence against this assumption, Stern (2000) indicates that he
and his colleagues have found empirical evidence that activism is
distinct from citizenship behaviors. But, the actual evidence he
presents does not support that assumption. In one of the two cited
studies (Stern et al., 1999), which used data from a general
household survey, they operationalized activism with only one
item tapping into participation of demonstration and protest. Not
only is the base-rate for this behavior very low in the general

16

S. Alisat, M. Riemer / Journal of Environmental Psychology 43 (2015) 13e23

population (7% reported engaging in this behavior in the last 12

month in the survey) but measurement theory suggests that a valid
scale needs to include multiple items that together span the range
of the latent dimension to be measured (Smith & Smith, 2004). That
is, it needs to include items that are easy to endorse and items that
are difcult to endorse. Thus, one item is not sufcient to measure a
complex phenomenon such as environmental activism, especially
in the general population. In the other cited study (Dietz, Stern, &
Guagnano, 1998), there are only two collective and political
behavior items (p.455) mentioned and both are actually citizenship behaviors in the 2000 version of Stern's typology (i.e.,
belonging to an environmental group and signing a petition).
guin et al. (1998) included both types of behaviors and found a
Se
one-factor model to be the best t in their explorative factor
analysis. Using the same six items in their study on environmental
activism and social identity, Dono et al. (2010) conrmed the one
factor. The use of only six items in a sample that has a relatively low
base rate of those behaviors may have prevented the discovery of
separate factors, though. Interestingly, in the study by Dono et al.
guin et al. (1998) to mea(2010) they used both the scale by Se
sure activism and the non-activist items from the Stern et al. (1999)
study as a measure of pro-environmental behaviors. The latter
guin et al. (1998)
included the citizenship behavior items, which Se
considered to be part of environmental activism. While Dono et al.
(2010) did not report that they conducted an omnibus explorative
factor analysis with all of the items together, the path analysis
showed a very strong relationship (.73) between the scores on the
Seguin et al activism scale and the Stern et al citizenship scale. Thus,
currently the evidence that these concepts are closely related is
stronger than the evidence that they are distinct.
We conclude that because environmental actions are environmentally focused activities of engaged citizens, they include both
types of Stern's (2000) typology, citizenship and activism behaviors, as those are both considered civic actions (Riemer et al., 2013).
We propose that the difference is a matter of degree in regard to
type of participation and the political pressure of the action. At the
low end of the spectrum are simple participatory citizenship actions such as keeping oneself informed about a societal issue and
engaging with others in conversations about those issues. These are
relatively uninvolved, simple behaviors that create almost no political pressure. Based on our experience with young people
becoming engaged in environmental action, this is where most
people start because these actions require relatively little action
competence. As people become more involved, they start to
participate in collective activities such as community events and
protests and joining an environmental group or organization.
While many individuals never go beyond this level, some take on
active leadership roles in organizing events and groups. These actions are more complex and require a different level of action
competence, especially related to leadership skills. At the end of the
spectrum then are leadership actions that are more political in
nature such as organizing a boycott or a protest. Stern et al., (1999)
point out that the boundary between those who are engaged in a
more supportive way and those who engage in leadership actions
(which they refer to as activists) is fuzzy and they refer to Snow,
Rochford, Worden, and Benford (1986) in noting that people
often move back and forth in these roles, which provides further
support to consider them to be different levels of the same
dimension of engagement in environmental action.
From this and the earlier conceptual consideration we conclude
that we can expect a oneedimensional scale of engagement in
environmental action that ranges from low-intense and simple
participatory civic actions to highly committed, organized, and
guin et al.'s (1998) review
political leadership actions. Following Se
we also consider environmental protection behaviors (i.e.,

