THE EXPERIENCE SAMPLING METHOD, APPLE IPHONES, AND INVESTIGATING
THE RELATIONSHIP BETWEEN VALUES, SUBJECTIVE WELL-BEING AND PROENVIRONMENTAL BEHAVIOR
By
Douglas Laurent Bessette

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE
Community, Agriculture, Recreation and Resource Studies
2011

!

!

ABSTRACT
THE EXPERIENCE SAMPLING METHOD, APPLE IPHONES, AND INVESTIGATING
THE RELATIONSHIP BETWEEN VALUES, SUBJECTIVE WELL-BEING AND PROENVIRONMENTAL BEHAVIOR
By
Douglas Laurent Bessette
Recent research shows that humans’ private-sphere behavior, or the purchase, use, and
disposal of personal and household products, is responsible for exacerbating climate change, biodiversity loss and environmental degradation. Some believe that these rates of change, loss and
degradation are unsustainable. As a result a great deal of effort has been spent in attempting to
identify the determinants of such behavior. While some researchers focus on attitudes and
intentions as potential determinants, others focus on worldviews, social- and personal-norms, and
values. Most, if not all, of this research relies on single-administration surveys, interviews and
focus groups to generate its data. These methods are often subject to recall error and socially
desirable responding (SDR) bias, as well as typically rely on self-reporting. The current study,
as its first primary objective, tests the joint use of a new method and instrument in proenvironmental behavior research: the Experience Sampling Method (ESM) and the Apple
iPhone, respectively. In four one-week periods, the study uses ESM and the Apple iPhone to
examine seventy-one university sophomores’ values, their subjective well-being, contributions of
certain phenomena to their well-being, and finally their private-sphere pro-environmental
behavior. In adopting this specific method and instrument the study also attempts to ameliorate
the errors and biases mentioned above. The study tests a theoretical framework that suggests
significant associations between certain value orientations and pro-environmental behavior
performance.

!

!

PREFACE

This study introduces the reader to pro-environmental behavior theory and more
specifically to recent research on the association between certain value orientations, individuals’
subjective well-being and pro-environmental behavior. In doing so it also argues that additional
attention and research are necessary, both to advance such theory and to provide an increased
array of tools to those who hope to mitigate the possibly catastrophic effects of climate change,
bio-diversity loss, and environmental degradation. This study’s contribution to such urgent
efforts is the adoption of both a new technology and a relatively novel method, neither of which
has been utilized within the field of socio-environmental research to date. The former is the
Apple iPhone (Apple Incorporated, Cupertino, California), a multi-use multimedia broadband
mobile phone and web-interface device. The latter is the Experience Sampling Method (ESM), a
multiple-administration survey method—as opposed to single-administration, that attempts to
capture the “subjective experience of persons interacting in natural environments in a way that
ensures ecological validity” (Csikszentmihalyi & Larson, 1987: 526).
While both technology and method complement each other well in carrying out tasks
such as measuring behavior and examining perceptions, beliefs and feelings, as will be seen
below, it is perhaps the possibilities that their conjoint use create beyond these initial warrants
that is most exciting.
This thesis will apply both method and instrument in an investigation of the relationship
between values, pro-environmental behavior and subjective well-being. Such a relationship as
presented is inherently socio-psychological, and many of the efforts used to examine it buck the
recent trend of adding or applying external or contextual variables to models that previously
depicted the internal workings of the human mind. The most obvious examples of this and the

"""! !

!

primary theoretical framework utilized in this study come from Shalom H. Schwartz (1992,
1994, 2000) and Tim Kasser (2010, 2011, in press). ESM however, as will be shown, is often
used to examine both the internal and external coordinates of experience. How the two—the
body of research and the method—work jointly will be of primary interest.
The concluding chapter presents results from research conducted at Michigan State
University in the winter of 2011 where the Experience Sampling Method (ESM) and the Apple
iPhone were used to investigate seventy-one university sophomores’ pro-environmental attitudes,
behavior and values. With support from the Sustainable Michigan Endowed Project (SMEP),
this research also aimed at identifying associations between students’ well-being and their proenvironmental behavior performance.
*

*

*

*

*

This thesis is structured around three chapters. The first chapter is an introduction to proenvironmental behavior theory in general and a specific look at Schwartz (1994) and Kasser’s
(2010, 2011, in press) research on values, well-being and pro-environmental behavior. It ends
with Brown and Kasser’s (2005) call for additional research into values and pro-environmental
behavior, research that they argue needs to specifically address recall error and socially desirable
responding (SDR). The second chapter introduces ESM and the benefits of incorporating Apple
iPhones into ESM and socio-environmental research. The final chapter presents the current
study and discusses the implications of its results, while also suggesting avenues for future
research using ESM and broadband mobile phones.

"#! !

!

TABLE OF CONTENTS

List of Tables

vii

CHAPTER ONE: Examining the Determinants of Pro-Environmental Behavior
1.0. Fifty Years of Pro-Environmental Behavior Research
1.0.1. Current Trends in Pro-Environmental Behavior Theory
and Methodology
1.1. Four Models of Pro-Environmental Behavior
1.1.1. Rational-Choice Theories of Behavior
1.1.2. Personal-Norm Theories and Pro-Social Behavior
1.1.3. The New Environmental Paradigm
1.1.4. Value Orientations and Pro-Environmental Behavior
1.2. Schwartz’s Structure and Content of Values
1.2.1. Self-Enhancement Values and Pro-Environmental Behavior
1.2.2. The Causes of Self-Enhancement Values
1.2.3. Self-transcendent Values and Pro-Environmental Behavior
1.3. Subjective Well-Being and Pro-Environmental Behavior
1.3.1. Value Orientations, Subjective Well-Being
and Pro-Environmental Behavior
1.4. The Need for Expanded Studies and Methodology
1.4.1. Socially Desirable Responding (SDR)
1.4.2. Recall Errors and Biases
1.4.3. ESM and the Apple iPhone: In Response

1
1

CHAPTER TWO: The Experience Sampling Method and the Apple iPhone
2.0. Introduction
2.0.1. Introduction to ESM
2.1. Types of Experience Sampling
2.2. Resource Requirements of ESM
2.3. Reliability of ESM
2.3.1. Internal Validity
2.3.2. External Validity
2.3.3. Ecological Validity
2.3.4. A Focus on Intra-Personal Variation
2.4. Critiques of ESM
2.5. ESM Embracing Technology
2.5.1. Larger Sample Sizes
2.5.2. Flexible Survey Instruments
2.5.3. ESM as a Hybrid Technology: Stated and Revealed Preferences
2.6. ESM and Technology: Expanding Beyond Previous Studies

34
34
34
37
38
40
40
43
45
47
48
50
51
52
54
56

#! !

3
6
7
9
13
15
16
17
19
19
20
24
25
25
29
32

!

CHAPTER THREE: An Application of the Experience Sampling Method
and the Apple iPhone: Investigating the Effect of Values and Well-being
on Pro-Environmental Behavior
3.0. Introduction
3.1. Overview of the Current Study
3.1.1 Method and Instrument: ESM and the Apple iPhone
3.1.2. Research Objectives
3.1.3. Values, SWB, and Pro-Environmental Behavior in Previous Studies
3.1.4. Values, SWB and Pro-Environmental Behavior in the Current Study
3.1.5. Science, Expertise and Values in Pro-Environmental Knowledge
3.1.6 Hypotheses
3.1.7. Rating-Scale
3.1.8. Analytical Methods
3.2. Sample Construction and Signal-Response Rate
3.3. Results and Discussion
3.3.1. The Apple iPhone and ESM
3.3.2. The Apple iPhone, ESM and Socio-Environmental Research
3.3.3. Hypotheses
3.3.3.1. Hypothesis 1: SWB & Pro-Environmental Behavior
3.3.3.2. Hypothesis 2: Values and Pro-Environmental Behavior
3.3.3.3. Hypothesis 3: Values and SWB
3.3.3.4. Hypothesis 4: Values and Pro-environmental Knowledge
Campaigns
3.3.3.5. Hypothesis 5: Values and Pro-environmental Knowledge
Campaigns II
3.4. Conclusions

59
59
60
61
62
64
66
68
69
71
73
75
80
80
82
84
85
89
95
98
101
105

APPENDICES
Appendix A: ESM Survey Questions
Appendix B: Values and Pro-Environmental Behavior Data
Appendix C: Values and Subjective Well-Being Data
Appendix D: Values and Environmental Knowledge Videos Data

111
112
118
121
125

REFERENCES

130

#"! !

!

LIST OF TABLES
Table 1. Study Sample and Response Rates

80

Table 2. Summary of Sample Characteristics

85

Table 3. SWB and Pro-Environmental Behavior Results

87

Table 4. Individual Value Orientations and Pro-Environmental
Behavior Results

91

Table 5. Individual Contributions to Satisfaction and Pro-Environmental
Behavior Results

92

Table 6. Individual Value Orientations and SWB Results

95

Table 7. Individual Contributions to Satisfactions and SWB Results

96

Table 8. Environmental Knowledge Video Results I

99

Table 9. Environmental Knowledge Video Results II

100

Table 10. Environmental Knowledge Video Results by Gender

101

Table 11. Individual Value Orientations and Environmental
Knowledge Videos

102

Table 12. Individual Contributions to Satisfaction and Environmental
Knowledge Videos

103

Table 13. Values and Pro-Environmental Behavior

118

Table 14. Values and Subjective Well-Being

121

Table 15. Values and Environmental Knowledge Videos

125

#""!!

!

Chapter One:
Examining The Determinants of Pro-Environmental Behavior
1.0. Fifty Years of Pro-Environmental Behavior Research
A great deal of research exists today suggesting human beings’ day-to-day activities have
a deleterious effect on the earth’s ecosystems, climate and natural resources.

Despite the

prevalence of such studies and their discussion in the media, however, individuals continue to
behave and often believe in a manner that sustains or even accentuates environmental decline.
The distinction between behavior and belief is and always has been an important one for
researchers, as are the distinctions between attitudes, values, preferences, beliefs and intentions.
For over a century behavioral scientists and social psychologists, along with behavioral
economists and sociologists, have investigated the determinants of human behavior, not only
theoretically but also empirically, often through the building of and reinterpretation of
mathematical models. The concepts listed above, e.g. attitudes, values and beliefs, are all
possible culprits. Until the 1960s and 70s however, most of these researchers’ efforts existed
outside the realm of socio-environmental study.
th

This was so because for most of the 20 century human beings were primarily portrayed
as egocentric gain-maximizers, or utility-maximizers. Their behaviors purportedly consisted
only of consuming resources with little or no concern for efficiency, passing waste and costs on
to others, and forming small groups that excluded and neglected the interest of others (de Young,
2000). Appropriate behaviors were those behaviors that were considered most expedient for the
individual decision-maker. Externalities—or consequences either favorable or adverse for those
not involved in the decision-making—were lumped into a residual category and for the most part
were considered either forgotten or only peripherally acknowledged by individuals (Heberlein,

$! !

!

1972). Leopold (1948) argued that land was still property and the land relationship was strictly
economic; it entailed only privileges, not obligations.
White (1967) argues however that in the 1960s and early 70s individuals’ attitudes and
beliefs began to change; suddenly, human beings were no longer seen as separate from or
superior to the natural world. A new awareness of the consequences of human actions was
growing. And decisions that were supposedly once solely based on economic efficiency and yet
had deleterious consequences for the environment were now being questioned in the public
forum. Heberlein (1972: 79) argues that such investigations were due in part to the “general
moral turbulence of the sixties,” and in part to a new ethical standard based on preserving the
“integrity, community and beauty of the biotic community” (Leopold, as quoted in Dunlap &
Van Liere, 1977: 204). Dunlap and Van Liere (1977) saw the situation differently. They argued
that it was traditional ethics with its focus on moral and interpersonal norms that was to blame.
People were not so interested in preserving the natural environment, as they were concerned for
their fellow human beings.
Regardless of the cause, socio-environmental researchers had three new phenomena to
examine: (1) in what manner one interprets the effect of his or her behavior on the environment;
(2) the process in which one reaches that interpretation; and (3) how one’s behavior depends on
that interpretation.

Deciphering each of these three phenomena, it was believed, would

ultimately lead researchers to be able to (4) predict whether or not certain individuals would—
and do—behave more environmentally consciously.
Regrettably, the links between what individuals believe, why they believe it, and how
those beliefs—and values—affect their behavior is not so obvious. Fifty years have passed and

%! !

!

socio-environmental researchers still do not yet know definitively what directs what, and why.
As such, divisions have arisen both in the theory and the literature.

1.0.1. Current Trends in Pro-Environmental Behavior Theory and Methodology
Before examining that division or the theory, methods and models of pro-environmental
behavior specifically, it is first important to understand exactly what comprises “proenvironmental behavior.” Stern (2000: 408) defines environmental behavior by its impact: “the
extent to which [that behavior] changes the availability of materials or energy from the
environment or alters the structure and dynamics of ecosystems or the biosphere itself.” That
behavior can have a direct or indirect effect; clearing forest or polluting causes direct
environmental change, while behaviors like international development policy or tax policy
impact the environment indirectly (yet perhaps much more significantly than the former). Stern
adds that humans’ environmental impact has primarily been a by-product of our desire for
“physical comfort, mobility, relief from labor, enjoyment, power, status, personal security,
maintenance of tradition and family” (ibid: 408). Yet, as was discussed previously, only after
what Heberlein (1972) and Dunlap and Van Liere (1977) argue was the acknowledgment of an
environmental ethic (or norm), did environmental protection become an important consideration
in human decision-making.
It was at that point that environmental behavior began also to be defined by the intent of
the actor, i.e. one’s intention to change the environment. This resulted in both highlighting
environmental intent as an independent cause of behavior and the possibility that positive
environmental intent, or what is referred to in this thesis as “pro-environmental attitudes” and
“pro-environmental behaviors,” may actually fail to result in positive environmental impacts

&! !

!

(Stern, 2000).

Yet regardless of this seemingly glaring incongruity, socio-environmental

researchers have continued to focus on investigating individuals’ pro-environmental intentions—
or more appropriately the lack thereof. Ecologists, climatologists, and engineers investigate the
impact of human behaviors. Sociologists, psychologists and economists examine that impact’s
effect, or how that effect is translated—or fails to translate—into social norms, as well as how to
predict or alter those human behaviors.
These disciplinary divisions aside, at this point two schools of thought exist in
explaining, predicting, and altering individuals’ pro-environmental behavior (Guagnano et al.,
1995). The first is that of the social and cognitive psychologists who study behavior as a
“function of processes internal to the individual” (ibid: 700).

The second is that of the

economists and sociologists who focus on external and contextual forces. There exists another
division and that is between the rational-choice theorists like Fishbein (1967) and Ajzen (1991;
Ajzen & Fishbein, 1980), as well as Tversky and Kahneman (1981), and the pro-social, personalnorm and value-orientation theorists like Schwartz (1972), Kasser (2010, 2011), Stern, Dietz and
Guagnano (Stern et al., 1995), and Geller (1995).
The methods upon which each respective school relies to investigate beliefs, attitudes,
values and behaviors vary as well. Obviously the socio-psychologists and the rational-choice
modelers typically focus on the internal processes of the individual, while the economists and
sociologists generally focus on the external context, or the structural, institutional, and socioeconomic constraints each individual face.

Regardless of which side of the mirror one

investigates, two methods typically dictate. The first is socio-environmental surveys, often
single-administration surveys, which evince individuals’ stated preferences, or individuals’ selfreported beliefs, attitudes, values and behaviors. The second method, often referred to as the

'! !

!

revealed preference method, relies on direct observation or monitoring the choices individuals
make through their participation in the marketplace. Each method has both advantages and
disadvantages, as will be discussed below.
The last two decades however have seen each of the two schools begin to integrate.
Behavioral economists, as well as neuroeconomists who study brain imagery during decisionmaking (e.g. Weber et al., 2007), are striving to reintroduce psychology into their work (Shogren
& Taylor, 2008).

And the rational-choice theorists like Izek Ajzen (1991) are offering

reinterpretations of their earlier models, i.e. the Theory of Planned Behavior, that take into
account both internal and external constraints. This dedication to examining individuals’ context
and constraints in socio-environmental research is for the most part an attempt to close the
“value-action gap,” or the empirical reality that attitudes and beliefs fail to predict behavior
(Blake, 1999).
Such attempts more than ever require the melding of methodologies. To understand the
determinants of pro-environmental behavior, one cannot simply monitor an individual’s market
behavior, or ask him or her questions regarding his beliefs or values about climate change. Often
one must do both, either working simultaneously, which can sometimes be difficult—if not
impossible, or else rely on multiple population samples, which generate validity and financial
concerns.

Thus, in order to successfully close the value-action gap, new methods and new

technologies may be necessary. Paper surveys, telephone interviews, and willingness-to-pay
(WTP) or accept (WTA) studies, while still very important, may now be complemented—if not
replaced entirely in some cases—by studies that rely on mobile technology and real-time
monitoring of individuals’ behavior, beliefs and values.

(! !

!

Certainly technology is moving at a faster speed than that which our core disciplines can
maintain. Peer-review and the replication requirement guide much of our research, and studies
that rely on or introduce new technologies and methods may find themselves without a rudder in
bodies of literature that are decades old. This is of course a necessary process, one that mitigates
the scientific and normative power of one-time and one-off studies that utilize unknown, unreplicable or unclear methods. At the same time, scientists must be careful not to rely too
heavily on only proven methods and technologies, as they may, in avoiding or resisting evolution
of their craft, miss the proverbial boat.

1.1. Four Models of Pro-Environmental Behavior
Predicting, and altering, human behavior is not easy. The number of determinants one
must consider is high. Those determinants can be internal, or specific to human beings in
general, such as one’s beliefs, attitudes, intentions and perceptions. Or, they can be external, i.e.
dependent upon individuals’ context, such as knowledge, income, opportunities to act,
appreciation of norms, or level of self-control. The number of theories and models used to
describe the process of human behavior is equally high. However three models and theories in
particular have come to inform much of today’s research in the socio-environmental arena, and
an introductory examination of each will situate the reader well to understand the theoretical
framework that was ultimately chosen for this thesis and the current project. Any investigation
of pro-environmental attitude and behavioral theory would also be incomplete without a brief
summary of the work’s history.
While each of the following models primarily investigates the internal sociopsychological determinants of behavior, each does so from a unique perspective. The first

)! !