environmental conservation efforts) as environmental actions even

though they may not be considered civic actions in the traditional
sense. There are, however, primarily collective in nature and they
are focused outside the individual and, thus, distinct from changing
personal practice. It could be argued that sometimes people change
personal practices in order to raise awareness about environmental
issue such as Matthew Luxon and Waveney Warth who tried to live
waste-free in their home in New Zealand for a year, which inspired
a lot of other people (Rubbisfree, n.d.). While these changes to
personal practice could be considered environmental actions based
on our denition, we decided not to include them because those
cases are relatively rare and could result in unnecessary confusion
when included on a measurement scale. We conducted explorative
psychometric analyses (see below) to investigate whether the
conceptual decisions presented in this review can be conrmed
with empirical data.
To summarize, our intention was to develop a comprehensive
psychometric scale that would reliably assess different levels of as
well as changes in engagement in intentional and conscious civic
behaviors (including conservation efforts) that are focused on
systemic causes of environmental problems and the promotion of
environmental sustainability through collective efforts. We chose
to use the term environmental actions for our scale for four main
reasons. First, the term behavior is closely associated with personal practice behaviors and is commonly used in that context.
Second, the term taking action is often associated with engaging in
collective civic actions that go beyond personal practices. Third, the
conceptual confusion regarding the term activism discussed above
warrants the use of a different term as well. This term should be
inclusive of both low-intense civic actions and political activism
behaviors, which the term environmental action provides. Fourth,
many people who engage in environmental actions are hesitant to
identify themselves as activists. In fact, Rouser-Renouf et al. (2014)
found that the identication with being an activist was the biggest
barrier to becoming engaged in climate change activism in the USA.
This negative connotation of the term activism was conrmed for
other countries as well in our discussion with the international
experts we consulted for the development of the EAS (see below).
Thus, a more neutral term was necessary for the purpose of what
the EAS is supposed to asses.
1.2. Scale development
While several of the articles discussed above include a list of
items that are indicators for different types of environmental action, none of them describe an actual measurement scale that
represents the full range of environmental actions and was developed using a rigorous psychometric measurement evaluation. Our
goal was to develop a comprehensive scale based on both traditional psychometric considerations (e.g., factor analysis) as well as
more modern measurement approaches (i.e., Item Response Theory). For this purpose, several steps were taken in the development
of this scale, following the guidelines described by DeVellis (2003)
and the procedures described by Riemer et al. (2012), who also
combined these two types of measurement approaches.
First, we determined that this scale should provide a measure of
environmental action (as dened above) for youth (16e26) and
adults, and should do so within a specied time frame, to allow for
the measure of changes in engagement in those actions over time. It
was determined that individuals would rate themselves in regard
to the extent of their past engagement in each action over the past
six month to account for the relative infrequency of some of the
behaviors.
Then, an initial item pool of 65 items were generated based on a
general review of the literature, previous research and in

S. Alisat, M. Riemer / Journal of Environmental Psychology 43 (2015) 13e23

consultation with environmental activists and researchers from

different countries representing both economically developing and
developed countries in the global North and South. In identifying
the initial item pool we considered multiple relevant aspects of
environmental actions. For example, we strove to generate items
which represented the full range from simple citizenship behaviors
that many people engage in (e.g., informing yourself about and
issue and having a conversation with friends and families about the
issue) to highly political leadership actions that only highly
engaged activists take on (e.g., organizing a protest). We also
wanted to ensure that the variety of environmentally focused actions in different countries are represented, such as conservation
efforts, participating in educational event, and being part of a group
that specically focuses on the connection of environmental issues
with other social issues such as an environmental justice organizations. Following the earlier conceptual discussion the main
criteria were to have sufcient items of different level of political
pressure and degree of active participation and leadership. Items at
each of these levels were oversampled in order to be able to select
the psychometrically strongest in the following steps. We also
included several personal practice behaviors to allow for the possibility that empirically they do not warrant a distinct scale, as we
hypothesized based on the discussion above.
After the initial pool and format of the items were established
they were reviewed, tested and rened in multiple steps. First,
items were rened based on cognitive interviews (Willis, 1999) to
evaluate the respondents' understanding and interpretation of the
scale items. Next, the items were reviewed by several researchers
who do research related to environmental action, and by scholars
from Bangladesh, India, Uganda the USA and Germany to ensure
their relevance in a broad range of cultural and political contexts. At
each stage, items were added, deleted, or reworded for clarity,
resulting in a preliminary 36-item test version of the scale.
Following this, we identied other measures which could be
used to validate the environmental action scale. For some individuals, the natural world provides an important focus for the
self; Clayton described this as an environmental identity (Clayton,
2003). Environmental identity may be particularly signicant in
relation to personal engagement in environmental activities and
actions (Whitmarsh & O'Neill, 2010). Identied environmental activists have been found to score signicantly higher than nonactivists on both Environmental Identity and on another (singleitem) measure of environmental interest, (Matsuba et al., 2012), so
both were included in the survey package.
Another step was to administer the survey package to participants, to evaluate and further rene the newly developed measure.
Two distinct samples were recruited. A student sample provided a
broad base of emerging adults, and a general non-student sample
added diversity in terms of age and background. We evaluated the
scale items using a comprehensive mixed-method procedure
described by Riemer and colleagues (Riemer & Kearns, 2010;
Riemer et al., 2013). Based on these analyses, the scale was
reduced to the nal 18 items. We are presenting some more details
regarding these analyses in the next section.
In the nal step we conrmed the quality and psychometric
properties of the 18-items version presented in this paper (see Scale
Evaluation below).
2. Preliminary analyses and initial evaluation of the scale
items
2.1. Methods
A convenience sample of 205 undergraduate university students
(69.8% female) attending a midsize Canadian university were

17

recruited from the psychology participant pool. Students were

between 16 and 62 years (mean age 18.88 years, SD 3.34). Most
were studying full time (96.1%), worked less than 5 h per week
(75.1%), and volunteered less than 5 h per week (86.3%). A more
diverse adult sample of 161 individuals (44.7% female) was
recruited using Amazon's Mechanical Turk (MTurk) website. Mechanical Turk is a crowdsourcing internet marketplace for work,
where computer programmers can post tasks for a network of individuals who perform those tasks on-line for monetary payment.
Participants in this sample lived in the United States, and were
between 18 and 68 years (mean age 33.49 years, SD 11.02).
Most were not students (80.1%), were working at least 15.5 h per
week (69.6%), and volunteered for less than 5 h per week (79.5%).
Participants from these two samples were combined for analyses,
resulting in an overall sample of 366 participants.
Participants were recruited using online research invitations.
The student sample was recruited using the psychology research
experience pool; students received credit towards their psychology
mark for participating. The MTurk sample was recruited via a
posting on Amazon's Mechanical Turk website. MTurk participants
received a payment of $.50 into their Amazon account, which is
comparable to other research studies which are posted there. The
research was described and participants were invited to complete
the survey by clicking on a link. The on-line survey was hosted on
Survey Monkey. All procedures were approved by the Research
Ethics Board of Wilfrid Laurier University.