!

focuses on attitudes and intentions; the second focuses on personal and social norms, and the
third focuses on worldviews (each of these terms will be defined below). While each involves
the examination of individuals’ values, none does so as explicitly as the fourth model, which
investigates individuals’ specific value orientations and searches for associations between those
orientations and the individuals’ pro-environmental behavior. It was this theoretical model
(Schwartz, 1994; Brown & Kasser, 2005; Kasser, 2010) that was ultimately chosen to inform this
thesis and the current study. A more in-depth examination of this theoretical construct begins in
Section 1.1.4.

1.1.1. Rational-Choice Theories of Behavior
The first models this thesis examines are that of the social and cognitive psychologists or
rational-choice theorists who examine behavior especially as a “function of processes internal to
the individual” (Guagnano et al., 1995: 700). These theorists, or at least those authors engaged
in socio-environmental research, typically date their models’ origins back to Fishbein and
Ajzen’s Theory of Reasoned Action (Fishbein, 1967; Ajzen & Fishbein, 1980). This sociopsychological model was based on both the presumption that individuals are rational and make
systematic use of the information around them in their decision-making, and that a person’s
intention was the immediate determinants of one’s action.

Intentions were defined as the

motivational factors that influence a behavior, i.e. how hard people are willing to try, or how
much effort they are willing to exert. The greater the intention, the greater should be that
individual’s performance (Ajzen, 1991).
Intentions, then, are the functions of two determinants themselves: one’s attitude toward
the behavior, and one’s subjective norm. An attitude is the degree to which an individual favors

*! !

!

or disfavors the behavior in question, and the subjective norm is a social factor that refers to the
implied or perceived social pressure to (or not to) perform the behavior. By determining the
relationship between attitudes and norms an investigator begins to understand and becomes able
to predict individuals’ behaviors (Ajzen & Fishbein, 1980). It is important to note that Fishbein
and Ajzen believed that external factors, while certainly influential, were internalized as
affecting an individual’s beliefs—which then determine his or her attitudes; i.e. “an external
variable will have effect on behavior only to the extent that it influences the determinants of that
behavior” (ibid: 9)
The belief amongst rational-choice theorists that external variables influence only an
individual’s determinants of behavior has held for decades. Even Ajzen’s (1991) reinterpretation
of the Theory of Reasoned Action maintains such a distinction, arguing that as a general rule the
stronger a person’s intention to engage in a behavior, the more likely should be that behavior’s
performance. Yet, he added, that intention can only find expression if the behavior in question is
under volitional control, i.e. “if the person can decide at will to perform or not perform the
behavior” (ibid: 182). The performance of most behaviors depends to some degree on nonmotivational factors like the availability of requisite opportunities and resources, e.g. time,
money, skills, cooperation, or climate. It is these factors that represent an individual’s actual
control over the behavior (ibid). Intentions then would be expected to influence performance
only to the extent that the person has behavioral control.
Ajzen’s Theory of Planned Behavior thus has in addition to (1) attitudes toward the
behavior and (2) subjective norms, (3) perceived behavioral control (PBC) as a central tenet.
This differs from actual behavior control, which as stated above would be determined by the
availability of necessary opportunities and resources. Instead, PBC is one’s perception of the

+! !

!

ease or difficulty of performing the behavior of interest (ibid). This is very close to Bandura’s
(1977; 1982) notion of self-efficacy, or “an individual’s confidence that he or she can perform a
particular behavior” (quoted in Bagozzi, 1992: 180).
Studies of PBC show that people’s confidence in their abilities to perform specific
behaviors strongly influences their choice of activities, preparation for an activity, the effort
expended during that activity, as well as thought patterns and emotional reactions (Bandura et
al., 1977; Bandura et al., 1980).

1.1.2. Personal Norm Theories and Pro-Social Behavior
The second model widely used by socio-environmental researchers is that of Schwartz’s
(1977) Norm-Activation Theory of Altruism.

Many pro-environmental behaviors are

characterized as altruistically motivated behaviors or pro-social behaviors. Altruism is defined
as feeling or acting on behalf of the welfare of others in cases where self-interest could not be
involved (Jencks, 1990). According to Schwartz (1977) it is especially in regards to these
scenarios that “personal norms” become key (ibid: 222). Such altruistic or helping behaviors are
initiated by exposure to the need of another, and Schwartz offers three explanations for the
process that connects that perception of need to the helping intention: (1) arousal of emotion, (2)
activation of social expectations, or (3) activation of self-expectations. It is only this last that is
codified as a personal norm.
To link back to Fishbein and Ajzen’s theories of reasoned action and planned behavior
(see above), it may help to think of (1) and (2) as being attitudinal (or intentional) and socialnormative, respectively. Or, as Schwartz states, the arousal of emotion may even be less its own
distinct variable and instead collapsible within (3). If (1) is a manifestation of empathy, then it

,! !

!

may be the need to relieve one’s own distress. Conversely, the greater one envisions the
perceived need of the other the more likely he may activate his own personal norms (ibid).
The activation of social expectations on the other hand is unique from their arousal.
Activation means a cognitively driven “directing of attention to expectations sufficient to bring
them in to the stream of information processing” (ibid: 225). These social expectations are
learned in the normal course of socialization and are activated by explicit or subtle
communications from others. They are also implicitly backed by socially mediated sanctions,
which despite being mild, remote or even improbable, are the ultimate reason why expectations
motivate behavior (Blake & Davis, 1964). It is not unfounded to suggest that the urge to meet
these social expectations may also ultimately collapse into self-expectations (Schwartz, 1977).
And self-expectations are instead experienced as feelings of moral obligations. Behavior then is
motivated by the “desire to act in ways consistent with one’s values so as to enhance or preserve
one’s sense of self-worth and avoid self-concept distress (ibid: 225).
To clarify, Schwartz’s theory is based on three basic propositions: an obligation
proposition, an activation proposition, and a defense proposition. They are listed below:
1. Altruistic behavior is influenced by the intensity of moral (personal) obligation,
which an individual feels to take specific helping actions.
2. Feelings of moral obligation are generated in particular situations by the activation of
the individual’s cognitive structure of norms and values.
3. Feelings of moral obligation may be neutralized prior to overt action by defenses
against the relevance or appropriateness of the obligation.
(Schwartz: 1977: 227)

$-!!

!

These propositions suggest three corollaries from which testable hypotheses can be
derived:
1. Individual differences in feelings of moral obligation to perform particular actions
lead to individual differences in overt behavior.
2. The impact of feelings of moral obligation on behavior is a function of conditions,
which influence the initial activation of the individual’s cognitive structure of norms
and values.
3. The impact of feelings of moral obligation on behavior is also a function of
conditions which influence defense against the relevance or appropriateness of the
activated obligation.
(ibid: 227)
The activation proposition and its corollary suggest that the impact of feelings of moral
obligation on behavior is a function of factors, the most notable two being the awareness of
consequences of one’s behavior for others or the environment (AC) and the extent to which one
ascribes responsibility to him or herself for changing the offending condition (AR) (ibid; Stern et
al., 1993). Schwartz argues that AR should be relabeled RD, or responsibility denial; as opposed
to a spontaneous tendency, it is instead generally a tendency to accept rationales for denying
responsibility (1977).
Because Schwartz stresses norms and values, as opposed to Ajzen and Fishbein’s beliefs,
attitudes and intentions, a proper definition of each is necessary. Norms are “what should or
should not be done by particular types of actors in given circumstances,” while values are
“standards of desirability that are more nearly independent of specific situations” (Williams,
1968: 284). Kasser (2010) describes values a bit differently; this will be discussed below.

$$!!

!

However for clarification, equality is a value, while feeling that all high school graduates should
have equal opportunities to go to university is a norm. Norms are either pre-existing based on
past circumstances from which individuals draw their current self-expectations (Pruitt, 1972), or
constructed by reference to general norms and values they have internalized in the past
(Schwartz, 1977). This process is experienced as feelings, as opposed to intellectual judgments
of right and wrong (ibid: 234).
Schwartz (ibid) presumes that the norm construction process can be instigated simply by
asking people how they feel they ought to behave, i.e., “under circumstances X, would you feel a
moral (personal) obligation to do Y?” Obligation is the term used because it refers to action, and
so is particularly beneficial in attempting to predict behavior. It is also a relatively neutral term
and so avoids positive or negative self-evaluations that may instead emphasize guilt-avoidance as
the primary motive for conforming to internal obligations (Fenichel, 1945).
Below is an adapted outline of Schwartz’s procedural model.
I.

Activation steps: perception of need and responsibility
a. Awareness of a person, object or condition in a state of need
b. Perception that there are actions which could relieve the need
c. Recognition of own ability to provide relief
d. Apprehension of some responsibility to become involved

II.

Obligation step: norm construction and generation of feelings of moral obligation
a. Activation of preexisting or situationally constructed personal norms

III.

Defense steps: assessment, evaluation, and reassessment of potential responses
a. Assessment of costs and evaluation of probable outcomes

IV.

Response step

$%!!

!

a. Action or inaction response
(Schwartz, 1977: 241)
One should immediately note the similarities between Schwartz’s model and Fishbein
and Ajzen’s model. They both attempt to internalize within the model what can be considered
and are considered by many today to be external constraints and effects.

As opposed to

measuring an individual’s ability to recycle in the distance between her residence and the
recycling facility, or her availability of transportation, both the Theory of Planned Behavior
(Ajzen, 1991) and Norm-Activation Theory (Schwartz, 1977) can internalize that ability and
availability as perceived behavioral control and the “recognition of [one’s] own ability to
provide relief” and “assessment of costs,” respectively (see Schwartz’s model above). Ajzen
(1991) does acknowledge of course, as mentioned previously, that these same conditions can
also be considered non-motivational factors and will typically attenuate the intention-behavior
relationship.

1.1.3. The New Environmental Paradigm
The third development in attitude-behavior research, as opposed to the two discussed
previously, was specific to the socio-environmental arena. In 1978, Riley Dunlap and Kent Van
Liere introduced a new concept, or worldview, of environmental concern, entitled the “New
Environmental Paradigm” (1978). They argued that many of the U.S.’s ecological problems
stemmed from its traditional values, attitudes and beliefs, namely, belief in abundance and
progress, devotion to growth and prosperity, faith in science and technology, a commitment to
laissez-faire economy, limited governmental planning and private property rights (ibid: 10).
These combined to form a “dominant social paradigm” (DSP) or worldview, i.e. a “way in which

$&!!

!

individuals interpret the meaning of the external world…[and] a mental image of social reality
that guides expectations in a society” (Pirages & Ehrlich, 1974: 43-4).
As opposed to Heberlein (1972), who argued for an earlier shift in the way individuals
saw the natural world, one inspired by an environmental or land ethic, Dunlap and Van Liere
(1978) argued that new ideas like the inevitability of “limits to growth” and the necessity of
achieving a “steady-state” economy, in addition to the rejection of the “anthropocentric notion
that nature exists solely for human use” inspired a challenge to the DSP (e.g. Daly, 1973;
Meadows et al., 1972). Such a challenge they entitled the “New Environmental Paradigm”
(NEP). They then created a 12-item scale that attempted to measure individuals’ acceptance of
the NEP, eventually arguing that the scale created was both an internally consistent and valid
instrument of measurement.
Today the NEP scale is one of the most widely used social-psychological measures of
environmentalism (Stern et al., 1999). Yet it is often treated as a measure of endorsement of a
fundamental paradigm or worldview, as well as of environmental attitudes, beliefs, and even
values, according to Stern et al. (1995).

This “ambiguity is a result both of the inherent

ambiguity in measuring these phenomena as well as Dunlap and Van Liere’s failure to ground
NEP in social-psychological theories of attitude structure” (Dunlap et al., 2000: 427). It is
lauded as a folk ecological theory from which beliefs about the adverse consequences of
ecological change can be easily deduced (Stern et al., 1995), while also being recognized as
dealing only with very general attitudes that very often fail to predict congruent behaviors (e.g.
Schuman & Johnson, 1976). According to Dunlap et al. (2000), it is reasonable to argue that
NEP taps “primitive beliefs” about the nature of earth and our relationship to it—Rokeach (1968:

$'!!

!

6) defines primitive beliefs as forming the inner core of a person’s belief system and representing
“his basic truths about physical reality, social reality and the nature of the self.”
Dunlap et al. (2000) revised their NEP scale in the year 2000, making it more
comprehensive and alleviating a lack of balance in the original scale’s item direction, also calling
it the “New Ecological Paradigm.”

1.1.4. Value orientations and Pro-Environmental Behavior
The preceding introductions provide the reader with both a background and a foundation
for examining this thesis and the current study’s theoretical framework regarding proenvironmental behaviors. Fishbein and Ajzen (Fishbein, 1967; Ajzen, 1991) focused on attitudes
and intentions; Schwartz (1977) focused on personal norms; and Dunlap and Van Liere (Dunlap
et al., 2000) focus on global worldviews. To be sure there are a great many more determinants
and theories, such as those theories focused on affective influences (e.g. sympathy: Allen &
Ferrand, 1999; emotional affinity: Kals et al., 1999; empathy: Schultz, 2000). Yet within each,
an individual’s values play a key role, either in informing the proto-determinants of behavior,
such as Fishbein and Ajzen (Fishbein, 1967; Ajzen, 1991) and Dunlap and Van Liere’s (2000)
“primitive beliefs,” or in constructing the norms that become activated in Schwartz’s (1977)
theory of altruism.

None of the three focuses specifically however on the importance of

individuals’ values in determining their pro-environmental attitudes and behavior. Nevertheless,
recent work by Kasser (2010), amongst others (e.g. Dietz et al., 1998: Stern et al., 1995) has
found that general theories of values, Schwartz’s (1994) in particular, can be extremely useful in
identifying individuals’ likelihood and willingness to engage in pro-environmental behavior
(Stern, 2000).

$(!!

!

Furthermore, Kasser also places individuals’ subjective well-being, or happiness, within
the domain of behavioral determinants, in particular those affecting pro-environmental behavior
(Brown & Kasser, 2005). Despite the small number of studies, there are research findings that
suggest that satisfying our desires, i.e. pursuing our subjective well-being—which also happens
to often be value-specific—is both compatible with and may even lead to increased proenvironmental behavior (DeYoung, 1996, 2000; Eigner, 2001; Sohr, 2001).
As a result, the current study—while also examining the joint application of ESM as a
method and the Apple iPhone as an instrument in socio-environmental research (as will be
discussed in Chapter Two)—examined the interaction of values and subjective well-being with
individuals’ pro-environmental behavior. The next four sections present Schwartz’s (1994)
structure and content of values, which inform Kasser’s (2010) research and ultimately that of the
current study.

1.2. Schwartz’s Structure and Content of Values
Schwartz acknowledges that the commonly accepted definition of a value, as opposed to
an attitude, is that it is a “belief pertaining to desirable end states or modes of conduct that
transcend specific situations, guides selection or evaluation of behavior, people and events, and
is ordered by importance relative to other values to form a system of value priorities” (Schwartz,
1994: 20, 1992; Schwartz & Bilsky, 1987, 1990). While helpful, this definition both presumes
there to be an infinite number of specific values and does not tell anything about what different
types of values there are (Schwartz, 1994). In order to move past simply comparing individual
values, which tells us little, Schwartz redefined values as “desirable trans-situational goals,
varying in importance, that serve as guiding principles in the life of a person or other social

$)!!

!

entity: (ibid: 21). He then argued that these values speak to three “universal requirements,” i.e.
the “needs of individuals as biological organisms, requisites of coordinated social interaction,
and requirements for the smooth functioning and survival of groups” (ibid: 21). In meeting these
requirements ten types of values materialize: power, achievement, hedonism, stimulation, selfdirection, universalism, benevolence, tradition, conformity, and security. Schwartz (ibid) adds
an eleventh value, spirituality; however he acknowledges that this last value does not speak to
one of the three universal requirements and thus may not be existent across all cultures.
Schwartz then, in attempting to structuralize the value-types, argued that, “actions taken
in the pursuit of each type of values have psychological, practical, and social consequences that
may conflict or may be compatible with the pursuit of other value types” (ibid: 23). For
instance, striving for power in general precludes one from seeking benevolence, whereas
pursuing hedonism is generally compatible with the seeking of stimulation. The value-types
then, according to Schwartz, could be classified more generally into four groups: (1) selftranscendence (universalism and benevolence); (2) conservation (security, conformity and
tradition); (3) self-enhancement (achievement and power); and (4) openness to change
(stimulation and self-direction, and to some extent hedonism) (ibid: 24). Self-transcendence is in
opposition to self-enhancement, and openness to change is, quite logically, in opposition to
conservation. These oppositions are of particular importance in socio-environmental research
(Kasser, 2010).

1.2.1. Self-Enhancement Values and Pro-Environmental Behavior
Kasser (2010) argues that one set of values in particular, those of Schwartz’s (1994) selfenhancement group, i.e. power and achievement, is responsible for much of the ecological crisis

$*!!

!

humans face.

He further states that this value group also aligns with an “extrinsic” or

“materialistic” cluster consisting of three types of goals: (1) financial success, or the desire for
money and possessions; (2) image, or the desire to have an appealing appearance; and (3) status,
or the desire to be popular and admired by others (Kasser, 2010: 90).
Studies abound documenting the association of self-enhancing values and extrinsic goals
with individuals caring less and doing less about environmental protection and stewardship.
Kasser (2010) lists a number: (1) Saunders and Munro (2000) and Good (2007) identified that
materialistic values are associated with lower biophilia, or the belief that there is an instinctive
bond between humans and other living beings; (2) Schwartz (1992, 2006) showed that selfenhancing values are associated with individuals caring less about protecting the environment
and attaining unity with nature; and (3) Schultz et al. (2005) identified that across nations high
achievement and power values were associated with worse environmental attitudes.
More importantly perhaps is that high self-enhancement values have been identified as
being negatively associated with pro-environmental behavior and positively with ecologically
destructive behavior. Richins and Dawson (1992), Brown and Kasser (2005) and Gatersleben et
al. (2008) all demonstrated results that support the former association, while Dechesne et al.
(2003), Kasser and Sheldon (2000), and Sheldon and McGregor (2000) all demonstrated results
that support the latter (see Kasser, 2010).

Futhermore, nations that exhibit high self-

enhancement value numbers, or more specifically, mastery values, i.e. manipulating the world to
serve one’s own interest, have higher ecological footprints than those nations with lower mastery
value figures (Kasser, 2010, in press)

$+!!

!

1.2.2. The Causes of Self-enhancement Values
Kasser et al. (2004) argue that the research (e.g. Ahuvia & Wong, 2002; Banerjee &
Dittmar, 2008; Kasser et al., 1995) suggests two primary pathways lead to high materialistic, or
extrinsic, values: social modeling and felt threats, or insecurity. Investigating these pathways is
important, as will be seen below. The first, social modeling, occurs when “individuals are
exposed to people or messages in their environment suggesting that money, power, possessions,
achievement, image and status are important aims to strive for in life” (Kasser, 2010: 92). As
opposed to Schwartz’s (1977) altruistically motivated behavior and pro-social norms, these aims
almost exclusively involve activating individuals’ self-interest.