2.2. Evaluation of the scale items

The ExpectationeMaximization algorithm (Schafer & Olsden,
1998) was used to impute values for missing scores on the EAS
scale (4 or less cases on any given item, for 36 items).3
In preliminary analyses, test-retest reliability of the scale was
assessed by evaluating the consistency of the responses of a subset
of the student sample over time. Twentyesix percent of the student
sample4 (n 27; 26% female) completed the follow-up portion of
the survey. Surveys were completed an average of 13.67 days after
they completed the survey (range 11e17 days). The intraclass
correlation coefcient between the EAAS scores on the rst survey
and the follow-up survey was very high (ICC .97).
Using elements of classical test theory and item response theory
(i.e., Rasch measurement; Smith & Smith, 2004) we were able to
assess the interrelationship of the items, investigate the dimensionality of the scale, and identify stronger and weaker items as
well as items that are redundant in regard to providing information
about the latent phenomenon. The goal was to further reduce the
number of items. Classical test theory (CTT) was used to examine
properties of the individual items, and item response theory (IRT)
was used to further assess the individual items and their relationship to the overall scale. Specically, we applied the rating scale
model (RSM) with polytomously scored items (Andrich, 1998) that
is part of the Rasch modeling approach. This analysis provides item
difculty ratings and item t statistics (int and outt). Item difculties show where an item is most precise in estimating the level
of engagement in environmental actions (on a logit scale). Fit statistics quantify how well an item ts with the proposed model of a
scale that ranges from the low-level of the latent dimension to the
high-level. Although the RSM is a 1-parameter logistic model,

3
Participants who responded to less than 85% of the items on the scale were
eliminated from the sample.
4
Because the survey was completed at the end of term, only a small portion of
the sample had completed the survey early enough to allow sufcient time for us to
assess consistency over time.

18

S. Alisat, M. Riemer / Journal of Environmental Psychology 43 (2015) 13e23

WINSTEPS 3.81.0 (Linacre, 2014) provides an estimate of each

item's discrimination, or its ability to differentiate persons with
high and low symptom severity. In addition, WINSTEPS provides
item-person maps that indicate how individual items are distributed across the range of the scale. The ideal scale has an equal
distribution of items across the complete range of the assessed
latent trait (in this case from low-intensity participatory action to
highly political leadership actions). Rasch modeling analyses
(Smith & Smith, 2004) were used to assess int, outt, and
discrimination scores.
Although some items were eliminated on the basis of their
statistical properties, items with a non-normal distribution were
not automatically removed from the data set, as it was expected
that some of the items referring to very engaged activities would be
applicable to very few respondents, and it was important to keep an
assortment of these items, to ensure there was a sufcient variety
of the very difcult actions which are characteristic of only the
most engaged activists. In general, retention of items was guided by
the following inclusion criteria: high inter-item correlation (>.4),
high itemetotal correlations (>.4), good int and outt (.6e1.4 for
MSQ), and have a discriminate score close to one (>.8). Items were
selected such that there was a distribution of items across the
measure scores, and such that redundancy in items in regard to
their contribution to the overall scale score was avoided. Two active
citizenship behaviors were found to be highly correlated, and
conceptually similar, so they were combined in later analyses, and
retained in the scale as one item (see item 5). Models that included
the personal practice items performed more poorly on several of
the indicators (e.g., factor analysis model t indicators and Rasch
model item t criteria). On the bases of these analyses, the personal
practice items were separated from the scale and other redundant
or poorer scale items were eliminated so that the initial item pool
was reduced from 36 items to 18 items.
Besides the overall factor of engagement in environmental action, two sub-factors emerged out of these explorative analyses.
One group of items comprised actions which tend to demonstrate
active environmental leadership such as organizing a boycott.
These actions align most closely with what is commonly understood as environmental activism and we will refer to as leadership
actions. The second group of items consists of actions that are more
participatory in nature and tend to require less effort and personal
risk. These items, such as participating in community events and
talking with others about environmental issues, may be described
as participatory actions. As discussed earlier, these may represent
initial stages of more involved activism or they could simply be a
different way of taking action for the environment.
3. Final scale evaluation
Next steps focused on an evaluation of the 18-item nal scale.
The scale was administered to students from several different
countries, and to a group of known environmental activists. Using
classical test theory and analyses of the international sample, the
interrelationships of the items and the internal reliability of the
overall scale were assessed. Item response theory analyses provided additional information about the relationship of individual
items to the overall scale in this international sample. Validity of
the scale was assessed in several different ways. Convergent validity of the scale was supported by the scale's signicant positive
relationship with other measures, including a self-rating of interest
in the environment, and environmental identity (Clayton, 2003).
Concurrent validity of the scale was assessed by the comparison of
the scale scores of a group of known environmental activists with
scores of a more random group of participants, who were not
recruited based on their environmental engagement. Discriminant

validity was established by demonstrating that this environmental

action scale is related to but distinct from measuring personal
practice behaviors.
3.1. Methods
An international sample of 281 participants was comprised of
individuals from Bangladesh (n 59, 47% female), Germany
(n 25, 36% female), India (n 74, 84% female), Uganda (N 102,
36% female), and the United States of America (n 21, 57% female);
see Table 1. Participants were between 18 and 45 years of age
(Mage 22.86, SD 3.66). The majority reported that their economic position within their country was average or better (88%),
and were studying full time (76.9%; 11.6% reported they were not a
student).
Participants were initially recruited by collaborators in each
country to participate in the validation study. They sent individuals
in their networks an e-mail invitation, which contained a link to the
survey information. Participants in Uganda completed the questionnaire on paper; for all others, the survey was hosted by Survey
Monkey. In each country, participants were included in a lottery
draw for a prize of $100 CAN. All procedures were approved by the
Research Ethics Boards of Wilfrid Laurier University and the
collaborating universities (Makerere University, University of
Dhaka, Western Kentucky University, Zavier's College, University of
Bombay, Saarland University).
An additional sample consisted of 18 known environmental
activists (50% female); see Table 1. These individuals were part of
the second author's network, and were actively involved in environmental organizations. Environmental activists were between 20
and 64 years of age (Mage 33.28 years, SD 11.27). Most indicated
that they were not students (55.6%), and were working at least
15.5 h per week (61.2%). Half of this sample indicated that they
volunteered for more than 5.5 h per week (50%). Individuals listed
in a database of environmental organizations were invited to
participate via an e-mail communication. Interested participants
completed the survey, which was hosted on Survey Monkey, electronically. The questionnaire package consisted of demographic
information, the shortened EAS scale, and the Environmental
identity scale which are described above.
3.2. Measures
3.2.1. Demographic questionnaire
Participants completed several demographic questions
including age, gender, student status, work hours per week and
engagement in volunteer activities. Participants also provided a
self-rating of environmental interest, by rating how interested are
you in environmental issues on a ve-point scale (from 1 to 5).
3.2.2. Environmental action scale
Participants were asked to rate the 18 items of the Environmental Action Scale (EAS) in response to the following question: In
the last six months, how often, if at all, have you engaged in the
following environmental activities and actions? Items were rated
on a 5-point scale from 0 (never) through 2 (sometimes) to 4
(frequently).
3.2.3. Environmental impact behaviors
Four additional environmental impact behaviors common in all
involved countries (e.g., Made environmentally conscious food
choices, Made a decision to reduce my consumption of material
goods by buying less.) were included with the environmental action items, and were rated on the same 5-point scale. These were
included to assess the EAS validity. Fourteen additional items

S. Alisat, M. Riemer / Journal of Environmental Psychology 43 (2015) 13e23

19

Table 1
Items, means, standard deviations, and Rasch measurement statistics.
Item

1. Educated myself about environmental issues

(e.g., through media, television, internet, blogs, etc.)
2. Participated in an educational event
(e.g., workshop) related to the environment.
3. Organized an educational event (e.g., workshop)
related to environmental issues.
4. Talked with others about environmental issues
(e.g., spouse, partner, parent(s), children, or friends).
5. Used online tools (e.g., YouTube, Facebook, Wikipedia,
MySpace Blogs) to raise awareness about environmental issues.
6. Used traditional methods (e.g., letters to the editor, articles)
to raise awareness about environmental issues.
7. Personally wrote to or called a politician/government ofcial
about an environmental issue.
8. Became involved with an environmental group or political party
(e.g., volunteer, summer job, etc.).
9. Financially supported an environmental cause.
10. Took part in a protest/rally about an environmental issue.
11. Organized an environmental protest/rally.
12. Organized a boycott against a company engaging in
environmentally harmful practices.
13. Organized a petition (including online petitions) for an
environmental cause.
14. Consciously made time to be able to work on environmental issues
(e.g., working part time to allow time for environmental pursuits,
working in an environmental job, or choosing environmental
activities over other leisure activities).
15. Participated in a community event which focused on
environmental awareness.
16. Organized a community event which focused on
environmental awareness.
17. Participated in nature conservation efforts (e.g., planting trees,
restoration of waterways).
18. Spent time working with a group/organization that deals with the
connection of the environment to other societal issues such as
justice or poverty.