The second, insecurity, as

demonstrated in the literature, leads individuals to orient toward materialistic aims. Kasser
(2010: 93) gives the example of children’s self-enhancement values being higher when they
experience poverty or a “cold, controlling mother” (e.g. Cohen & Cohen, 1996; Kasser et al,
1995; Rindfleisch et al., 1997; Williams et al., 2000).
Even considering insecurity, such as economic hardship, poor interpersonal relationships,
hunger, personal self-doubt, social exclusion, or fearing one’s own death, increases the priority
that individuals place on materialistic aims or ecologically destructive behavior and attitudes (see
Kasser, 2010: 93 for a complete list of studies).

1.2.3. Self-transcendent Values and Pro-Environmental Behavior
Kasser (2010) and Schwartz (1994) argue that certain values and goals necessarily lie in
opposition to one another. Self-enhancement values, like power and achievement, or extrinsic
goals, like image or status, are psychologically, socially and actively opposed by two selftranscendent values: benevolence and universalism, and the intrinsic goals of self-acceptance,

$,!!

!

affiliation and community feeling. Empirical research demonstrates that these oppositional
values and goals are positively associated with pro-environmental attitudes and behavior.
Schultz et al. (2005) demonstrated that self-transcendent values were positive predictors of
individuals having engaged in a set of 12 environmentally helpful behaviors, and Gatersleben et
al. (2008) and Kasser (2005) found that generosity, which Kasser (2010) argues is akin to
universalism and community feeling, predicted more pro-environmental attitudes.

Finally,

nations that report high harmony values, again akin to the value and goal of universalism and
community feeling respectively, have significantly lower CO2 emissions than those countries
who rate materialistic goals as important (Kasser, 2010, in press).

1.3. Subjective Well-Being and Pro-Environmental Behavior
Just as an introduction to value orientation theory was necessary to locate the current
study, an introduction to the role subjective well-being (SWB) has played in pro-environmental
behavior research is also needed, as the current study examines both.
DeYoung (2000) argues that the two most popular motivations used to induce individuals
to act pro-environmentally are financial, or material, incentives and focusing on altruism, ala
Schwartz’s (1977) Norm-Activation theory. DeYoung (2000: 510) identifies that such “singledetermination theories” remain popular, despite centuries of evidence demonstrating that there
are likely multiple determinations. There are problems with both motivations. Incentivizing,
while often able to initiate pro-environmental behavior, seems unable to produce durable
behavioral change (e.g. Dwyer et al., 1993; Katzev & Johnson, 1987) and can sometimes
discourage behavior as seen in the Israeli Day-Care experiment (Gneezy & Rustichini, 2000).

%-!!

!

Even if the incentivizing continues indefinitely, which is rare, participation rates ultimately begin
to drop (Katzev & Johnson, 1987).
Reactance—the phenomena of participants doing the opposite of what is asked of them
(Brehm, 1966)—is often to blame in cases involving altruistic motivations (Schwartz & Howard,
1981). Reactance in these cases can be caused by individuals’ suspiciousness following a highpressure appeal, psychological reactance, or over-justification. Similar to Gneezy & Rustichini’s
(2000) results, “external pressures to provide aid [may] undermine the internalized motivation to
perform altruistic actions” (Schwartz & Howard, 1981, quoted in DeYoung, 2000: 512).
Focusing on individuals’ self-interest, however, as opposed to altruistic or financial
motivations, may be a potential solution to the problem of pro-environmental behavior, notes
DeYoung (ibid). In a study of individuals’ multiple motives for volunteering, it was discovered
that those who participate because of self-oriented motives do so longer than those who
volunteer for social issues, or because of community-based concerns (Omoto & Snyder, 1995).
It is important to note that they also volunteer longer than those who do so because of “valuebased reasons” (ibid, quoted in DeYoung, 2000: 515).
Borden and Francis (1978, as noted in Lehmann, 1999: 34) hypothesized some time ago
that people who have “satisfied their personal needs are more likely to act ecologically because
they have more resources to care about bigger, less personal social and pro-environmental
issues.”

And Geller (1995) proposed that environmentally friendly behavior can only be

motivated by individuals actively caring (Allen & Ferrand, 1999), i.e. individuals must focus
beyond themselves and be concerned about others in a larger community before they will
actively care or act on behalf of the environment.

Yet this focus only occurs when an

individual’s need for self-esteem, belonging, personal control, self-efficacy, and optimism have

%$!!

!

been satisfied. This last part is key and separates Geller’s (1995) model of active caring from
earlier models of norm-activation (e.g. Schwartz, 1977) by acknowledging that the needs of the
individual may mediate social norms.
As mentioned previously, De Young (2000) argues that a prominent approach to
facilitating pro-environmental behavior has been to focus only on altruism as the crucial motive.
Kaplan (2000: 491), however, acknowledges that this focus results in several “inadvertent
consequences, including contributing to helplessness and stressing sacrifice rather than qualityof-life-enhancing solutions.”

He suggests an alternative approach, a product called the

Reasonable Person Model. This model stresses three goals: (1) providing a durable source of
motivation, (2) reducing the sense of helplessness, and (3) generating innovative solutions that
people do not see as threatening (ibid).
Altruism as defined in section 1.1.2 entails sacrifice, since there cannot be a
compensating benefit to the self. Kaplan suggests that such a ‘negative pay-off’ creates serious
motivational issues that bring into question the “strategic usefulness of the concept” (ibid: 494).
Encouraging altruism communicates a powerful message that adopting pro-environmental
behaviors reduces one’s quality of life by acknowledging, perhaps inadvertently, the positive
relationship between resource use and happiness, and that materialism and waste are more fun.
As a result incentivizing altruistic behavioral change is particularly problematic. An example is
Schwartz and Howard’s (1981) study mentioned above that reported situations in which there
was a significant presence of those factors most conducive to activating norms favoring helping,
and yet decreased rates of helping behavior.
Are there then only two positions to adopt, that of the altruist who must put aside the
“issue of gain, of self-interest, in human behavior” and the “economic man” who argues that gain

%%!!

!

is all that matters” (Kaplan, 2000: 496)? If so, then neither position is satisfying, Kaplan argues.
Instead, he suggests, we should focus on four aspects of human cognition, or information
processing, each of which likely has strong behavioral and motivational implications (Kaplan &
Kaplan, 1989):
•

people are motivated to know, to understand what is going on; they hate being
confused or disoriented

•

people are motivated to learn, to discover, to explore; they prefer acquiring
information at their own pace and in answer to their own questions

•

people want to participate, to play a role, in what is going on around them; they hate
being incompetent or helpless

•

people avoid contexts that they consider conducive to helplessness
(Kaplan, 2000: 498)

The last motivation is key; Levin (1993) reported that more information often leads not
only to greater concern but also to a greater sense of helplessness. Allen and Ferrand (1999) in
their test of Geller’s theory of active caring found that personal control, or the “opposite of
helplessness” in Kaplan’s words (2000: 499) was the only significant predictor variable
associated with pro-environmental behavior.
Thus it seems it is not selfishness, but self-interest that is the motivator, i.e. not an
individual “selfishly consuming resources or creating waste without concern for others,” but
instead an individual “taking care of [oneself] and maintaining [one’s] ability to function
effectively in a challenging and frequently chaotic world” (DeYoung, 2000: 515). Furthermore,
happiness need not be inherently or necessarily selfish. Although it is experienced personally, it
can be derived from “such things as enhancing the well-being of another person or the

%&!!

!

sustainability of an ecosystem” (ibid: 515). Wallach and Wallach (1983: 201) argue that “we are
satisfied or pleased if we attain what we (really) want; we are made happy if something that we
(really) wish for comes to pass.” Thus happiness, and in turn our subjective well-being is not
necessarily egoistic, but instead intrinsic. Kasser (2002) reviews a substantial body of research
that suggests individuals who are intrinsically oriented, as opposed to those who report being
extrinsic or materialistically oriented, report greater personal well-being (see Kasser, 2002). And
Richins and Dawson (1992) report that extrinsic individuals also have lower subjective wellbeing and engage in fewer pro-environmental behaviors.

1.3.1. Value orientations, Subjective Well-Being and Pro-Environmental Behavior
Brown and Kasser (2005) in their study of psychological and ecological well-being
examined the relationship between individuals’ subjective well-being and their proenvironmental behavior (or ERB—ecologically responsible behavior), as well as the relationship
between their value orientation, i.e. intrinsic or extrinsic, and their mindfulness, and their proenvironmental behavior.

They discovered that both, SWB & ERB, and intrinsic value

orientations and ERB, along with mindfulness, were positively correlated. Furthermore, they
were able through their analysis to find support for their propositions, stating quite emphatically
that, “happy people live in more ecologically responsible ways because such individuals hold
intrinsically oriented values and are more mindful of their experience and behavior” (ibid: 360).
The two offered hypotheses regarding the reasons behind these connections, suggesting
that:
“Intrinsic values are, by their very nature, not dependent on material goods
for their fulfillment; thus, energy invested in intrinsic pursuits may mean less

%'!!

!

energy devoted to some of the consumption-based activities reflected in the
ecological footprint analysis and certain of the environmentally friendly behaviors
assessed here. For example, people holding more intrinsic values are unlikely to
be very interested in large ‘‘trophy’’ homes or gas-guzzling vehicles that often
reflect ostentatious displays of wealth or image enhancement. Further, the focus
on community that is a component of an intrinsic value orientation (Kasser &
Ryan, 1996) might lead individuals to try to decrease the ecological impacts of
their behavior so as to benefit future human generations as well as other species.”
(Brown & Kasser, 2005: 361)

1.4. The Need for Expanded Studies and Methodology
1.4.1 Socially Desirable Responding (SDR)
Brown and Kasser (2005) acknowledge that social desirability responding (SDR) may
have had a part to play in their study on SWB, mindfulness and pro-environmental behavior.
Edwards (1957: vi) defines SDR as “the tendency of subjects to attribute to themselves in self
description, personality statements with socially desirable scale values, and to reject those with
socially undesirable scale values.” Paulhus (2002: 49) defines the SDR bias as the “systematic
tendency to answer questionnaire items” with “overly positive self-descriptions,” and breaks
SDR into two components: self-deception and impression management (ibid: 56).

Self-

deception takes the form of either enhancement, i.e. promoting positive qualities, or denial, or
the disavowing of negative qualities (Paulhus & Reid, 1991). Impression management involves
consciously or deliberatively distorting self-descriptions to fool an audience (Sackeim & Gur,

%(!!

!

1978), or else the habitual presentation of a specific positive public impression (that could be
construed as an aspect of personality) (Paulhus, 2002).
A great deal has been made of SDR outside socio-environmental research and a number
of operationalizations exist (e.g. Edwards, 1957, 1970; Wiggins, 1959; Eysenck & Eysenck,
1964; Crowne & Marlowe, 1964; Paulhus, 1984). Paulhus (2002) argues that because of the
large number of alternate constructs it is easy for skeptics to discount SDR as empirically
invalid. Yet despite such skepticism, there exist theoretical grounds for and particular instances
in which one might expect distorted reporting (Goffman, 1959; Schlenker & Weigold, 1989).
One of those instances is in the answering of socio-environmental survey questions.
Previous socio-environmental researchers have admitted the limitations in relying on selfreported measures of environmental attitudes, intentions and pro-environmental behavior (Stern
& Oskamp, 1987; Tarrant & Cordell, 1997). Tarrant and Cordell (ibid) argue that what people
report on a survey is often inconsistent with their real actions (e.g. Bickman, 1972; Heberlein,
1981; Weigel, 1983), and greater attitude-behavior correspondence occurs when actual versus
self-reported behavior is measured (Hines et al., 1987). Yet Milfont (2009) points out that Stern
and Oskamp (1987) and Tarrant and Cordell (1997), the two papers most often referenced to
support SDR in socio-environmental surveys, do not argue for SDR specifically as one of those
limitations. Furthermore Milfont (ibid) argues that there are very few studies that test for SDR
directly in socio-environmental research and those that have examined it indirectly have found
the effect of SDR to be low or non-existent (Kaiser et al., 1999; Hartig et al., 2001; Schahn,
2002; Wiseman & Bogner, 2003; Mayer & Frantz, 2004; Pato et al., 2004).
To get a sense of a few of these studies, a brief summary follows. Kaiser et al. (1999)
used an SDR scale (Amelang & Bartussek, 1970) consisting of 32 items (e.g. “I never claim to

%)!!

!

know more than I actually do”) and found all five of their measures, environmental knowledge,
environmental values, ecological behavioral intentions, responsibility factor and general
ecological behavior, were marginally influenced by SDR. Hartig et al. (2001) used the MarloweCrowne SDR scale (Crowne & Marlowe, 1960) and found that perceptions of restorative quality
and reports of ecological behavior were weakly but significantly correlated with SDR item
responses. Mayer and Frantz’s (2004) investigation of their “connectedness-to-nature” scale
(CNS) found both measures of the CNS and new ecological paradigm (NEP) independent from
SDR measures. Milfont’s own study, which claims boldly that both environmental attitudes and
ecological behavior are completely free from social desirability effects (2008: 266), in reality
only examines the relationship between Paulhus’s (2002) measure for impression management
and the NEP and Milfont and Duckitt’s (2004) ecological behavior scale. This approach of
determining individuals’ likeliness to distort answers by asking them questions about past
behaviors appears to be the only one available to socio-environmental researchers.
Based on these studies and measures, Schahn and Bohner (2002) offer an interesting
interpretation of SDR and why it is less problematic than it may appear initially. They argue that
most socio-environmental research is conducted to evaluate interventions and not to assess
representative population statistics.

Since few researchers focus on the absolute values of

relevant variables, SDR’s effect is mitigated. Yet many socio-environmental researchers are
trying to determine empirically what is being bandied about theoretically in the literature, namely
are general attitudes regarding nature shifting, i.e. are individuals adopting a New Ecological
Paradigm or pro-social ethic? An important distinction exists that were one to adopt Schahn and
Bohner’s (ibid) interpretation one may miss outright; i.e. are those attitudes shifting, or is merely
the impetus to commit SDR increasing?

Such questions demand attempts to determine a

%*!!

!

representative statistic, particularly as climate change and resource depletion become more
pronounced both within the public consciousness and as environmental realities.
Schahn and Bohner also argue that in studies of self-reported pro-environmental
behavior, conscious distortions of individuals’ past behavior “do not appear plausible, since
respondents are typically not motivated to distort responses in the context of a research study”
(ibid: 24). This tautological argument may seem unsatisfying, particularly to those researchers
attempting to determine if respondents distort responses. Instead, Schahn and Bohner (ibid: 24)
argue that because of “the necessary effort and the lacking availability of appropriate
formulation,” those special cases where concern about response distortion is high are often
abandoned. Admittedly the effort to examine SDR is substantial, and in certain arenas is perhaps
not seen as a valuable pursuit.
Yet this thesis argues the opposite. All of the studies above offer valuable insight into the
world of impression management, SDR, and pro-environmental attitudes and behavior, yet such
attempts to measure an individual’s likeliness to distort his answers by asking him how often he
distorts his answers may prove precarious. It is this wariness that motivates the current study,
which was an attempt to avoid SDR entirely by both couching the research as a consumerbehavior study and avoiding any reference to the research’s true socio-environmental aims. This
also aligns with Kasser’s (2011) assertion that by simply acknowledging certain value
orientations, individuals’ behavioral inclinations—and thus their self-reports—can shift.
Brown and Kasser (2005) acknowledge that though SDR may have played a part in their
study, they did not consider SDR a serious concern. They report that this was in part because of
Kaiser’s (1998) research that showed, using Amelang and Bartussek’s (1970) social desirability
scale, that self-reported environmental behavior is not susceptible to social desirability.

%+!!

!

Nevertheless, they list SDR as one of the study’s limitations and argue that future studies should
pay attention to SDR.

1.4.2. Recall Error and Bias
Brown and Kasser (2005) also relied solely on self-report methods and argued that
methods not based on self-reporting are difficult to implement in this type of research.
Nevertheless, the authors call for future studies that “request reports over a limited time frame
(e.g. the past week), or use diary-based reports of ecological behavior and well-being to help
circumvent biases due to retrospective reporting” (ibid: 364). These biases due to retrospective
reporting are often termed recall errors or recall biases. A brief examination of the most
common recall errors and biases and how and why they occur follows.
When comprehending questions, individuals must retrieve from memory whatever
information is necessary to construct their answer. And according to cognitive psychologists,
they rely on two major memory systems: long term memory and working memory—Yarmey
(1979) argues for three: sensory, short-term and long-term. Long-term memory encompasses
semantic and episodic (or autobiographical) memory and procedural knowledge like how to ride
a bike (Tourangeau & Bradburn, 2010: 320). Working memory is the information we are
consciously aware of and that information from long-term memory that is currently being used
by some ongoing cognitive process, like the comprehension of language. In order to retrieve the
information necessary to answer a question, an individual must move information from the longterm memory, which has limitless capacity but is accessed slowly, to working memory, which is
limited in capacity but rapidly accessed. This process is called “remembering” (ibid: 321).
Surveyors often use retrieval cues to activate or direct this process.

%,!!

!

Tourangeau, Rips and Rasinki (2000) argue that when answering survey questions
individuals do one of four things: directly retrieve an existing answer; retrieve general
information from which they construct or infer an answer; retrieve specific memories from
which they derive an answer, or retrieve an impression. Most survey questions do not ask for
existing answers, but require the inference or derivation of an answer from an individual’s
general or specific knowledge. It is during this process of inference or derivation where many
recall errors occur. For one, and of particular relevance to environmental attitude judgments,
survey respondents may be influenced by that information which has already been transferred
from their long-term memory to their working memory based on previous questions (Schwarz &
Bless, 1992). Schwarz et al. (1991) call these “part-whole effects,” or the effect that previously
answering a specific question may have on answering a general question. If an individual is
asked about his marriage, and then asked about his general life-satisfaction, his answer may vary,
either positively or negatively, based on the increased accessibility of the information about his
marriage. Because individuals are unlikely to retrieve all information that may bear on a
judgment, they will naturally truncate the process as soon as enough information has come to
mind to form a judgment with sufficient subjective certainty (Bodenhausen & Wyer, 1987).
Part-whole effects are of particular concern when investigating individuals’ attitudes;
however, there are other sources of recall error that have relevance for both investigating
attitudes and monitoring behavior.

Tourengeau and Bradburn (2010) identify three: (i)

information gets harder to retrieve over time; (ii) retrieval fails to come up with the information,
even though the information exists; and as previously mentioned (iii) attempts to infer what
happened or fill in missing details introduce error. The first is fairly obvious; it is harder to

&-!!

!

remember something the more time that has elapsed since it occurred. The second requires an
explanation.
Yarmey (1979) and Bartlett (1932) argue that we do not remember events by
constructing an exact replication or reproduction, but instead remember general features of
experience. In Bartlett’s words, we start with some schematized representation of what we wish
to remember—which has important implications when considered alongside social desirability
response bias. These schemas reflect our “individual understanding of phenomena and are based
upon our own personalities, values, attitudes, and general plans of thought” (Yarmey, 1979: 63).
Johnson et al. (1993: 14) argue that, “reality is not given directly in perceptual and memory
representations but is a product of judgment processes.” In undergoing these processes we often
incorporate cues that are guesses about what actually happened. Sometimes we incorporate
fiction as fact, we believe a self-generated product was of the moment when in fact it was
generated or perceived much earlier, and we incorporate misleading suggestions into our
accounts of actual events (ibid: 11-13). Time-loss and truth-loss become one, as the retrieving of
our ‘guesses’ becomes more difficult when the event happened a long time ago or we encoded
the information poorly.
Not only can we forget and recount fictional events or behaviors, but we can also
remember too many events. Neter and Wajsberg (1964) call this “telescoping.” Telescoping
occurs when we are asked questions about our behavior within a bounded time period, and we
report behavior that took place before the period in question. “Constant wave responding” is
another strategy that we employ when our memory is poor and we are asked to reconstruct
events (Tourangeau & Bradburn, 2010). Since we typically have the best recall for the most
recent period, we may reconstruct earlier periods by extrapolating backward from the most

&$!!

!

recent period. For example, if asked how often we recycle, we may think only about how often
we recycled this week, or even today, and then report that figure as if it were a constant. This is
of particular interest in efforts to monitor pro-environmental behavior as most of these pro-social
acts do not take much time or effort and are easy to forget. Thus a strategy of anchoring our selfreports on our most recent behavior is an attractive one, and may even result in underreporting.
The implications of recall error are profound.

Regardless of whether self-reports of

individuals’ past pro-environmental behaviors are socially biased, they are likely skewed from
the very instance of encoding, which may have taken place long ago. There are two possible
solutions to this faulty encoding that Brown and Kasser (2005) argue in favor of and that the
current study hoped to incorporate. The first is direct monitoring (of behavior), and the second is
decreasing the time between the event (or behavior) and the act of the survey question. A study
that uses ESM and the Apple iPhone as method and instrument respectively—as will be
discussed in the next chapter—may even in some cases be able to reduce the recall time to the
point that the event may still be retained in an individual’s working memory.

1.4.3. ESM and the Apple iPhone: In Response
Kasser and his colleagues’ (e.g. Kasser, 2010, in press; Brown & Kasser, 2005) work
examining Schwartz’s (1994) value orientations, subjective well-being (SWB) measures and proenvironmental behavior demands additional attention, not only because of its theoretical value
and the empirical evidence in support, but also because of the urgency of many of the
environmental dilemmas humans face. Nevertheless, this thesis and the current study are not
attempts solely to validate or invalidate that work. Instead this project attempts primarily to
accomplish three objectives, only two of which are specifically complementary to Kasser’s

&%!!

!

(2010) work.
First and foremost, the current study is an application of the joint use of ESM and the
Apple iPhone as a socio-environmental survey method, the implications of which may be
profound. Next, the study uses that application to examine Kasser (2010, 2011) and Schwartz’s
(1994) theoretical framework regarding value orientations, SWB and pro-environmental
behavior. And third, in doing so, both the method and the instrument are presented as an
ameliorating response to the two errors and biases listed in Section 1.4, i.e. socially desirable
responding (SDR) and recall error.

Each of these objectives, especially the first, will be

discussed in the following chapter, and the current study will then be presented in Chapter Three.

&&!!

!

Chapter Two:
The Experience Sampling Method and the Apple iPhone
2.0. Introduction
Chapter One concluded with Brown and Kasser’s (2005) call for additional research into
values, well-being and pro-environmental behavior. It also claimed that ESM, along with the
Apple iPhone, was a complementary method that may be able to ameliorate the effects of
socially desirable responding (SDR) and recall error. This chapter introduces ESM and its
advantages and disadvantages, and examines how researchers might use ESM, along with the
Apple iPhone, to investigate the three variables listed above.

2.0.1. Introduction to ESM
It is generally considered that Mihaly Csikszentmihalyi and Reed Larson created the
Experience Sampling Method at the University of Chicago in 1975. At the time it was both an
effort to satisfy a need and a useful adoption of the latest technology. The need derived from
what Csikszentmihalyi and Larson (1987: 526) identify as a “large body of research
demonstrating the inability of people to provide accurate retrospective information on their daily
behavior and experience” (e.g. Yarmey, 1979).

It was also an attempt at systematic

phenomenology, which, inspired by the German philosopher Edmund Husserl’s (1962) pure
phenomenology, attempted to reliably measure the events that occurred in individuals’ stream of
consciousness over time (Hektner et al., 2007). As opposed to the behaviorists who only
measure overt actions and consign all mental processes to a “black box,” phenomenologists on
the other hand examine only the mental processes (ibid). Yet systematizing a method in which to

&'!!

!

do so would be difficult—Csikszentmihalyi has spent the last thirty years attempting to do so
(Csikszentmihalyi, 2000).
Perhaps surprisingly, beepers were the technology that allowed for this type of
investigation. Beepers allowed researchers to alert respondents and elicit self-report data at
randomized points in time (Csikszentmihalyi & Larson, 1987). Respondents would receive a
page and then enter their answer to the proposed question in a journal or research booklet. As
opposed to the diary method, which was in wide use before the initial ESM studies and provided
“dry and generalized” results “without much discrimination” and “according to predictable
scripts,” ESM typically generates rich and unique data (Hektner et al., 2007: 34). Beepers were
eventually replaced by mobile-phones, but only as signaling devices, and PDAs (personal digital
assistants, e.g., Palm Pilots), often only as recording devices.
Not only did ESM both meet a need and need a new technology, but it was also
purported to meet that need while addressing others in socio-psychological research. ESM is a
“means for collecting information about both the context and content of the daily life of
individuals,” a way “to capture daily life as it is directly perceived from one moment to the next,
affording an opportunity to examine fluctuations in the stream of consciousness and the links
between the external context and the contents of the mind” (Hektner et al., 2007: 6). Its more
general purpose is to study subjective experience and ensure ecological validity, capture
respondents in various life situations and record their psychological reactions (Csikszentmihalyi
& Larson, 1987). It attempts to identify regularities in the stream of consciousness and then
relate these regularities to characteristics of the person (e.g. age, aptitude, physiological arousal,
well-being), of the situation (e.g. challenges of a job, being a university student), or of the

&(!!

!

interaction between the person and the situation (e.g. the dynamics of one’s personal
relationships) (ibid).
ESM also addresses a number of the shortcomings found in alternate methods aimed at
capturing similar phenomena. Csikszentmihalyi & Larson (ibid: 527) list a few. First, scientists
engaging in laboratory studies that ask individuals to evoke feelings, behavior and imagery often
receive responses that are not typical of experience encountered in real-life situations. Second,
in quality-of-life studies, complex phenomena that are often temporally or spatially contingent
can only be presented as global assessments. Third, the data is often gathered in retrospect, and
outside the context of the life-situation, and thus permits distortions and rationalizations to
become important.

Finally, studies of time-budgeting often present unclear links between

behavior and psychological states.
In attempting to address these shortcomings, Scollon et al. (2003) argue that ESM has
over the decades begun to excel in five areas especially. First, ESM allows researchers to better
understand the contingencies of behavior. Second, ESM removes the respondent from the lab
and places her back into real-life situations. Third, ESM allows for the investigation of “withinperson processes” (ibid: 8). Fourth, ESM avoids memory and recall bias and the use of
heuristics. And finally, ESM alleviates the need for multiple methods to study psychological
phenomena.
This thesis and the current study suggest that ESM is able to go beyond merely the study
of social-psychological phenomena and assist in economic and socio-environmental research as
well. The rest of this chapter will continue to examine tenets of ESM and discuss the advantages
that the adoption of current broadband mobile-phone technology, such as the Apple iPhone,
affords future researchers.

&)!!

!

2.1. Types of Experience Sampling
Three distinct types of experience sampling exist (Scollon et al., 2003). The first is called
interval-contingent sampling and refers to the type of data collection in which participants
complete self-reports after a designated interval for a pre-set amount of time (e.g. Nowlis &
Cohen, 1968). Event-contingent sampling occurs when participants complete self-reports when
a pre-designated event occurs (e.g. reporting after a physical activity, sport, or social interaction,
Cote & Moskowitz, 1998). The last, and what will be the primary focus of this thesis and the
current study, is signal-contingent sampling, which requires participants to complete self-reports
when prompted by a randomly-timed signal.
Scollon et al. (2003) also acknowledge that there are three other similar types of study
that are distinguishable from ESM based on the type of data generated. The first is called
thought-sampling, and was developed by Hurlburt (1997) and independently by Klinger (19781979). Thought-sampling focuses on recording one’s inner thoughts and generally dispenses
with external events.

Hurlburt (1997) distinguishes descriptive-experience sampling; this

sampling is used only for gathering qualitative data and typically forgoes any of the
psychometric scaling that is commonly found in ESM studies. Stone et al. (1999) distinguish
ESM from ecological momentary assessment, or EMA. EMA measures both the “participant’s
momentary subjective experience” and “the elements of the environment related to the
momentary experience” (Scollon et al., 2003: 8). This last distinction is likely the most tenuous,
as many ESM studies are EMA studies.

&*!!

!

2.2. Resource Requirements of ESM
ESM has for most of its history required a great deal of resources, i.e. time, money and
human capital. As opposed to single administration surveys, ESM studies require researchers to
engage, sometimes continuously, with their respondents for days and weeks, if not months or
years. This engagement is often in the form of signaling, which as stated above relies on an
electronic instrument that emits stimulus signals according to a random schedule. For most of
ESM’s history, beepers or pagers have been responsible for this signaling. However more
recently ESM has relied on mobile phones as signaling devices and PDAs (personal digital
assistants) as recording devices. For the current study the Apple iPhone, a broadband web-ready
mobile-phone, was used, both as a signaling device and a recording device. The advantages of
such a device that can serve as both a signaling and recording device will be discussed in later
sections.
Due to this high level of resources and commitment, it has been argued that ESM studies
are not appropriate for all types of investigation. ESM is designed to capture individuals’
representation of experience as it occurs, within the context of daily life, and thus is best suited
to measure dimensions of experiences that are contingent on context, i.e. time, place, space and
activity (Hektner et al., 2007).

Thus global and retrospective questions are not generally

considered useful or cost-effective. However as mobile technology becomes more popular and
more advanced, the costs of conducting such a study are decreasing, making examinations that
were once thought appropriate for single-administration surveys suitable for ESM-investigation
(e.g. Killingsworth & Gilbert, 2010).

&+!!

!

Kahneman and Krueger (2006) provide a summary of the typical ESM experience from
the viewpoint of the respondent. In this example the participant is using a handheld computer or
PDA (ibid: 9):
“Participants in experience sampling studies carry a handheld computer
that prompts them several times during the course of the day (or days) to answer a
set of questions immediately. Participants are shown several menus, on which
they indicate their physical location, the activities in which they were engaged
just before they were prompted and the people with whom they were interacting.
They also report their current subjective experience by indicating the extent to
which they feel the presence or absence of various feelings, such as feeling angry,
happy, tired and impatient.”

The above description is derived from research that Kahneman and Krueger (2006) and
Kahneman et al. (2004) conducted in which they attempted to supplant ESM with an alternate
version entitled the Day Reconstruction Method, or DRM. DRM relies on a combination of
experience-sampling and time diaries and was “designed specifically to facilitate accurate
emotional recall” (Kahneman & Krueger, 2006: 10). Kahneman (ibid) and others argue that
DRM, because it is cheaper and easier than ESM and provides similarly accurate data, is the
preferred method. Hektner et al. (2007: 277) however argue that DRM is subject to the same
biases that more standard diary-reliant methods are, and that DRM very well “might provide
information that distorts reality.”
What is perhaps lost in such an argument over DRM and ESM is that technology such as
the Apple iPhone is capable of reducing the cost of ESM studies to a point that make Kahneman

&,!!

!

and Krueger’s (2006) concerns unnecessary (e.g. Killingsworth & Gilbert, 2010). And the
adoption of broadband mobile-phones not only decreases financial costs. Previous ESM studies
required researchers to provide significant amounts of instruction to respondents regarding the
signaling device and the ESM form, or the booklet in which the respondents were to record their
answers. While respondents surely still require an introduction to the study, its intentions, and
instructions on how to respond to signals, using mobile-phones as both signaling and recording
devices reduces the amount of instruction required by a considerable degree—presuming the
respondent is already knowledgeable about using certain features of her phone. There are other
advantages to reducing the necessary introduction and instruction, namely, reducing the impact
of the study on the respondents’ day-to-day activities. The benefits of such a move will be
discussed later.

2.3. Reliability of ESM
2.3.1. Internal Validity
Two fundamental questions must be addressed by any research study. Are confounding
variables controlled so that the conclusions are warranted? And can generalizations be made
beyond the individuals and moments sampled? (Hektner et al., 2007: 104). These are questions
of internal and external validity, respectively.
ESM, because it focuses on everyday life, emphasizes external or ecological validity over
internal validity (ibid). That does not mean that ESM studies lack internal validity. Internal
validity is defined as the correlation between a measure and the true value of the attribute, taken
on a set of individuals (Groves, 1987). It is “whether a difference exists at all in any given
comparison…whether or not an apparent difference can be explained away as some

'-!!

!

measurement artifact” (Webb et al., 1966: 10-11, quoted in Bailey, 1994). Evidence suggests
that ESM studies are more internally valid than single-administration surveys (SAS); however
Csikszentmihalyi & Larson (1987: 531) admit that because the measure is designed to measure
the effects of life situations on psychological states, “perfect reliability would in fact defeat
[ESM]’s purpose.”
ESM studies benefit from, as Zuzanek (1999) and Kubey et al. (1996) argue, the
immediacy of the questions being asked, which reduces recall error and bias, along with the
tendency to answer questions using socially desirable responding (SDR) (see Section 1.4.1) for
an explanation of SDR and SDR bias). Recall error is of particular concern to psychologists,
who argue that monitoring individuals in situ is necessary versus asking a participant to recall an
event a week or more after the fact (Hicks et al., 2010). These errors can be caused by the
emotional state of the participant, the length of time the participant is asked to recall, or simply
by the foreign environment in which the participant is located (ibid). Furthermore, these errors
have been shown to be systematic (Shiffman et al., 1997). For a more in-depth analysis of recall
error see Section 1.4.2.
The randomness of ESM signals also reduces reflexivity bias, or attempts of the
respondents to discern the purpose of the research and tailor their responses accordingly (also
reducing SDR). Hektner et al. (2007) also argues that because of the large number of signals
received, the respondents often habituate to the recording of their personal behavior, and thus
forget to some extent that they are being monitored or engaging in a research project. Overreporting and underreporting are also more likely in SAS (e.g. Lee, 2005), and events and
behaviors deemed less important by respondents are often ignored, like idling—or doing nothing

'$!!

!

at all, which accounts for a greater amount of time than perhaps thought previously
(Csikszentmihalyi & Larson, 1987).
With a method that some might infer as intrusive, there are of course concerns over
reactivity, or the phenomenon of respondents answering questions accurately, but changing their
behavior because of their participation in the method. Hektner et al. (2007) present evidence
suggesting that these concerns are unfounded. Csikszentmihalyi and Larson (1987) found that
the vast majority of respondents reported having a normal week and that ESM “captured their
week well” (Hektner et al., 2007: 106). Hormuth (1986) found that among German adults only
14 percent considered the signals a bother; however 22 percent said ESM disrupted their daily
routine.

Larson and Richards (1994) found that over half of the families they surveyed

responded, “not at all” to the question, “Do you think the family’s week was different because of
the study?” And Hufford et al. (2002) found that undergraduate problem drinkers maintained the
same level of drinking before, during and after their ESM study period.
Using Apple iPhones as both a signaling and recording device in ESM studies likely has
a mitigating effect on each of the errors and concerns introduced above. Simply merging the two
devices, signaling and recording, quickens the participants’ ability to respond and reduces the
amount of hardware, as well as the cognitive load, of participating in the study. While this
increase in efficiency may have negligible effects on recall error—as compared to other ESM
studies, it surely aids in decreasing reactivity bias. Moving from beepers and research booklets,
to PDAs, and finally to mobile-phones certainly decreases the real and perceived intrusion of
ESM studies on respondents’ daily activities and impressions of normalcy. And because most
individuals carry their mobile-phones on their persons at all times, concerns over delayed
responses or missed signals should be mitigated as well.

'%!!

!

2.3.2 External Validity
Webb et al. (1966: 10-11, quoted in Bailey, 1994) argue that, “external validity is the
problem of interpreting the difference [in any given comparison], the problem of generalization.
To what other populations, occasions, stimulus objects, and measures may the obtained results be
applied?” Hektner et al. (2007: 107) identify that, unique to ESM, the “population” that Webb is
speaking about can be thought of in two senses: “as a collection of people or of moments of
experience.” While all researchers and their projects must consider how well their sample
represents the sampled population, ESM researchers must also consider how well their questions
represent the population of lived experiences (ibid). Often this is done through comparing ESM
and non-ESM studies. Diaries can be compared, as Csikszentmihalyi and Graef (1980) and
Robinson (1985) have done. Hoover (1983) compared ESM reports and the readings from heartrate and activity monitors.
Often situational validity is used instead of standard external validity measurements.
Situational validity is derived from examining the internal logic of a reported situation, the
convergence of time, context, and activity (Hektner et al., 2007).

For example, personal

grooming times are likely to be reported at predictable times of day, and participants report
talking when they are known not to be alone (ibid). When it comes to measuring internal states
of mind, which was of course ESM’s original intention, it must be noted that
phenomenologically speaking only the individual can truly know whether her response is an
accurate representation. However, again, situational validity is often used. People are expected
to report internal states that make sense when compared to the activities reported. Hektner et al.
(2007: 113) argue that because most ESM studies do report situationally valid results (e.g.
Csikszentmihalyi & Larson, 1984; ibid, 1987; Csikszentmihalyi & Schneider, 2000; Wells,

'&!!

!