Classical test theory

Rasch measurement

SD

Item-total r

Measure score

2.88

1.08

.43

1.93

.63

1.11

.77

1.67

1.31

.65

.66

.94

.95

.97

.64

1.06

.63

.63

.98

1.14

1.19

2.93

1.06

.50

1.98

.73

1.00

1.40

1.68

1.33

.60

.66

.69

1.17

1.21

.80

1.04

.65

.39

.95

.95

.86

.40

.83

.57

1.18

1.01

1.12

1.02

1.35

1.42

.80

.31

1.31

.81

.72

.82
.94
.30
.43

1.10
1.23
.76
.93

.56
.71
.59
.43

.35
.20
1.47
1.08

.75
1.05
1.13
.83

1.24
1.02
1.00
1.65

1.23
.96
1.04
1.74

.47

.98

.62

1.02

1.06

1.29

.97

1.22

1.34

.80

.15

1.26

.78

.69

1.48

1.33

.77

.44

1.19

.75

.69

.62

1.08

.76

.67

1.20

.81

.59

1.57

1.38

.71

.55

1.12

.95

.91

1.35

1.41

.71

.30

1.02

1.07

1.17

describing environmental activities as well as other scales were

also included as part of this survey package, but they are not relevant to the current analysis.
3.2.4. Environmental identity scale
The Environmental Identity Scale (Clayton, 2003) is a 12-item
scale which measures a personal feeling of connection with the
natural world (Clayton, 2003). Items such as I think of myself as
part of nature, not separate from it were rated on a 1 (not true of
me at all) to 7 (completely true of me) scale. Past research has
supported the relationship of environmental identity with ecocentrism, universal values, environmental behaviors, environmental decision-making (Clayton, 2003), and environmental
activism (Matsuba et al., 2012). Cronbach's alpha in the present
study (total sample, containing all participants) was .86.
3.3. Results and discussion
3.3.1. Scale properties
The nal 18-item version of the scale was examined to assess its
factor structure and to establish its reliability and validity. The
mean EAS score was computed for participants across all 18 items.
The mean response on the nal scale was 1.29 (range 0 to 3.89,
SD .75). The distribution was slightly skewed towards the right
(.87, SE .15) with negligible kertosis (.39, SE .29). Means for the
individual items were between .30 (item 7) and 2.93 (item 4). Most
items, especially those representing the scale extremes, were
signicantly non-normally distributed. Retained items, means,

Discrimination

Int MSQ

Outt MSQ

standard deviations, and itemetotal correlations are shown in

Table 1.
3.3.2. Reliability
Internal consistency of the EAS was evaluated using elements of
CTT analyses and Rasch measurement. Coefcient alpha for the EAS
scale was high, at .92. Itemetotal correlations were between .43,
and .80.
Results from application of the Rasch model to the data for each
EAS respondent are also found in Table 1. Items were relatively
evenly distributed across the scale with difculty scores ranging
from a low of  1.98 (item 4, talked with others') to .67 (item 16,
organized a community event) on a logit scale. In this case, lower
scores mean that people with a lower level of engagement are more
likely to endorse this item with a high rating. Generally, items
tting well with the Rasch rating scale model will have Mean
Square (MNSQ) t statistics between .6 and 1.4 (Linacare & Wright,
1994). In this case, all items fall within this range with the exception
of items 1, and 12, which are slightly above 1.4 (1.77 and 1.74 MSNQ
respectively for the outt index and 1.65 MSNQ for item 12 on the
int index). Item 1 also had the lowest discrimination index (.63),
which indicates that this item is the least likely to contribute to the
discrimination between those engaged in environmental action
and those who are not, which is not surprising given that it is an
entry level action.
The person-item map (see Fig. 1) suggests that the items are well
aligned along the range of possible levels of engagement and that
there is little redundancy in the items in regard to the information

20

S. Alisat, M. Riemer / Journal of Environmental Psychology 43 (2015) 13e23

Fig. 1. Person map of items.

they contribute for the overall scale score. This map also indicates
that the scale is especially good in assessing and differentiating
people at the mid-levels of engagement. The dimensionality map
indicates that 62.0% of the raw variance is explained by the measure, suggesting that the overall primary dimension of environmental engagement is well represented by the scale but there may
also be some additional secondary dimensions as conrmed by the
factor analysis reported below.

3.3.3. Validity
Conrmatory factor analysis (CFA) using maximum likelihood
estimation was conducted using AMOS 20 (SPSS, 2011) to evaluate
the proposed structure of two sub-factors, one reecting environmental leadership actions, and one representing participatory civic
actions, both loading onto a general engagement factor representing the primary dimension (See Fig. 2). To prevent an underidentied model, the loading of one item on each factor was

Fig. 2. Conrmatory factor analysis of proposed factor structure.