1985), such as being relaxed while watching television, this easily recognized “normal” or
“obvious” representation of the world speaks to ESM’s validity.
Hektner et al. (2007) also argue that a common use of ESM instruments is to validate
non-ESM measures of aggregated person-level characteristics or traits. Such use of ESM was
the primary motivation for the current study, which examined value orientations, proenvironmental behavior, and individuals’ subjective well-being. All were hypothesized to be
time and context-dependent, as well as subject to recall error and faulty generalization by
respondents in single-administration surveys.
Admittedly, while mobile-phones and iPhones in particular have a great deal to
contribute regarding internal validity concerns, they have less impact on the external—and
situational—validity of ESM studies. This is because the questions asked would likely remain
unchanged regardless of the specific technology being used—though not necessarily. One area
in which iPhones can assist is in the linking of time and context to the internal state or behavior
that is the subject of inquiry. This can be done primarily through the use of the mobile-phone’s
camera. Taking pictures, while increasing the amount of data analysis and perhaps the amount
of coding required by the researcher and her assistants, does help to erase doubts or the vagaries
found in individuals’ self-reports.
For example, Hektner et al. (2007) provide an example of two individuals giving
conflicting reports about their being with their partner. A husband reported to be alone, while
the wife reported being with her husband. While taking a picture would certainly increase the
burden and the invasiveness of any ESM study, it is possible that having the husband and wife
take a picture of any persons in their space may have provided a unique answer from that which
was previously reported. While the husband and wife may continue to consider their purported

''!!

!

state of aloneness differently, one’s picture of the other would give the researcher an objective
picture of the situation.

2.3.3. Ecological Validity
Ecological validity, while similar to external validity, varies in important ways.
Ecological validity is the extent to which findings can be generalized to the naturally occurring
situations in which the phenomenon that is investigated takes place (Brunswik, 1949).

This

generalization is not only spatial, but also temporal and experiential. ESM allows for the
researcher to be able to investigate the various situational contingencies of any psychological
phenomenon; in doing so ESM allows her to ecologically validate her theoretical concepts and
empirical findings in real life settings (Scollow et al., 2003). It may be easier to think of
ecological validity as external validity reversed. If ensuring external validity is ensuring that
specific results can be generalized, ensuring ecological validity is ensuring that generalized
results hold up in specific situations and contexts, in situ. For example, if Csikszentmihalyi’s
(1988) theories of flow and optimal experience are right, then at certain points of each day and
during specific activities, most humans are more cognitively engaged and more highly motivated
then at other points. Thus the time, location and specific act of being surveyed would have a
considerable effect on an individual’s responses to questions regarding his mood.
The ecological moment in which an individual finds himself does not only affect his
mood. It may also have a significant effect on his beliefs, values and reported behaviors,
especially environmental attitudes and pro-environmental behavior.

For example, single-

administration research instruments can only measure individuals’ concerns regarding pollution,
reductions in green-space, or climate change in the context of the survey-act. This type of

'(!!

!

engagement necessarily generalizes the individual’s response, and leaves out the emotional
components of past experience (Hektner et al., 2007: 23). If asked about her level of concern
regarding pollution, the individual must remove herself from the social, cultural or even
biological context in which that concern more typically materializes. For example, if she is in a
laboratory or participating in an online survey, she is likely not able to consider all of the
variables and constraints that contribute to her level of concern. On a scale of 1-7, with 1 being
“not concerned at all” and 7 being “extremely concerned,” she may answer “6.” However were
she to answer while sitting in a boat on a sunny day in a lake that is isolated from other people
she may answer “4.” Were she sitting in traffic on a hot, humid day in southern California, she
may answer “7.” None of these responses is inaccurate or wrong. Instead they capture the
natural oscillations that occur in each of our daily lives. Yet single-administration surveys miss
these oscillations. ESM on the other hand allows us to examine the context as well as the
content of our attitudes and behaviors.
As regards phenomena like personal well-being, ESM allows researchers to examine not
only the accuracy but also the robustness of individuals’ stated measures of well-being. Instead
of asking an individual to generalize his current level of satisfaction (as regards his daily life),
one can ask him to report his level of satisfaction at multiple times and places and during
different activities during the week. The researcher can then investigate whether or not these
levels, which are affected by any number of contexts and constraints, predict the individual’s
general level of satisfaction. Furthermore one can investigate what exactly those contexts and
constraints are, as well as what the specific contributors to that level of satisfaction are, as was
done in this study. In this way researchers can get a better picture of not only how individuals
behave, but also why individuals behave and believe the way they do.

')!!

!

2.3.4. A Focus on Intra-Personal Variation
Hektner et al. (2007: 9) admit that scientists are “justifiably leery of putting too much
credence in what people say about themselves, especially when they describe their thoughts and
feelings.” They also argue however that subjective experience is the most objective datum we
have access to and “trying to reduce it to a more objective standard only decreases its
objectivity” (ibid: 10). They argue that if somebody reports that she is happy, then she is happy.
But what exactly “happy” means to that individual is difficult to determine without examining
intrapersonal, as opposed to interpersonal, comparisons.

Scollon et al. (2003) agree that

psychologists—amongst others—have called for more idiographic research, i.e. intrapersonal
research, over nomothetic, i.e. interpersonal research, since the 1990s (e.g. Lamiell, 1997;
McAdams, 1995; Pelham, 1993), yet the latter has formed the mainstay of research. ESM
however allows researchers to investigate within-person patterns that may mask patterns that are
not visible at mean levels (Scollon et al., 2003: 9).
For example an individual might report that she is happy both while driving a car and
while watching television. This may not tell researchers much of anything. However were she
to report on a scale of 1-7 with 7 being “extremely happy” and 1 being “completely miserable”
that while driving a car she was a “4” and while watching television she was a “6,” then
researchers would know a great deal more about her definition of happiness, or at least what
experiences are representative of her happiness.

At a within-person level, researchers are

primarily interested in what states go together at any given moment, and not only between
emotions, such as guilt and happiness as Scollon et al. (ibid) suggest, but also between activities
and emotions, and activities and attitudes. In the previous example, our subject is obviously
happier watching television than she is happy driving. Knowing this allows researchers to then

'*!!

!

construct a “Z-score” or a mean score on a particular variable and judge how she deviates from
the mean in different circumstances (Hektner et al., 2007: 11).
An interesting example of the benefits of ESM intrapersonal investigation is a study done
by Oishi et al. (2002) in which the researchers found that across cultures, being with friends
increased pleasant mood and decreased negative mood. Other global affective traits existed
among samples, but the degree to which situations exerted an influence on the absolute level of
affective experience varied across cultures. The benefit of being with others was qualified by
culture and gender such that people in some cultures, as well as men in general, received a
greater boost from being with friends.
ESM has also been used to investigate the role of resources in subjective well-being
examinations (e.g. Diener & Fujita, 1995), something the current study does in its investigation
of the effect of well-being and values on pro-environmental behavior and attitudes.

2.4. Critiques of ESM
ESM is obviously not without its disadvantages, as discussed by Kahneman and Krueger
(2006) and others above. Hormuth (1986) argues that ESM studies may place too great a
responsibility on participants to accurately report the objective features of each moment. They
may lack the ability or become ignorant to such features over the duration of the experiment.
There are also concerns over reactivity. While reactivity, a methodological “confound that
occurs when respondents report accurately but change their behavior or internal experiences as a
direct result of their participation in a method” (Hektner et al., 2007: 106), is a concern, the use
of broadband mobile-phone technology, as suggested above (in Section 2.2) should work to
mitigate such concerns.

'+!!

!

The sharpest critique of ESM is the rate of data loss through refusal, non-response or
attrition. As stated above, gaining a sample that is large enough to be representative and
generalizable is difficult because of the intrusive and burdensome nature of some ESM studies.
Lawton (1999) argues that high volunteer rates and signal response rates are difficult to attain
and the self-selection bias of those who do volunteer needs to be considered. For example,
studies have shown that those individuals who volunteer for ESM have higher grades than those
who do not (Mulligan et al., 2000) and college students who volunteered were less anxious, less
likely to employ pathological defensive styles, and more well-adjusted than those who refused to
participate (Waite et al., 1998).
Scollon et al. (2003: 14) identify self-selection bias as one of the most difficult hurdles
facing ESM studies, arguing that ESM studies are indeed onerous and the alarms may “disrupt
one’s activities, conversations, and work, and may not only annoy oneself but surrounding others
as well, such as in church, classrooms, or meetings.” They add that filling out the forms may
also take over the course of an entire study one hour or more. They inquire as to who would
volunteer for such a study, arguing that ESM studies may over-represent participants that may
show greater motivation, conscientiousness and agreeableness, participants who have more time
(e.g. unemployed people or college students), and participants who are familiar with and
physically able to respond to signal technology.
While these are all legitimate concerns, the introduction of broadband technology may
alleviate a large number of them. As will be presented in Chapter Three, data-loss and refusal
attrition rates were low to non-existent in the current study. Furthermore, using participants’
own mobile phones makes the survey process much less invasive and reduces the probability that
participants will stop carrying the research instrument at all times—as opposed to studies which

',!!

!

use PDAs or research journals. And while monetary incentives may certainly improve volunteer
rates (e.g. Lynn, 2001), one of the focuses of the current study was to intentionally present the
examination as novel, enjoyable and minimally invasive. This was hoped to entice individuals
who may have forgone a typical ESM-style project. Additional research is necessary in order to
confirm that ESM studies using broadband mobile-phones entice less motivated or less willing
participants.

2.5. ESM Embracing Technology
For the past two decades ESM has relied primarily on signaling technology, or alerting
participants via pagers or telephone calls. Participants would then answer by writing their openor closed-end answers in a journal or research booklet. Over time, written answers have been
exchanged for computerized responses as is demonstrated in Kahneman and Krueger’s (2006)
portrayal above, and a few social scientists have moved from using personal data assistants
(PDAs) to using mobile phones (e.g. Hicks et al., 2010; Killingsworth & Gilbert, 2010). Such
advances have allowed researchers to gather data about the specific time and place respondents
receive signals, drive triggers and sensors, i.e. ask questions at specific times and places (Hicks
et al., 2010), and have participants respond to signals more easily and accurately.
To this point however, there have only been a few forays into using broadband mobilephones themselves, such as the Apple iPhone, to study socio-psychological phenomena.
Killingsworth and Gilbert (2010) used ESM and the iPhone to study mind-wandering and its
deleterious effects on individuals’ happiness, and Karapanos et al. (2010) used the DayReconstruction Method, an alternate form of ESM that asks participants to keep a diary of their
day’s activities, to examine individuals’ user-experiences with iPhones—this study combined

(-!!

!

DRM and iPhones, but did not use iPhones as the survey instrument. Just last year Hicks et al.
(2010) developed a Droid-powered personal data collection system called AndWellness. This
system has yet to be deployed.
Most of the literature regarding the use of mobile phones in socio-psychological research
does not reside in socio-psychological or methodological journals, but instead resides in
computer science, information systems, and consumer research journals. As a result the focus
remains on the development of the mobile technology and the purported advantages such
technology will afford social scientists. Little empirical evidence or supposition however exists
regarding the second-order advantages, or the advantages of these advantages.

2.5.1. Larger Sample Sizes
The most immediate advantage of using broadband mobile-phones in ESM study is the
ability to acquire much more data at a significantly reduced cost. This is not to imply that costsavings is the key advantage, but instead what that cost-savings allows.
Killingsworth and Gilbert (2010) acknowledge that laboratory experiments have revealed
a great deal about the cognitive and neural bases of mind wandering, the aim of their latest ESM
study; however, little has been discovered about its emotional consequences as they affect
individuals in their normal day-to-day activities. “The most reliable method for investigating
real-world emotion is experience sampling,” they admit; however, “unfortunately, collecting
real-time reports from large numbers of people as they go about their lives is so cumbersome and
expensive that experience sampling has rarely been used to investigate mind wandering” (ibid:
932). And any investigations that have occurred have been limited to very small samples.

($!!

!

Instead of using a different method however, one that lacked both the high internal and
external validity of ESM, the authors developed a Web application for the Apple iPhone. They
then used this app to create a large database of real-time reports on thoughts, feelings and
actions. The database at publication contained nearly a quarter of a million samples from about
5000 people from 83 different countries ranging in age from 18 to 88 and who, collectively,
represent 86 major occupational categories (ibid: 932). This is a staggering amount of data, but
even more so for an ESM study, which when relying upon PDAs or research journals, often
samples between 50 and a few hundred individuals.
Again, it is not simply the amount of data this Web application provided that is
intriguing, but also the new arenas of research that it suggests are available for ESM-style study.
ESM need no longer simply be a tool for verifying non-ESM data, but instead, because of the
reduced costs that mobile-phone technology provides, can be a cost-effective alternative to nonESM studies altogether.
The current study, while not nearly as expansive or exhaustive as Killingsworth and
Gilbert’s (2010) study, suggests that ESM may be a cheaper and easier method to examine proenvironmental attitudes and behavior than administering an SAS.

2.5.2. Flexible Survey Instruments
Another advantage to incorporating technology into ESM studies is real-time data
collection and analysis. The Killingsworth and Gilbert (2010) study collected—and continues to
collect—data in real-time; however, they did not analyze it at that time in order to adjust the
instrument.

The advantages of such immediacy cannot be overstated.

Because ESM is

necessarily longitudinal in format, instantaneous data collection and analysis allows for the

(%!!

!

creation of a flexible research-instrument. Researchers cannot only monitor participants as if
they were typing their answers into a computer in a laboratory, but they can also alter the
composition of the survey instrument so as to personalize it to each respondent or time period—
analysis concerns of course must be considered. While computer programming allows for such
on-the-fly composition, because SAS examinations typically “exist” for only a few minutes or
hours, adjustments generally cannot take into account contextual variables like time or social or
cultural events, especially those that are unpredictable.
Using mobile phones as both the signaling and recording device allows for such
incorporation. In studies such as the current one, researchers are able to ask questions regarding
pro-environmental attitudes and behaviors and well-being while at the same time incorporating
difficult-to-predict weather-events like blizzards or heat-waves and social-events like the
Japanese tsunami and nuclear near-catastrophe (in 2011). For example, researchers could ask
individuals about their beliefs regarding climate change during a heat wave in March or during a
random cold-spell in July, and these events would likely have a considerable effect on the
resulting data (see Section 2.3.3 on Ecological Validity). Yet coordinating the study so that the
specific “hot” day of the week or even the “cold” week of the year coincides perfectly with the
actual weather event would be impossible for non-ESM or DRM studies. Having control over
the survey instrument provides researchers with the ability not only to guess better, but also to
investigate multiple events in single research periods and sample populations. This results in
ESM researchers not only being able to examine the reliability of non-ESM studies, but also to
study social and environmental events and individuals’ reactions to each in the same way that
ESM researchers previously examined the effect of different physical or behavioral states on
individuals’ internal or psychological states.

(&!!

!

Such a marriage of ESM and broadband mobile telephones creates a great deal of lowhanging fruit for not only socio-psychological researchers, but also for those researchers who
have relied primarily on other methods and technologies.

2.5.3. ESM as a Hybrid Technology: Stated and Revealed Preferences
Using mobile technology such as Apple iPhones to implement ESM studies increases the
number and variety of arenas appropriate for experiential research. Such devices also allow for
ESM and ESM-style approaches to move beyond personal experience and psychological and
phenomenological investigation. The current study is an example of an investigation that, in line
with previous studies, examines both the content and context of individuals’ psychological states,
e.g. their well-being, and hybridizes the repeated-interaction approach of ESM to investigate
individuals’ behavior and values.
In economics terms, ESM and mobile phones allow for investigations of both stated and
revealed preferences, and may provide greater detail and accuracy by accounting for the
ecological context of both. Research abounds that relies upon one or the other or, much more
common recently, the combination of both stated preference (SP) and revealed preference (RP)
models. Both types of experiments have been the subject of considerable criticism. SP data has
been critiqued because it is thought consumers react differently to hypothetical experiments than
they would were they facing the same alternatives in a real market (Brownstone et al., 2000);
also, actual behavior is not observed (Adamowicz et al., 1994). Another issue is that some
attributes for new products might be too novel for consumers to completely understand them,
and socially desirable options, a particular problem in socio-environmental research, may cause
respondents to misrepresent their choices (Brownstone et al., 2000).

('!!

!

RP methods, or indirect methods (SP methods are considered direct methods), are not free
from critique either. They often suffer from high co-linearity and limited variation among
attributes in real markets (ibid), and cannot be used to examine non-use values (Adamowicz et
al., 1994). Combining RP and SP is a way to avoid many of these concerns; Adamowicz’s et al.
study of recreation sites used RP to determine the respondents’ market behavior, and then an SP
model to explain the choice of one alternative over the other as a function of the attributes, e.g.
travel distance. Brownstone et al. (2000) in their study examined both SP responses and RP
behavior in two waves, with the SP data preceding the RP observations by 15 months. Both
waves of data were gathered via survey, either telephone or mail.
Certainly these joint-model studies provide data that offer advantages over SP or RP
studies alone. Yet in doing joint-model studies, researchers face higher costs and higher resource
requirements. Many of these types of studies are longitudinal or require the use of multiple
samples. The validity concerns regarding the use of multiple samples are known and require
considerable effort on the part of the researcher to mitigate. ESM when combined with iPhones
or other broadband mobile technologies offer a low-cost alternative or at least can supplement
many joint-model studies. No longer do all researchers need to rely on conducting separate
investigations—admittedly many researchers still do and would not benefit from ESM-style
studies. However, for many ESM surveys allow for simultaneous investigations, both surveying
to acquire stated preferences and monitoring market behavior to acquire revealed preferences.
Such investigations also allow the researcher to leave the laboratory behind. Regarding
contingent valuation studies, surely the context of both the question within the survey and the
context of the individual within his daily activities have significance. ESM and mobile phones
used jointly have two advantages over single-administration surveys. They resist part-whole

((!!

!

effects by isolating the questions asked from each other and can, through the use of geographic
and temporal triggers, potentially isolate and align the individuals’ experiences purposely with
specific questions (Hicks et al., 2010). Instead of asking individuals to recall their purchases
from three days prior, or think back to how often they recycle in a week, or imagine how much
they would be willing to pay for mass-transit, researchers using ESM and mobile phones can ask
participants to describe or photograph their purchases in real-time, text each time they recycle an
item over the period of a week, or relate willingness-to-pay (WTP) for mass-transit during a
rush-hour commute (remember that texting while driving is illegal in Michigan!)

2.6. ESM and Technology: Expanding Beyond Previous Studies
Hektner et al. (2007: 125) argue that ESM research centers on “the contexts of daily life,
the experiential content of life, and the links between context and content.” They add that a third
dimension is the specific characteristics of the sample of participants. This is obvious in the
decades of ESM studies Hektner et al. describe in their text. Each study ranges from monitoring
a specific behavior, like sleeping (Stickgold et al., 2001), climbing mountains (Delle Fave &
Massimini, 2003), or listening to music (Sloboda et al., 2001) to monitoring a specific
population, like widows aged 69 or older (Hnatiuk, 1991), teenagers (Csikszentmihalyi &
Larson, 1984) or college students with concealable stigmas (Frable et al., 1998).
No other group has been studied more with ESM than adolescents and teenagers (Hektner
et al., 2007: 133). Csikszentmihalyi & Larson (1984) examined adolescents internal and external
worlds, measuring what they were doing, how often they were doing it, where, whom with, and
the effect that their behavior had on their internal states of mind. Their findings showed that
teenagers follow “astonishingly similar patterns,” and those patterns vary little based on age, sex,

()!!

!

and social class (ibid: 78). Futhermore, certain activities, such as sports, games, hobbies and
conversations merged high motivation with opportunities for learning cognitive discipline (ibid).
Only ESM is able to monitor these psychic peaks and valleys that occur throughout
individuals’ days. And Csikszentmihalyi has used decades of these studies to develop what he
calls the theory of “flow experience,” or “intrinsically motivated states of consciousness,” which
often lead to discoveries of new ideas, artifacts and technologies (1992: 29). Csikszentmihalyi
argues that these optimal states allow for the “evolution of consciousness…the evolution of
culture…and ultimately the evolution of the human species,” all of which hinges on our
“capacity to invest psychic energy in goals that are not modeled exclusively on the teleonomy of
genes or culture” (ibid: 29). Teleonomy means simply “that the person identifies his or her goals
with the genetic instructions programmed in his or her organism” (ibid: 24).
While these flow, or optimal experience, studies are obviously outside the purview of this
thesis, any investigation of ESM would be incomplete without their mention. Furthermore, it is
this theory of peaks, valleys and flows that makes ESM and ESM-like studies of values and
behavior, particularly values that are linked to environmental behavior, so interesting. This
thesis argues that such a belief in context and content, or ecological validity, makes the inclusion
of ESM studies in socio-environmental research necessary. And ESM works best when in
conjunction with other methods, argue Scollon et al. (2003).
With this in mind, it is important to note where else ESM-type studies have and are
evolving. While certainly most ESM efforts are spent in socio-psychological investigations of
specific groups or specific activities and remain paper-and-pencil or PDA driven (e.g. McCance,
2010; Verduyn, 2009), some of the most recent studies using ESM, aided by broadband mobilephones, (e.g. Killingsworth & Gilbert, 2010) or those developing the technology (e.g. Hicks et

(*!!

!

al., 2010; Uy et al., 2010), are expanding beyond purposive samples. This, according to Zuzanek
(1999), would be a significant challenge. And certainly in 1999 it would have seemed so. Yet
the advance of mobile technology and broadband capable devices has made the once small
sample size and purposive necessity of ESM a thing of the past. Not only increased sample
sizes, but more heterogeneous samples are now possible, as seen in Killingsworth and Gilbert’s
study (2010). Greater heterogeneity both in populations and in activities, events or experiences
means greater generalizability and greater external validity across ESM studies. While ESM
remains an excellent tool for investigating specific populations and activities, it need not be
restricted to those investigations. iPhones and broadband mobile-phones make ESM available to
all the social sciences.
It was with this in mind that the current study was proposed, and conducted. As stated
previously, Brown and Kasser (2005) argued for additional research into values, well-being and
pro-environmental behavior, research that could not only reduce recall error and SDR, but also
move past self-reporting. This thesis argues that ESM, along with the iPhone as its signaling and
recording device, is a particularly good fit for such efforts, despite having never been used
previously in similar investigations. The current research was conducted thus not only because
ESM may be able to address recall error, SDR and self-reporting as was requested, but also
because it may provide complementary benefits to a great number of socio-environmental
investigations. Chapter Three presents the results from the current study in hopes of supporting
such claims.

(+!!

!

Chapter Three:
An Application of the Experience Sampling Method and the Apple iPhone:
Investigating the Effect of Values and Well-being on Pro-Environmental
Behavior
3.0. Introduction
The previous chapter introduced the Experience Sampling Method (ESM), and presented a
number of rationales for incorporating it into the field of socio-environmental research. It also
argued that using broadband-capable mobile phones, the Apple iPhone in particular, would assist
not only in the deployment of ESM, but also in gathering more and more accurate data across
socio-environmental methodologies. Chapter One described a few of the theoretical models used
to investigate pro-environmental behavior as well as identify that behavior’s determinants. The
introductory chapter’s latter half focused specifically on Schwartz’s (1994) theories of clustered
value orientations and Kasser’s (2010) theory that self-transcendent value orientations lead to, if
not at least correlate positively with, increased pro-environmental behaviors—for the rest of
Chapter Three these theories will be jointly referred to as the “value orientation theory of ProEnvironmental Behavior,” or VO-PEB. The chapter ultimately concluded with Brown and
Kasser’s (2005) call for additional research into value orientations and pro-environmental
behavior, in particular research that takes account of recall errors and biases and socially
desirable responding (SDR), and if possible moved past self-reporting.
Chapter Three and the study described herein incorporate ESM and broadband mobilephones and directly respond to Brown and Kasser’s (ibid) call for additional value-oriented
research. It will be argued here that ESM and the iPhone when used conjointly are particularly
adept at reducing recall errors and biases, as well as conducting direct monitoring—as opposed
to relying solely on self-reporting. Great efforts were also taken in the current study to reduce

(,!!

!

SDR by couching the study as an investigation into consumers’ behavior (not participants’ proenvironmental behavior).

While the study’s original title, “Determining the Relationship

between Attitudes, Well-being and Private-Sphere Environmental Behavior,” may suggest
otherwise, the data collected was less specific to participants’ attitudes and much more specific
to their individual value orientations and those factors that most contribute to their subjective
well-being. It also investigated their pro-environmental behaviors, as well as their attitudes
regarding renewable energy, sustainability efforts and climate change.
Chapter Three begins with an overview of the study and its specific rationales and
objectives. The chapter will then present the study’s hypotheses followed by an analysis of the
data collected. The chapter will then conclude by presenting the results, a brief discussion of
their implications, and suggestions for additional research.

3.1. Overview of the Current Study
The data presented in this chapter comes from the study entitled “Determining the
Relationship between Attitudes, Well-being and Private-Sphere Environmental Behavior,” a
project sponsored by the Sustainable Michigan Endowed Project (SMEP) and Michigan State
University. It consisted of an examination of 71 Michigan State University sophomores, all of
whom were individually surveyed using the Experience Sampling Method (ESM) and the
students’ Apple iPhones. These surveys were conducted during four one-week periods across
January and February of 2011. Researchers initiated this study in September of 2010 with three
objectives in mind: first, (RO:1) develop and test a robust, rapidly deployable, and near real-time
instrument for measuring individuals’ attitudes, values, well-being, and private-sphere proenvironmental behavior; second, (RO:2) use the instrument in the examination and interpretation

)-!!

!

of the relationship between attitudes, values, well-being and behavior; and third, (RO:3) attempt
to reduce the errors and biases often found in previous socio-environmental studies, studies
which identified either through explicit means or implicitly through their content their intended
pro-environmental goal, vis-à-vis measures of pro-environmental attitudes and behavior.

3.1.1 Method and Instrument: ESM and the Apple iPhone
In order to achieve these objectives the study used ESM, which asks individuals to
provide responses to both open- and closed-end questions at several random points throughout
each day for the period of one week (Hektner et al., 2007). Instead of asking participants to log
their responses in a journal or enter them into a computer, the study adopted the Apple iPhone as
its survey instrument. The iPhone is not merely a technologically savvy signaling device; it can
act as the interactive medium in which researchers communicate with participants, and the
participants communicate their responses to the researchers. The iPhone and ESM complement
each other well, as was discussed in Chapter Two; however, a brief summary of their conjoint
benefits follows.
One of the critiques mentioned in previous chapters regarding ESM is the burdensome
nature of its investigation. ESM requires a great deal of effort on the part of the respondent (and
the researcher). That effort is lessened by the use of the respondent’s own mobile phone as the
signaling and recording device, as was accomplished in this study. As opposed to carrying a
journal, PDA, or beeper, the respondent need only carry her cell-phone, a device she likely
carries regularly. This study also purposely used SMS messaging and psychometric Likert-scale
questions as the primary mode of question and response. This allowed the participant to quickly
engage and disengage with the instrument and the survey. Other studies (e.g. Hicks et al., 2010;

)$!!

!

Killingsworth & Gilbert, 2010; Uy et al., 2010) have focused on the development of web-based
applications which require respondents to log-in or engage physically and cognitively with the
survey instrument for greater periods of time.
Respondents generally carry their iPhone with them at all times. As a result such an
instrument also encourages more accurate self-reporting of attitudes and activities, phenomena
that researchers consider to be time-sensitive measures. Using iPhones as the survey instrument
also allows for bilateral communication, something that many scholars who study post-normal
science (e.g. Funtowicz & Ravetz, 2003; Batie, 2008; Rittel & Webber, 1973) argue is necessary
in order to pursue—if not achieve—sustainability. While this form of communication was not
used in the current study, much of the media and value-oriented inquiries used in evincing
participants’ attitudes and the specific manner in which they were used would likely work well in
such pursuits. This study certainly suggests that ESM need no longer be one-dimensional—
information can be exchanged between the researcher and respondent and vice-versa, regardless
of either’s location—and may be a great tool in post-normal deliberative research.
The iPhone also allows for more in-depth queries that use and specialize in interactive
multimedia like videos, web-pages, news articles and subject-initiated photographic—and even
video—reports. Finally, and specific to this study, the iPhone was chosen as the primary
technology because it accentuated the purposely deceptive intention of the experiment,
“Determining the relationship between attitudes, well-being and consumer behavior.”

The

deceptive element was key in addressing RO-3, as will be seen below.

3.1.2. Research Objectives
The objectives of this study, listed previously in Section 3.1, are as follows:

)%!!

!

RO: 1—Develop and test a robust, rapidly deployable, and near real-time
instrument for measuring individuals’ attitudes, values, well-being, and private-sphere
pro-environmental behavior.
RO: 2—Use the instrument in the examination and interpretation of the
relationship between attitudes, values, well-being and behavior.
RO: 3—Attempt to reduce the errors and biases often found in previous socioenvironmental studies, studies which identified either through explicit means or
implicitly through their content their intended pro-environmental goal, vis-à-vis
measures of pro-environmental attitudes and behavior.

Each of these three objectives speaks to the particular method that was chosen for this
study. RO:1 called for both ESM as the primary method and the Apple iPhone as the primary
survey instrument.

ESM is not typically used in the examination of individuals’ global

perspectives, such as values, attitudes, or subjective well-being; this is primarily because of the
significant costs typically involved (Hektner et al., 2007). However, by incorporating the iPhone
as the survey instrument, both the financial costs and the human and capital resources necessary
were reduced substantially, certainly to the level of mailed questionnaires and interviews, if not
much lower. This allows such inquiries to be made alongside the “snapshot style” questions
more commonly found in ESM research.
RO:2 speaks specifically to the current study’s examination of VO-PEB, in particular
individuals’ value orientations, subjective well-being, and their propensities to engage in proenvironmental behavior. RO:3 speaks to Brown and Kasser’s (2005) proposition for additional

)&!!

!

research that accounts for recall error and biases and socially desirable responding (SDR). In
addition the current study’s construction as a consumer behavior study avoids inadvertently
activating the participants’ values, a phenomena that Kasser (2011) argues results in an
automatic shift in their attitudes and behavior.

3.1.3. Values, SWB, and Pro-Environmental Behavior in Previous Studies
Schwartz’s work (1992, 1994, 2006) identifies 10 types of basic priorities people have
across the world, two of which Schwartz identifies as self-enhancement values, i.e. power and
achievement, and two types of which Schwartz identifies as self-transcendent values, i.e.
benevolence and universalism (Kasser, 2010). Kasser and his colleagues (e.g. Grouzet et al.,
2005; Kasser & Ryan, 1996; Ryan et al., 1999; Schmuck et al., 2000) identify three types of
goals they call extrinsic or materialistic, which align with Schwartz’s (1994) self-enhancement
values: financial success, image and status. They identify three types of goals they call intrinsic,
which align with Schwartz’s (ibid) self-transcendent values: self-acceptance, affiliation, and
1

community feeling.

Past research shows that self-enhancing and extrinsic values have been associated with
attitudes and values that are inconsistent with environmental stewardship (Saunders & Munro,
2000; Good, 2007; Schwartz, 1992, 2006; Schultz et al., 2005) and behavior that is ecologically
detrimental (Richins & Dawson, 1992; Brown & Kasser, 2005: Gatersleben et al., 2008; Kasser,
2005, in press). On the other hand, self-transcendent and intrinsic values and goals have been
associated with attitudes and values that are consistent with environmental stewardship (Schultz
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
1

For the remainder of this paper, “self-enhancing, extrinsic and
materialistic” will be used interchangeably, as will the terms “selftranscendent and intrinsic.”
)'!!

!

et al., 2005; Gatersleben et al., 2008; Kasser, 2005) and behavior that is ecologically sensitive or
pro-environmental (Sheldon & McGregor, 2000; Brown & Kasser, 2005; Kasser, in press). In
addition, Ahuvia and Wong (2002), Banerjee and Dittmar (2008) and Kasser et al. (1995) argue
that the two primary causes of high materialistic or extrinsic values are social modeling, i.e.
being exposed to people or messages suggesting that “money, power, possessions, achievement,
image and status are important aims to strive for in life” (Kasser, 2010: 92), and insecurity, i.e.
economic hardship, poor interpersonal relationships, hunger, personal self doubt, or social
exclusion (ibid: 92). Thus, individuals exposed to a high degree of materialistic social modeling
or insecurity are more likely to hold materialistic or extrinsic values than are those who are
exposed to a lesser degree of the same. And as such, those individuals will exhibit little proenvironmental behavior.
Regarding subjective well-being, research has shown (Kasser & Ryan, 1993, 1996;
Sheldon & Kasser, 1995; Sheldon et al., 2004) that intrinsically oriented individuals report
greater well being than those who focus on extrinsic values. And Brown and Kasser (2005)
report positive correlations between subjective well-being and pro-environmental behavior. The
authors ultimately conclude that “happy people live in more ecologically responsible ways
because such individuals hold intrinsically oriented values” (ibid: 360).
All of this research suggests one fairly straightforward supposition: individuals who hold
intrinsically oriented values lead happier lives than and perform more pro-environmental
behavior than those individuals who hold extrinsically oriented values.

)(!!

!

3.1.4. Values, SWB and Pro-Environmental Behavior in the Current Study
In examining the supposition above, the current study asked each participant questions
about their values, subjective well-being, and pro-environmental behavior. In order to examine
both their values and their SWB, questions were asked about their level of satisfaction, both for
the day and in general, and about the specific contributions to that level of satisfaction. For
example, participants were asked about the effect that their personal relationships, their personal
finances, or their health had on their level of satisfaction. Asking each individual about these
effects was intended to examine whether or not each effect or value contributed to their overall
satisfaction. If, as Rokeach (1973) suggests, values are the psychological representations of what
people believe is important in life, then those individuals who report no effect for a particular
question can be thought not to hold or be pursuant of that particular value. As opposed to asking
an individual “how much he or she values a particular [goal or value],” asking “what effect does
[this value] have on your level of satisfaction” releases the individual from having to self-identify
his or her motivations and instead should provide responses that more closely represent his or her
real attachment to values. More will be said about what effects align with what values (i.e. selftranscendent or self-enhancement) below.
Each individual was also asked two hypothetical questions regarding what type of vehicle
they would like to own and what city they would like to live in were cost not a concern. These
two questions were aimed at ascertaining both the individuals’ value orientations and level of
environmental concern.
Regarding pro-environmental behavior, the current study focused on the individuals’ diet,
transportation and housing choices, as these are typically identified as the most ecologically
consequential human activities (Brown & Leon, 1999). Brown and Kasser (2005: 358) identify

))!!

!

that “meat eaters who drive many miles per week and live in large houses have larger ecological
footprints than do vegetarians who use public transportation and live in small homes.” The study
also focused on respondents’ recycling behavior, as many previous studies have used recycling
as their primary pro-environmental variable (e.g. Guagnano et al., 1995; Hopper, 1991; De
Young, 1988-89; Corral-Verdugo, 1997). The study also inquired about the participants’
technology choices, e.g. the age and type of their computers, televisions, gaming systems,
automobiles, etc.

These questions were intended to examine the frequency with which

participants exchanged and upgraded technology.
The study also measured how often the individuals dined outside their homes and
identified the foods and beverages each individual kept in her refrigerator (through
photographs)—photographs were also used to identify participants’ recycling behavior. The
study also asked about the individuals’ transportation choices on multiple days throughout the
week. These self-reports and photographs were thought to be more accurate than would be those
reports gained through reflective single-administration surveys that suffer from recall error and
biases, such as telescoping, part-whole effects, and satisficing. Only the current approach, i.e.
that of the Experience Sampling Method and other methods similar, allows for this type of
longitudinal behavioral monitoring.
Certainly monitoring individuals’ behavior over the period of one week leaves the
research open to extrapolation errors, i.e. the faulty assumption that this single week of behavior
is representative of the individual’s typical activities. However only through additional research
or replicating longitudinal studies can those assumptions be verified or rejected.

)*!!

!

3.1.5. Science, Expertise and Values in Pro-Environmental Knowledge
The current study asked the participants to watch three short videos (See Appendix A for
the video’s hyperlinks) and respond to three questions immediately afterward. Each video used a
unique approach to inform its viewers about renewable energy, sustainability and climate
change. The first, a National Geographic video starring Natalie Portman and Chloe Sevigny,
argues that compact fluorescent light bulbs can help reverse the effects of global warming. It
was theorized by the researchers to advance a self-transcendent value orientation. The second
video, a Plastics Industry Trade Association video, starring Andrew Winston, the founder of
Winston Eco-Strategies, argues for the durability of sustainability as a theme in both the
consumption and production of industrial products. He does so from a self-enhancement value
orientation. The final video, from 1958, starring Dr. Frank Baxter, examines global warming
from neither a self-transcendent or self-enhancement perspective and instead uses fear, or
insecurity, a proposed cause of self-enhancement values (Kasser, 2010), as its primary
motivation.
Upon conclusion of each video, the participants were asked to measure the video’s
scientific content and expertise and were then asked about the accuracy of statements made
during the video. The intent of these questions was to determine the effectiveness of each
message and to examine the participants’ reactions to each, i.e. what level of trust the narrator
was able to achieve. It was theorized that those individuals who showed self-transcendent value
orientations may react differently than those with self-enhancement value orientations. More
will be said about this below.