S. Alisat, M. Riemer / Journal of Environmental Psychology 43 (2015) 13e23

constrained to be one, the loadings of these secondary factors onto

the general environmental factor were constrained to be equal, and
the variance of the general factor was constrained to be 1. Items 2
and 3 were conceptually similar, as were items 15 and 16, so the
errors for these item pairs were allowed to co-vary in this analysis
(Kline, 2005).
In testing for the normality assumption, the multivariate
normality critical ratios were greater than 10.0 indicating signicant non-normality of the data, so each model was reran using a
bootstrapping method to obtain a BolleneStine corrected probability (p) value. This approach allows for estimating standard errors
without making distributional assumptions (Chernick, 1999). The
number of bootstrap samples drawn for each analysis was set to
200. Researchers have proposed cut-offs for good model t of a
relative chi-square of less than 2 or 3 (Kline, 2005), RMESEA of less
than .08 (Hu & Bentler, 1995), and CFI of greater than .93 (Byrne,
1994). The CFA for the proposed hierarchical model, with one
general environmentalism factor and two subfactors suggested that
the proposed structure was a relatively good t for the data,
(c2 329.55, df 132; c2/df 2.50; BolleneStine bootstrap c2
p .005; CFI .90; RMSEA .07), with only of the indices slightly
below the proposed minimum value. In comparing a variety of
alternative models (e.g., a simple one-factor model) this factor
structure clearly emerged as the superior t with the data.
These results support the proposed scale structure of two factors, one comprised of items assessing active environmental leadership, and one representing participation. The standardized
estimates of the factor loadings of the two secondary factors onto
the general environmental engagement factor were high; the
standardized estimate for leadership actions was .71, and the
standardized estimate for participatory actions was 1.20. The
standardized estimates of the factor loadings of the items on the
two secondary factors were between .37 and .83 (see Table 2).
Convergent validity was assessed by comparing the scores on
the EAS with scores on other measures, which have been shown to
be related to environmental activism. Scores on the EAS were
signicantly correlated with Environmental Identity (r .36,
p < .001) and with self-rated interest in the environment (r .49,
p < .001).
Analyses of the scale's concurrent validity focused on the sample
of environmental activists. This sample was actively interested in
the environment (Menvironmental interest 4.83, range 4e5;

Table 2
Factor loadings, variance accounted for and eigen values.
Item

Standardized loadings
Participatory actions

1.
2.
4.
5.
8.
9.
14.
15.
17.
18.
3.
6.
7.
10.
11.
12.
13.
16.
Loading on environmental action

Leadership actions

.37
.63
.46
.53
.82
.51
.83
.77
.71
.70

1.20

.62
.53
.64
.70
.70
.42
.66
.80
.71

21

SD .38) and strongly identied with the natural world (Menvir 1.53, SD .39). The mean EAS score was 2.07
(SD .74). A tetest comparison indicated that the environmental
activist sample scored signicantly higher than the international
validation sample on the EAS (t 3.46, df 297, p < .01; d .41),
supporting the concurrent validity of the scale.
We expected that the mean EAS score would be related to but
distinct from the mean score of the four personal practice behavior
items. While many who engage in environmental actions would
very likely engage in these types of behaviors, these behaviors may
also be motivated by a variety of non-environmental reasons such
as nancial savings and social norms and do not represent a form of
environmental action as dened in this article. The relatively
moderate correlation of r .34 (p < .05), conrms our expectation
and supports the discriminant validity of the scale.

onmental identity

4. Discussion and conclusions

Our goal with developing the EAS was to provide a comprehensive and psychometrically sound scale that assesses the level of
engagement in environmental actions. Our hope is that such as
scale will contribute to the empirical evaluation of approaches and
programs intended to foster environmental actions, which we
dened as intentional and conscious civic behaviors that are
focused on systemic causes of environmental problems and the
promotion of environmental sustainability through collective efforts. Based on the literature review, we consider these actions to be
qualitatively distinct from personal practices (e.g., green consumerism), which was conrmed in our empirical study. Programs that
focus on fostering environmental actions are often very different in
nature compared to campaigns promoting changes in personal
practice. The former often focus on root causes and community and
political participation, include participatory activities related to
raising critical consciousness, foster system thinking, and promote
action competence, and rely on smaller group processes such as
peer mentorship (Jensen & Schnack, 1997; Dittmer & Riemer, 2013;
Riemer et al., 2013). Personal practice campaigns and strategies, on
the other hand, tend to focus more on one or two specic behaviors
that are targeted with specic manipulations derived mostly from
social psychology experiments and targeted often at a large group
of people simultaneously (McKenzie-Mohr, 2011). In many cases
such behavioral manipulations do not require specically environmental intentions but could be based on other motives such as
personal health, monetary incentives, or social norms. People
engaged in environmental actions on the other hand tend to have a
strong environmental identity suggesting that the environment
plays a more central role in their lives and serves as a general driver
for their action (Clayton, 2003; Matsuba et al., 2012; Whitmarsh &
O'Neill, 2010). It is therefore prudent to evaluate these qualitative
different approaches to engaging people in environmentally
focused change with different types of scales. Our conceptual
considerations and empirical analyses provide support to that
assertion.
Our psychometric evaluations also conrm our conclusions
from the literature review that environmental actions can be
measured on one continuous dimension ranging from low-intense
citizenship behaviors to very involved political leadership actions.
But, the two secondary sub-factors that emerged also lend support
to those who believed that there are some relevant differences
between leadership actions and more participatory or supportive
actions. While the leadership actions tend to cluster at the upper
end of the scale, the secondary dimension suggests that the differences are more than just a matter of level. As such, the resulting
factor structure of one primary with two secondary sub-factors
could provide some conceptual and empirical clarity regarding

22

S. Alisat, M. Riemer / Journal of Environmental Psychology 43 (2015) 13e23

the categorization and measurement of environmental actions.