)+!!

!

3.1.6. Hypotheses
This study is primarily the application of a method and instrument to VO-PEB, which
argues that individuals who hold intrinsically oriented (or self-transcendent) values lead happier
lives than and perform more pro-environmental behavior than those individuals who hold
extrinsically oriented (or self-enhancement) values (Brown & Kasser, 2005). The method and
instrument were discussed principally in Chapter Two. VO-PEB will be examined below as this
thesis’s hypotheses.
The first hypothesis is informed by Brown and Kasser’s (2005) argument that happier
people, i.e. people with high subjective well-being (SWB) perform more pro-environmental
behaviors than those with low SWB. This can be tested in the current study by comparing the
participants’ SWB measures and their pro-environmental behaviors (or lack of).

Null

Hypothesis 1 (Ho1) then states that there is no association between individuals’ reported SWB
measures and their pro-environmental behaviors. Whereas alternative Hypothesis 1 (H1) argues
that there is a positive relationship between participants’ self-reported SWB and participants’
pro-environmental behaviors.
The second hypothesis is informed by the VO-PEB, which states that individuals who
hold intrinsically oriented values perform more pro-environmental behaviors than those who
hold extrinsically oriented values.

Null Hypothesis 2 (Ho2) thus states that there is no

association between self-transcendent or materialistic value orientations and pro-environmental
behaviors. Whereas alternative Hypothesis 2 (H2) argues that individuals who demonstrate selftranscendent value orientations will engage in more pro-environmental behaviors than those
individuals who demonstrate materialistic value orientations.

),!!

!

This can be tested by examining the participants’ answers to “contributions to
2

satisfaction” questions and then comparing those answers to the individuals’ pro-environmental
behaviors. One of the contributions of this study specifically is that both negative and positive
effects on an individual’s level of satisfaction are considered telling of his or her value
orientation. For example, if an individual reports that her financial situation has caused a great
positive effect to her level of satisfaction, that contribution is interpreted as a sign of an extrinsic
value orientation. An individual who recently lost her job and reports her financial situation, or
lack of income, as a great negative effect to her level of satisfaction is still considered to be
evincing an extrinsic value orientation. Only the lack of an effect constitutes the lack of that
particular value orientation. For example, an individual who reports her personal finances
having little to no effect on her level of satisfaction is considered to be void of that particular
value orientation. Thus Hypothesis 2 suggests that individuals who consistently report great
positive or negative effects to their level of satisfaction from materialistic or extrinsic valueoriented phenomena will exhibit less pro-environmental behavior than those who report great
positive or negative effects to their level of satisfaction from intrinsic value-oriented phenomena.
The third hypothesis (and its corollary) examines whether individuals who hold
intrinsically oriented values are happier than those individuals who hold extrinsically oriented
values. Null Hypothesis 3 (Ho3) states that there is no association between participants’ value
orientations and their self-reported SWB. Whereas, the alternative Hypothesis 3 (H3) states that
a positive relationship exists between participants’ self-transcendent value orientations and self-

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
%!See

Appendix A for a list of the “contributions to satisfaction”
questions.!
*-!!

!

reported SWB, and a negative relationship (H3A) between participants’ materialistic value
orientations and their self-reported SWB.
The four and fifth hypotheses are informed by Kasser’s (2010, 2011) work on the
effectiveness of pro-environmental information campaigns. Kasser argues that campaigns that
focus on power and achievement as values (i.e. success, image and status) fail to evince
sustained pro-environmental behavioral change in individuals.

Campaigns that focus on

universalism and benevolence however are comparatively successful.

The current study

presented the participants with three pro-environmental knowledge videos, one of which focused
on self-transcendent values, and two which focused on either self-enhancement values or the
cause of self-enhancement, i.e. fear and insecurity.
Null Hypothesis 4 (Ho4) argues that there is no association between individuals’
attribution of expertise and the video’s value orientation, while Null Hypothesis 5 (Ho5) argues
that there is no association between individuals’ value orientations and their attribution of
expertise to the knowledge campaign. Alternative Hypothesis 4 (H4) states that individuals will
attribute greater expertise to a pro-environmental knowledge campaign that relies on selftranscendent value orientations than a campaign that relies on self-enhancement values.
Alternative Hypothesis 5 (H5) states that individuals who demonstrate self-transcendent value
orientations will attribute greater expertise to a pro-environmental knowledge campaign that
relies on self-transcendent value orientations.

3.1.7. Rating-Scale
For most of this study and many previous studies (e.g., Bailey, 1994; Ferrer-i-Carbonell
& Gowdy, 2007), seven-point Likert scales were used. These scales seem to be best in terms of

*$!!

!

reliability, percentage of undecided respondents, and respondents’ ability to discriminate
between the scale values (Schwarz et al., 1991). Only the endpoints of the Likert scale were
labeled with words, e.g. very satisfied or very unsatisfied, extremely happy or completely
miserable, for the satisfaction and contribution to satisfaction questions. It has been suggested
that reliability and validity can be improved if all points on the scale are labeled with words
(Krosnick & Berent, 1993; Peters & McCormick, 1966). Respondents also report being more
satisfied if all points are labeled (Dickinson & Zellinger, 1980)—the video-response questions in
this study labeled each scale point. Yet this level of detail and accuracy was simply not
appropriate in the context of MMS messages sent via broadband phones. It would also be
difficult if not superfluous to label the numbers on some of the rating-scales, in particular those
that dealt with a single continuum, e.g. “1: not satisfied at all to 7: completely satisfied.”
Nevertheless, there has been recent discussion concerning the selection of the numbers
themselves, suggesting that the numbers used, e.g. 1 to 7 as opposed to -3 to 3, can produce
shifts in responses (Schwarz et al., 1991). Krosnick (1999) suggests that the numbers should
reinforce the meanings of the words, and Ajzen (1991) blames uni-polar, as opposed to bi-polar,
scaling of belief items for the low correlations found in studies of Fishbein and Ajzen’s (1975)
model of expectancy value. Regardless of these findings, uni-polar scales were used in the
current study to prevent, if not at least encourage participants away from anchoring their
responses at zero.
This study relied primarily on single-response measures.

The reliability of these

measures varies considerably, “depending on the particular judgment required of the subject”
(Fishbein & Ajzen, 1975: 108). Because of this unreliability, variables measuring a construct are
often computed as an average of responses to several related items shown to be highly correlated

*%!!

!

(Hektner et al., 2007). In ESM however, repeated measurement often takes the place of multiple
items. Still, it is important to note that during ESM periods, respondents will often tend to find a
stable anchor around which they vary their responses. This leads to a decrease in variability, or a
decrease in the standard deviations of their responses (ibid). As stated above, to discourage
anchoring at a value of zero, uni-polar scales were used.

3.1.8. Analytical Methods
The five hypotheses listed above are concerned with the presupposed associations
between value orientations, SWB and PEB, and thus this study relied primarily on chi-square
tests. Chi-square tests are principally used to simultaneously classify data with respect to two or
more traits that are observed for each sample element (Bhattacharyya & Johnson, 1977). For
example, students may be classified, and were for this study, by their self-reported SWB and as
having one of two value orientations. It is possible using contingency tables to then infer
whether or not particular characteristics appear to manifest independently or whether certain
levels of one characteristic tend to be associated or contingent with some levels of another (ibid).
To clarify, chi-square tests are able to demonstrate that high levels of SWB are independent of,
or are associated more readily with one of two, value orientations.
For this study, individuals were classified as either having a self-transcendent value
orientation or a materialistic value orientation—more will be said about this and the following
classifications in Sections 3.3.3.1 through 3.3.3.5. Individuals were also classified as either
engaging or not engaging in particular pro-environmental behaviors. Finally their self-reported
subjective well-being, responses which ranged from a minimum of 1 to a maximum of 7, were

*&!!

!

added, averaged and recoded onto a scale of 0 to 1. Using chi-square tests the study was able to
determine if associations between these variables as measured exist.
Once the sample data were collected, the researchers coded it using Microsoft Excel.
More will be said about the coding of specific datasets in the results sections below. Once the
data were coded, IBM SPSS Statistics Version 19 was used to analyze those results. Using
SPSS’s crosstabs function and chi-square test statistic function, each of the classifications
introduced above were analyzed and associations or the lack of associations between them were
examined. The study relied upon SPSS’s ability to separate variables to the extent allowed for
by the number of different results attained. To clarify, individuals’ SWB was represented by a
code; that code could have been .075, .350, .465, or any number between zero and 1, to the
thousandth decimal place. Instead of breaking down these SWB codes into quartiles, as is
sometimes done, SWB was separated into categories that equaled the number of different
responses received. This process allowed for much greater detail and should produce more
accurate chi-square statistics.
Because chi-square tests only test for independence, Pearson correlation coefficients were
also created for some of the classification variables—these coefficients are depicted in the data
tables in the results sections below. This was done to determine whether there existed a positive
or negative correlation between certain variables, for example between responses to certain
contribution to satisfaction questions and SWB. SPSS was used to analyze and calculate these
statistics as well.
All of the data analyzed are represented in data tables in the following sections, and in the
appendices. Both the chi-square critical value (to a significance of 95%) and the chi-square
statistic attained are presented, alongside the Pearson correlation coefficient and that

*'!!

!

coefficient’s significance. The number of observations (N) and degrees of freedom are also
presented.

3.2. Sample Construction and Signal-Response Rate
External validity is a concern with any analytical method; however, there are specific
concerns regarding ESM. Because ESM is typically targeted at a specific population or specific
event, generalizing such a study’s results to a larger population can be difficult (Hektner et al.,
2007). ESM also typically requires more resources per participant than most methods, so sample
populations are often small in comparison—Zuzanek (1999) argues that attempting to change
from this manner of purposive sampling to a more representative sampling technique in ESM
will be a significant challenge.
This study was initially intended to sample a number of specific populations so as to
retain some level of external validity and to compare and contrast the results between samples.
Twelve iPhones were purchased by the researchers, prepared and made ready to be distributed to
three distinct populations. These populations were designated as Michigan State University
students, staff members, and members of the faculty. It was hypothesized that the primary
distinction amongst the three would be their level of income and their respective level of jobsecurity—an economic constraint that is said to often affects individuals’ subjective well-being
and one’s willingness to behave pro-environmentally. Security concerns are also considered
causes of self-enhancement value orientations (Kasser, 2010).

In addition, distributing the

devices to the members of each population would avoid the self-selection bias inherent in
sampling only those individuals who already owned iPhones. Due to time constraints, Internal
Review Board (IRB) restrictions and concerns regarding the retrieval of distributed phones

*(!!

!

however, the issuing of phones was substituted for sampling only those individuals who already
owned the device. However, the number of individuals in each population who owned iPhones
was too small to be viable, and so only university students were sampled.
As a result it could be argued that this study fails to test the iPhone as a generalizable or
universal ESM instrument because the study’s participants were already familiar with the
instrument and its functions. However, this argument fails to consider that most individuals’
phones are broadband-capable and capable of interacting with the current study in its original
form. As a result, while it may be true that the iPhone itself was not tested as a capable
instrument for non-iPhone owners, broadband-capable mobile phones were tested.
Before presenting the response rate for this study it is important to note the distinction
between response rates for single-administration surveys (SAS) and ESM surveys. Whereas
SAS typically have only one response rate, ESM trifurcates that rate. SAS response rates refer to
the proportion of recruits for the study sample who actually agree to participate. ESM, because
of its complexity, identifies this rate as the volunteer rate. The signal-response rate then
describes the proportion of signals for which responses are completed, and the attrition rate
describes the proportion of participants who do not complete subsequent waves of data
collection. Attrition rates only apply to those studies that are longitudinal, i.e. contain more than
one data collection period per population. One of the reasons that this study used iPhones as
opposed to paper-and-pencil or a computerized entry method was that it was hypothesized that
signal-response rates would be greater in a study that relied on the former than those studies that
relied on the latter. This was found to be the case as will be reported below; however, because
each individual was required to own this specific technology volunteer rates were extremely low.

*)!!

!

These low volunteer rates remain a sticking point. Due to IRB regulations, the language
used in any document must match the language in all documents. Because this project used a
novel method (ESM) and instrument (iPhones), the review process was more comprehensive and
extended than was initially proposed or predicted, and thus the evolution of documents, method
and instrument was stifled. This produced both positive and negative results. Positively, the
implications of this study are specific to Apple iPhone users exclusively. Negatively, this study
was unable to sample students who had phones comparable in power and function to iPhones,
e.g. Google Android and Blackberry mobile-web devices, as discussed above. While there may
exist anecdotal evidence that iPhone users are unique from others, little empirical evidence exists
delineating what exactly those differences are. This study did not attempt to determine these
differences. The homogeneity of the sample however may be useful in other respects, which will
be discussed below. It also remains to be seen whether the homogeneity of this sample provides
advantages in quality or power as compared to a higher volunteer rate and a larger sample.
Additional studies comparing the results of this study with those that use multiple instruments
are necessary.
To construct this study’s sample, 7862 students were initially contacted by email with
instructions that researchers at MSU were looking for individuals to participate in an iPhone
Study. The participants’ primary tasks were described within the email as answering questions
sent to them as text messages, taking pictures with their iPhone of their environment, of what
they purchased at the store, or of events and activities in which they had participated or were
participating. They were also told that they would be sent links to short film-clips that they
would be asked to view and comment upon. They were instructed that approximately five
questions per day would be sent to them no earlier than 8am and no later than 9pm for a period

**!!

!

of seven days. Filling out a short paper survey at the conclusion of the survey was also required.
They were asked to respond to the email if they were interested in learning more about
participating in the study.
103 students replied to the initial email, and 11 students were recruited in person. These
114 students were then sent via email a letter of consent, which contained more detailed
instructions, the specific dates and times that their examination period would begin and end, and
notice that each would receive $20 as an incentive for participating. They were instructed that
their receiving the incentive was not dependent upon their answering every question. Of the 114
students, n=71 (62.3%) students agreed to participate.

Hektner et al. (2007) identify that

volunteer rates for ESM studies vary greatly, as low as 12 percent for unskilled, blue-collar
workers to 91 percent for fifth and eighth graders (Csikszentmihalyi & Larson, 1987). This
study’s volunteer rate was 0.89 percent.

This was expected based on the technological

requirements of the study, and would likely have been much higher if the researchers were able
to recruit all students with mobile-broadband camera-equipped phones. Regardless of the low
volunteer rate, samples of this size are not uncommon for ESM studies (e.g. Feldman Barrett,
2004) and studies with as few as 5 or 10 participants can still provide rich data that can be used
reliably in statistical analyses (Hektner et al., 2007: 34).
Each student received 37, 39 or 41 questions each by text messaging, depending on
which examination period he or she participated, spread across seven days (mean = 5.57
questions/day). In total 2777 questions were texted to the participants with a signal-response rate
of 98.13 percent. Only 52 questions were not answered, and of these 44.2 percent (n=23) were
video-responses and 17.3 percent (n=9) were photograph-responses, those questions that required
the most effort of the participants.

*+!!

!

This signal-response rate of 98.13 percent is much higher than those rates reported from
previous ESM studies, providing immediate support for the researchers’ early conjecture
regarding the benefits of mobile-phone based ESM studies. Typical signal-response rates range
from 70 to 80 percent (e.g. Csikszentmihalyi & Larson, 1987; Hormuth, 1986; Larson et al.,
2002; Zuzanek, 1999; Mannell et al., 2005) and rates amongst university students can range from
50 to 75 percent depending on both hardware and whether or not follow-up signals were used
(Zuzanek, 1999; Feldman Barrett, 2004). Follow-up signals were used in this study.
None of the studies mentioned used mobile phones as either signaling devices or as
reporting devices.

In addition, this study was the only one to ask respondents to take

photographs or to watch and react to videos. While the latter is not likely to be incorporated
quickly into more typical ESM studies, the former could be a valuable addition, helping
researchers to gain insight into the ecological context of participants’ responses or even their
psychological states, e.g. through providing photographs which represent one’s mood. While the
photograph-response questions had the lowest signal-response rate (95.77%), it was still a rate
much higher than those seen in previous studies. See Table 1 below for all sample and response
rates.

*,!!

!

Table 1. Study Sample and Response Rates
Initial Contacts
% Possible Volunteers
By Email
7862 100%
103
Face-to-Face Contacts
11 100%
11
Questions Asked
Total
2777
Video-Response
639
Photograph-Response
213
Text-Response
1925

%

Volunteer Rate %

1.31%

70

0.89%

100%

1

9.09%

Questions Answered

Signal-Response Rate %

2725

98.13%

616

96.40%

204

95.77%

1905

98.96%

3.3. Results and Discussion
3.3.1. The Apple iPhone and ESM
This study was a joint application of ESM and broadband mobile-phones, iPhones
specifically, to Schwartz (1994) & Kasser’s (2010) value orientation theory of pro-environmental
behavior. That method and instrument’s successful application suggest broadband mobile-phones
are not only capable but preferable to PDAs, written research journals and beeper technology in
ESM studies. The extremely high signal-response rate (98.13 percent) achieved in this study
suggests that the burden of such studies is reduced significantly with the incorporation of mobilephone technology. Such a high signal-response rate alone would be significant; however, when
combined with the incorporation of multimedia and direct behavioral monitoring, as well as the
considerable reduction in financial and human resources necessary to enact such research, this
study suggests that mobile-phone technology-driven ESM studies are an extremely attractive

+-!!

!

alternative to single-administration surveys, focus groups, interviews, and many lab-setting
experimental designs.
There are thus in reality two implications of this study’s methodological and instrumental
application.

The first concerns ESM and the second concerns ESM’s setting in socio-

environmental research as a whole. The first implication is that mobile-phones make ESM
easier, cheaper and possibly more accurate. Replacing PDAs, beepers and research journals with
mobile-phones likely reduces the participants’ burden, and as seen in this study certainly reduces
the researchers’ burdens. And such reductions should allow for greater flexibility, larger samplesizes and more accurate data, as respondents no longer need carry additional materials during the
research period. They need only carry the one device from which they are rarely free. Yet what
respondents and their researchers can accomplish with that device is much greater in comparison
to what those previous instruments afforded. The broadband mobile-phone’s audio and video
capabilities, as well as photographic ability, increase the type of questions and interactions
available to the researcher. Furthermore, instead of relying on self-reported attitudes, emotions,
times, locations, and behaviors, respondents can now provide researchers with real, objective
evidence of the same (or what may prove not to be the same!).
Another result of the current study’s implementation was the significantly reduced
financial and human resources necessary to field it. Previous ESM efforts required not only
considerable human resources both to develop and analyze the study, but often times the
preferred instruments must be purchased or constructed.