Based on our earlier literature review and our ndings we
expect that scores on the EAS are related to Environmental Identity
(Clayton, 2003), action competence, self-efcacy beliefs related to
action competence, collective efcacy beliefs as well as social and
contextual factors such as friends who are also engaged in environmental actions. Future research will be able to provide some
additional insight into this. For example, in a recently completed
multi-national study evaluating a program intended to engage
youth in environmental actions that relies on the promotions of
such determinants of environmental action (e.g., connection to a
local person with personal experience of environmental injustice,
environmental participation in action projects, peer facilitators, and
links with a local environmental organizations) we found that the
EAS was sensitive to change over time (assessments were done at
baseline and after 3, 6 and 12 months) and differentiated between
those who participated in the program and a comparison group
(Authors).
In conclusion, the present evaluation of the EAS is encouraging.
The scale shows strong psychometric properties and appears to be
a reliable and valid measure of environmental actions that correspond with our theoretical conceptualization. The EAS can be used
by researchers to empirically assess changes in environmental
engagement over time and for creating a strong evidence-base for
strategies and programs targeted at environmental actions.
Despite the rigorous testing described above, we anticipate that
in the future, it will be necessary to periodically evaluate the scale
items in terms of the changing face of environmental action; for
example, as technology changes, we anticipate that environmental
actions may change to incorporate new technologies. Also, we
acknowledge that the process of engaging in environmental action
is highly complex and includes a vast variety of possible behaviors,
which makes it challenging to be assessed adequately with a general and relatively short quantitative scale. However, the empirical
ndings reported in this paper do suggest that general trends of
engagement in environmental actions can be captured by using the
EAS.
Acknowledgments
The authors equally contributed to the paper, and are listed in
alphabetical order. This research was supported the Social Sciences
and Humanities Research Council of Canada under Grant #4102011-2506 as part of the international study Youth Leading Environmental Change. This work was supported by the Tri-Council
Research Support Fund. The authors thank the international experts and environmental activists who helped us develop the scale
and Dr. Mark Pancer for his thoughtful comments on an earlier
version of this document. Communications regarding this manuscript may be directed to Susan Alisat, Psychology, Wilfrid Laurier
University, Waterloo, ON Canada, N2L 3C5, [email protected].
References
American Psychological Association. (2010). Psychology & global climate Change:
Addressing a multifaceted phenomenon and set of challenges. A Report of the
American Psychological Association Task Force on the Interface Between Psychology & Global Climate Change. Downloaded from http://www.apa.org/
science/about/publications/climate-change.aspx. November 23, 2010.
American Psychological Association (no date). Retrieved from http://www.apa.org/
research/action/glossary.aspx?tab1).
Andrich, D. (1998). A rating formulation for ordered response categories. Psychometrika, 43, 561e573.
Byrne, B. M. (1994). Structural equation modeling with EQS and EQS/Windows.
Thousand Oaks, CA: Sage Publications.
Chernick, M. R. (1999). Bootstrap methods: a practitioner's guide. New York: John
Wiley & Sons.
Clayton, S. (2003). Environmental identity: A conceptual and an operational

denition. In S. Clayton, & S. Opotow (Eds.), Identity and the natural environment
(pp. 45e65). Cambridge, MA: MIT Press.
DeVellis, R. F. (2003). Scale development: Theory and applications (2nd ed.). In
Applied social research methods series (2nd ed.), (Vol. 26). Thousand Oaks, CA:
Sage.
Dietz, T., Stern, P. C., & Guagnano, G. A. (1998). Social structural and social psychological bases of environmental concern. Environment and Behavior, 30(4),
450e471.
Dittmer, L. D., & Riemer, M. (2013). Fostering critical thinking about climate change:
applying community psychology to an environmental education project with
youth. Global Journal of Community Psychology Practice, 4(1). Available electronically at http://www.gjcpp.org/en/.
Dono, J., Webb, J., & Richardson, B. (2010). The relationship between environmental
activism, pro-environmental behaviour and social identity. Journal of Environmental Psychology, 30(2), 178e186.
Gifford, R. (2008). Psychology's essential role in alleviating the impacts of climate
change. Canadian Psychology, 49(4), 273e280.
, N. (2011). Psychology for a better world. Auckland: Department of Psychology,
Harre
University of Auckland. Downloaded from http://www.psych.auckland.ac.nz/
psychologyforabetterworld. February 26, 2012.
Hegarty, K., Thomas, I., Kriewaldt, C., Holdsworth, S., & Bekessy, S. (2011). Insights
into the value of a stand-alonecourse for sustainability education. Environmental Education Research, 17(4), 451e469.
Hu, L. T., & Bentler, P. M. (1995). Evaluating model t. In R. H. Hoyle (Ed.), Structural
equation modeling: Concepts, issues and applications (pp. 76e99). Thousand
Oaks, CA: Sage.
Jensen, B. B., & Schnack, K. (1997). The action competence approach in environmental education. Environmental Education Research, 3(2), 163e178.
Kenis, A., & Mathijs, E. (2012). Beyond individual behavior change: the role of power, knowledge and strategy in tackling climate change. Environmental Education Research, 18(1), 45e65.
Kline, R. B. (2005). Principles and practice of structural equation modeling. NY:
Guildford Press.
Kollmuss, A., & Agyeman, J. (2002). Mind the gap: why do people act environmentally and what are the barriers to pro-environmental behavior? Environmental Education Research, 8, 239e260.
Leiserowitz, A., Maibach, E., Roser-Renouf, C., Feinberg, G., & Howe, P. (2012).
Americans' actions to limit global warming in September 2012. New Haven, CT:
Yale University and George Mason University. Yale Project on Climate Change
Communication
http://environment.yale.edu/climate/publications/BehaviorSeptember-2012/.
Linacre, J. M. (2014). WINSTEPS 3.81.0 [Computer software]. Retrieved from http://
www.winsteps.com/index.htm.
Linacre, J. M., & Wright, B. D. (1994). Chi-square t statistics. Rasch Measurement
Transactions, 8(2), 350.
Lubell, M. (2002). Environmental activism as collective action. Environment and
Behavior, 34, 431e454.
Matsuba, M. K., Pratt, M. W., Norris, J., Mohle, E., Alisat, S., & McAdams, D. (2012).
Environmentalism as a context for expressing identity and generativity: Patterns among activists and uninvolved youth and midlife adults. Journal of Personality, 80(4), 1091e1115.
McKenzie-Mohr, D. (2011). Fostering sustainable behavior: An introduction to
community-based social marketing. Gabriola Island, BC, Canada: New Society
Publishers.
Ockwell, D., Whitmarsh, L., & O'Neill, S. (2009). Reorienting climate change
communication for effective mitigation: forcing people to be green or fostering
grass-roots engagement? Science Communication, 30(3), 305e327.
Riemer, M., Athay, M. M., Bickman, L., Breda, C., Kelley, S. D., & Vides de
Andrade, A. R. (2012). The peabody treatment progress battery: history and
methods for developing a comprehensive measurement battery for youth
mental health. Administration and Policy in Mental Health and Mental Health
Services Research, 39(1e2), 3e12.
Riemer, M., & Kearns, M. A. (2010). Description and psychometric evaluation of the
youth counseling impact scale. Psychological Assessment, 22(10), 259e268.
Riemer, M., Lynes, J., & Hickman, G. (2013). Engaging youth in environmental
change. A model for developing and assessing youth-based environmental
engagement programmes. ahead of print Journal of Environmental Education
Research, 1e23.
Riemer, M., & Schweizer-Ries, P. (2012). Psychology and sustainability science:
complexity, normativity, and transdisciplinarity in meeting sustainability
challenges. Umweltpsychologie, 16(1), 143e165.
Rouser-Renouf, C., Maibach, E. W., Leiserowitz, A., & Zhao, X. (2014). The genesis of
climate change activism: from key beliefs to political action. Climatic Change,
125, 163e178.
Rubbishfree (no date). About us. Retrieved from http://www.rubbishfree.co.nz/
information.php/info_id/98.
Schafer, J. L., & Olsden, M. K. (1998). Multiple imputation for multivariate missingdata problems: a data analyst's perspective. Multivariate Behavioral Research, 33,
545e571.
guin, C., Pelletier, L. G., & Hunsley, J. (1998). Towards a model of environmental
Se
activism. Environment and Behavior, 20(5), 628e652.
Smith, E. V., Jr., & Smith, R. M. (2004). Introduction to Rasch measurement. Maple
Grove, MN: JAM Press.
Snow, D. A., Rochford, E. B., Worden, S. K., & Benford, R. D. (1986). Frame alignment
processes, micromobilization and movement participation. American

S. Alisat, M. Riemer / Journal of Environmental Psychology 43 (2015) 13e23

Sociological Review, 51, 464e481.
SPSS Inc. (2011). Amos (Version 20.0) [Computer program]. Chicago: SPSS.
Stern, P. C. (2000). Toward a coherent theory of environmentally signicant
behavior. Journal of Social Issues, 56(3), 407e424.
Stern, P. C., Dietz, T., Abel, T., Guagnano, G. A., & Kalot, L. (1999). A value-belief-norm
theory of support for social movements: the case of environmentalism.
Research in Human Ecology, 6(2), 81e97.
Tindall, D. B., Davies, S., & Mauboules, C. (2003). Activism and conservation

23

behavior in an environmental movement: the contradictory effects of gender.

Society & Natural Resources, 16(10), 909e932.
Willis, G. B. (1999). Cognitive interviewing: A How to guide. reducing survey error
through research on the cognitive and decision process in surveys. Short course
presented at the 1999 Meeting of the American Statistical Association. (Rachel
A. Caspar, Judith T. Lessler, and Gordon B. Willis e Research Triangle Institute.)
Downloaded from http://appliedresearch.cancer.gov/areas/cognitive/interview.
pdf. on November 6, 2011.