This study not only used the

respondents’ own technology and thus reduced technology costs—or at least shifted those costs
to the respondents unwittingly, but was also free from having to train or provide extensive
instruction to the respondents regarding that technology.

+$!!

Minimal physical contact was

!

necessary, which both reduces the burden for the respondent and reduces the human resources
and preparation necessary for the research team.
Other studies (e.g. Hicks et al., 2010; Killingsworth & Gilbert, 2010; Uy et al., 2010)
have advised developing web-based applications to assist in the design, deployment and analysis
stages of mobile-phone-driven ESM research. This thesis and research suggests that such efforts
may backfire. Not only can the up-front costs, both financial and human, be significant—though
certainly not in every case, but the cognitive requirement of the respondents may also be
increased, requiring greater physical or cognitive engagement with the device. Such an increase
would erode one of the significant advantages afforded by mobile-phone instrumentation, i.e. the
reduced burden such technology affords the respondent.

3.3.2. The Apple iPhone, ESM and Socio-Environmental Research
The first proposition of this study is that mobile-phones, specifically Apple iPhones, and
SMS messaging—as opposed to web-applications—streamline and expand ESM research,
allowing for more and more accurate deployments of the method within socio-psychological
research.

The second supposition is that ESM and mobile-phone technology need not be

confined to socio-psychological research and instead are a cost-effective, high-quality and highaccuracy alternative to many of the methods currently used in socio-environmental research.
This supposition is supported by the current study’s implementation, yet not necessarily
supported by the current study’s data analysis and hypothesis-testing—more will be said about
this in following sections.
ESM specialists like Joel Hektner, Jennifer Schmidt and Mihalyi Csikszentmihalyi
(Hektner et al., 2007: 12) argue that ESM is best at examining, for example, the psychology of

+%!!

!

adolescence, family dynamics, the experience of psychopathology, the experience of media,
gender differences, and the “optimal experience of flow.” Yet this study suggests that ESM need
not be confined to the field of psychology. While socio-environmental researchers are now and
will continue to be engaged in studies of pro-environmental behavior, investigating the
determinants of that behavior will increasingly transcend any single discipline.

Pro-

environmental behavioral research requires understanding both the internal and external
coordinates of experiences—as ESM aids in accomplishing—as well as the internal and external
constraints affecting individuals’ pro-environmental attitudes and behaviors—something ESM
has not yet been applied to in force.
This latter form of investigation has long been accomplished with much cheaper and
more global or generalized examinations of individuals’ resources, constraints and surroundings,
and for good reason. Using ESM to investigate what type of vehicle a person owns would not be
warranted previously due to the high costs of implementing such a study. And certainly to some,
ESM will never be, and should never be, the preferred method to acquire such data—this study
used a paper survey to gain such information. However, ESM does provide more accurate data
about what type of vehicle a person drives, or drives in (as opposed to which vehicle he simply
owns).
For example an individual may own a certain vehicle, but may often car-pool or catch a
ride with a friend in the summer. He may drive his vehicle more in the winter because of the
snow. His car may be temperamental and he must often drive a rental car. Certainly, singleadministration surveys, focus groups and interviews are all capable of investigating such
behavior to a similar extent; however, each is limited by recall bias and error, and the specific
effort of the individual to generalize his or her own behavior. Furthermore, these efforts lack

+&!!

!

ecological validity, or how the specific setting of our experiences, behaviors and attitudes affect
both our conception and reporting of such phenomena. ESM allows for such investigation.
However, until mobile-phone technology could be incorporated, as is presented in this thesis,
such investigations were deemed too costly. The current study suggests that such limitations
may no longer exist.

3.3.3. Hypotheses
Schwartz (1994) and Kasser’s (2010) value orientation theory of pro-environmental
behavior has been investigated for over two decades, often through single-administration
surveys. Yet Kasser and Brown (2005: 364) argue for future research that, “requests reports over
a limited time frame…to help circumvent biases due to retrospective reporting” and moves
beyond self-reports (e.g. uses direct observation). Sections 3.3.3.1 through 3.3.3.5 report such
results from the current study. Table 2 below also presents a summary of the data collected—
certain aspects of Table 2 will not be addressed or explained until later sections.

+'!!

!

Table 2. Summary of Sample Characteristics

Variable
Individuals
Males
Females
ST Individuals
SE Individuals
Unclassified Individuals
ST Males
SE Males
ST Females
SE Females

N
40
31
27
37
7
14
20
13
17

Variable
N
Subjective Well Being (0-1)
TOT SWB
71
GEN SWB
71
SNAP SWB
71
Pro-Environmental Behavior
REC
(0-1)
57
TRANS (0-1)
71
PEB
(0-2)
71

TOT
SWB

SWB
ST DEV

0.785
0.723
0.767
0.744
0.792
0.797
0.770
0.736
0.714
MEAN

0.117
0.094
0.109
0.12
0.07
0.111
0.136
0.098
0.095

PEB
1
1.05
1.14
0.96
0.86
1.167
0.947
1.115
0.971

PEB
ST DEV
0.6
0.71
0.66
0.66
0.48
0.617
0.621
0.74
0.717

ST DEV

0.893
0.770
0.749

0.111
0.141
0.137

0.456
0.641
1.007

0.503
0.398
0.657

3.3.3.1. Hypothesis 1: SWB & Pro-Environmental Behavior
H1 stated that a positive relationship exists between participants’ self-reported SWB and
participants’ pro-environmental behavior. In order to test this hypothesis the participants were
asked two types of questions regarding their subjective well-being. The first type of question
3

asked them to report their general level of satisfaction. This question was asked twice, once on
the first day of the research period and once on the last day. The other SWB questions asked
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
&!All

questions are listed in Appendix A and asked for Likert-scale
responses on a scale of 1 to 7.!
+(!!

!

respondents about their levels of satisfaction on particular days—these were randomly asked
throughout the week and at different times of day. The participants’ responses to these questions
were then combined to form three variables: total SWB (TOTAL SWB), general SWB (GEN
SWB), and snapshot SWB (SNAP SWB). TOTAL SWB was a weighted average of all level of
satisfaction questions; GEN SWB was a weighted average of only the general level of
satisfaction questions (asked at the beginning and end of each research period), and SNAP SWB
was a weighted average of only the random daily level of satisfaction questions. GEN SWB
represents the level of subjective well-being that would most likely be reported during a single
administration survey (SAS). TOT SWB then can be considered a combination of the SAS and
the ESM self-reports.
A respondent’s pro-environmental behavior (PEB) was measured using two methods.
The first asked the respondent to report on certain behaviors that he or she had performed during
the day. These ranged from manner of primary transport (i.e. walking, the bus, automobile,
biking) (asked twice) to the length of time the individual spent online, talking on the phone or
watching television, or the number of meals the individual ate at a restaurant or at home. The
second method of inquiry required the individual to photograph his or her behavior. The
respondent was asked to photograph her recycling and the inside of her refrigerator, as well as
the type of television set and video-game system she owned. At no point was a participant asked
if or why he did or did not engage in a particular behavior.
To examine the relationship between participants’ SWB and PEB and to determine if that
relationship was positive, i.e. high SWB inferred greater performance of PEB, Chi-Square tests
were performed on each of the SWB variables and on the participants’ recycling behavior
(REC=1 if participant recycles; 0 if not) and transportation behavior (TRANS=1 if participant

+)!!

!

walks, bikes or uses mass-transit; 0 if participant drives automobile), as well as a weighted
average of the two (PEB=2 if participant recycles and does not drive; 1 if participant recycles
and drives or does not recycle and walks, bikes or uses mass-transit; 0 if participant does
neither).

Other behaviors either did not demonstrate sufficient variance or demonstrated

variance that was dependent upon the specific day, date or time the signal was received, e.g. time
spent online was much higher on weekdays then on weekend days, across all examination
periods. As a result these behaviors were left out of the analysis. The results of analysis are
shown below in Table 3.
Table 3. SWB and Pro-Environmental Behavior Results

Variable
1
TOTAL
SWB
TOTAL
SWB
TOTAL
SWB
GEN
SWB
SNAP
SWB

Variable
2

df

N

Pearson
2
!
Critical
Value

REC

23

56

35.172

23.166

0.451

-0.119

0.384

TRANS

46

70

62.830

35.800

0.861

0.077

0.526

PEB

92

70

115.390

81.331

0.779

-0.048

0.694

PEB

36

70

50.998

47.414

0.097

-0.072

0.551

PEB

60

70

79.082

49.136

0.841

0.002

0.986

Pearson
2
!
Statistic

Asymp.
Signif.

Pearson's
R
(Correl.)

Approx.
Signif.

The results in Table 3 show that there is no significantly positive (or negative)
relationship between the respondents’ self-reported SWB and their pro-environmental
behaviors recycling or mode of transportation, thus Null Hypothesis 1 (Ho1) cannot be
rejected. There are many possible reasons no association was found; however, the most
obvious is that extrapolating individuals’ general behavior from their daily behavior is

+*!!

!

problematic. Brown and Kasser (2005) used retrospective reporting of individuals’ PEBs to
investigate the relationship between PEB and SWB. Yet individuals reporting their PEBs
are subject to a number of errors and biases, most notably socially-desirable responding
(SDR) and recall error. Regardless of whether respondents are aware of the researchers’
intentions, it is possible that individuals over-report PEB, despite Milfont (2009) and Schahn
and Bohner’s (2002) assertions that the effects of SDR are insignificant in socioenvironmental research.
Regarding recall errors, this study’s investigation of PEBs took only a snapshot of the
respondents’ behavior. And thus it is difficult to generalize (even within the sample). Were the
signals sent on different days or the examination period to take place in a different month, it is
very possible that the participants’ PEB reports would be very different. Yet these contextual
variables, or external coordinates of experience (Hektner et al., 2007), are certainly important to
recognize and raise questions regarding individuals’ own reports of their generalized behavior.
The very fact that snapshots are so variable, or may be so variable, raises questions about the
veracity of individuals’ retrospective and generalized reports.

Do they report their PEB

performance on average or in average situations? The former is more accurate, though not
necessarily more interesting, than the latter, especially if the latter means only that they recycle if
there is a bin nearby or walk to work if it is sunny and 72 degrees. This average may rarely
occur, yet in remembering it may be this behavior or behavioral context that dominates. ESM
studies, and their use of randomized signaling, are particularly helpful in discerning these
inconsistencies.
These variables are subject to a great deal of underlying constraints and effects,
particularly in a sample of university sophomores, some living in dorms. For example, an

++!!

!

individual living in a one-room dorm without a car may rarely consider the impact of his PEB,
but due to resource constraints may appear to be engaging in PEB. Or an individual living at
home with his parents and commuting twenty miles to school each day may still very much wish
to be environmentally conscious, yet cannot behave accordingly. These concerns certainly need
to be considered and may imply only that a much weaker relationship than hypothesized exists
within this sample regarding SWB and PEB. Or, it may be that resources, situations and
constraints external to the individual dominate variables like SWB elsewhere as well, making
such correlations between SWB and PEB superfluous. More research is of course necessary.

3.3.3.2. Hypothesis 2: Values and Pro-Environmental Behavior
H2 stated that individuals who demonstrate intrinsic value orientations will engage in
more pro-environmental behaviors than those individuals who demonstrate extrinsic or selfenhancing value orientations. In order to examine this hypothesis the participants were asked
two types of questions; the first were in regard to their PEB and are described in Section 3.3.3.1.
The second type of question was about specific contributions to respondents’ overall level of
satisfaction (these questions are referred to as “contributions to satisfaction”). For example,
individuals were asked questions about their personal appearance, their feelings of personal
success, and how they believed others viewed them. Each question asked the respondents to
report the effect that each phenomena had on their level of satisfaction (for the specific day in
which the signal was received, not in general). The intent of these questions was to determine
the individuals’ value orientations as a function of these effects, or more importantly the strength
of these effects. As stated previously, Rokeach (1973) considers values the psychological
representations of what people believe is important in life, thus those individuals who report no

+,!!

!

effect for a particular question can be thought not to hold or be pursuant of that particular value.
As stated previously both positive and negative effects are considered representative of a valueladen position! Only the lack of an effect suggests the corresponding lack of its representative
value orientation.
As described in H2 above, Schwartz (1994) and Kasser (2010) argue that selftranscendent value orientations lead to pro-environmental attitudes and behavior, while selfenhancing value orientations lead to a lack of PEB and pro-environmental attitudes. The specific
goals, beliefs and values that make up these broader categories sometimes vary. However
Kasser argues that in general, image, status and financial success generally denote the selfenhancing values, while community feeling, self-acceptance and affiliation denote selftranscendent values. To capture self-enhancing (SE) value orientations then, this study asked
each respondent to rate the effect of his (1) personal appearance, (2) personal finances, (3)
feelings of success, (4) purchases and (5) how others viewed them on his overall level of
satisfaction for the day. To capture self-transcendent (ST) value orientations, this study asked
each respondent to report the effect of her (1) religious beliefs (or a religious event) (asked on
Sunday), (2) relationships, (3) local environmental quality, and (4) the day’s weather, on her
overall level of satisfaction for the day. Each participant was also asked about her (1) education
(or experience at school) and (2) health. These responses to health and education were not
thought to be ST or SE value-laden.
At the end of each examination period, each participant was then classified as a selftranscendent (ST) or self-enhancing (SE) individual based on his or her answers.

Those

individuals who consistently reported effects from SE-value satisfaction contributions were
designated SE individuals (1=SE), and those individuals who consistently reported effects from

,-!!

!

ST-value satisfaction contributions were designated ST individuals (0=ST). Two methods were
used however to determine which classification each individual received. The first counted all
effects within an ST or SE category; i.e. the answers 1, 2, 3, 5, 6, and 7 all contributed to the
classification. The second counted only strong effects; i.e. the answers 1, 2, 6, and 7 only
contributed to the classification. Responses of “4” were deemed no effect and were considered
to represent no value-attachment to the specific contribution question.

To examine the

relationships between value orientations and PEB, Chi-square tests were conducted on each of
the individual contributions and the PEBs, as well as on the SE/ST individuals and PEBs. Tests
included both variables representing all effects and those representing only strong effects. A
sample of the results is presented in Tables 4 and 5 below (the full table of results is in Appendix
B).
Table 4. Individual Value Orientations and Pro-Environmental Behavior Results

Variable 1

Variable
2

SE=1/ST=0 REC
SE=1/ST=0 TRANS
SE=1/ST=0 PEB

df N

Pearson
2
!
Critical
Value

Pearson
2
!
Statistic

Asymp.
Signif.

Pearson's
R
(Correl.)

Approx.
Signif.

1
2
4

3.841
5.991
9.488

0.184
4.237
1.973

0.668
0.120
0.741

0.060
-0.192
-0.064

0.675
0.128
0.621

51
64
63

,$!!

!

Table 5. Individual Contributions to Satisfaction and Pro-Environmental Behavior
Results
Pearson
2
Pearson
!
Pearson's
2
!
Variable
Critical
Asymp. R
Approx.
Variable 1 2
df N
Value
Statistic Signif.
(Correl.)
Signif.
Self-Enhancement Values (Extreme Effects Only)
Personal Appearance
REC
1
57 3.841
2.263
TRANS
2
71 5.991
1.115
PEB
4
70 9.488
4.145
Personal
Finances
REC
1
57 3.841
0.524
TRANS
2
71 5.991
0.344
PEB
4
70 9.488
4.966

0.132
0.573
0.387

-0.199
-0.113
-0.165

0.137
0.346
0.172

0.469
0.842
0.291

0.096
0.038
0.088

0.478
0.753
0.47

0.731
0.764
0.883

0.045
-0.067
-0.042

0.737
0.582
0.731

0.514
0.385
0.940

-0.087
-0.018
-0.085

0.522
0.882
0.488

0.562
6.079*
3.012

0.453
0.048
0.556

0.182
-0.454*
-0.213

0.485
0.034
0.353

0.040
2.364
1.803

0.841
0.307
0.772

0.036
-0.148
-0.037

0.847
0.361
0.823

0.032
6.053*
2.453

0.858
0.048
0.653

0.026
0.315*
0.170

0.862
0.014
0.195

Self-Transcendent Values (Extreme Effects Only)
Religious
Events
REC
1
57 3.841
0.118
TRANS
2
70 5.991
0.537
PEB
4
69 9.488
1.170
Relationships
REC
1
57 3.841
0.427
TRANS
2
70 5.991
1.908
PEB
4
69 9.488
0.786
Self-Enhancement Values (All Effects)
Purchases REC
1
17 3.841
TRANS 2
22 5.991
PEB
4
21 9.488
How
Others
Viewed
REC
1
31 3.841
TRANS
2
40 5.991
PEB
4
39 9.488
Value-free?
Education REC
TRANS
PEB

1
2
4

48
61
60

3.841
5.991
9.488

* denotes 95% Significance
** denotes 99% Significance

,%!!

!

As with Hypothesis 1, the data do not demonstrate that a statistically significant
relationship exists between the value orientations and the pro-environmental behaviors as
measured in the study. As a result, Ho2 cannot be rejected. Neither self-transcendent individuals
nor self-enhancing individuals engage in more or less PEB consistently. While transportation
decisions appear to correlate positively with self-transcendence (R= -0.192, "= .128), they do not
do so significantly.
Value orientations were tested both on an individual participant basis, i.e. by the
classification of respondents as self-enhancing and self-transcendent (SE=1 and ST=0) and on a
value basis by each individual variable representing one of the two value orientations. Table 5
shows a sample of this data. What is immediately apparent is the lack of significant associations
between the contribution to satisfaction responses and the PEBs measured in the study, with only
2

2

purchases (! = 6.079, df= 2; R= -.454) and education (! = 6.053, df=2; R= .315) as the lone
exceptions (both associating with individuals’ transportation choices). The possible reasons for
such a lack of correlation are numerous; however one in particular stands out.
The contribution to satisfaction questions may successfully identify individuals’ value
orientations only with repeated exposure. This study aimed at capturing momentary snapshots of
individuals’ daily life and subjective well-being. As a result, questions were asked attempting to
attain what specifically contributed to their overall level of satisfaction each day. Yet certainly
these contributions vary, as do most likely the strength of each contribution. Earlier an example
was given that argued that an individual who has just lost his job would regardless of this event
identify himself as self-transcendent or self-enhancing (i.e. the event would not mitigate or
accentuate personal finances as a major determinant of his SWB). Despite the results of this
study, this may still very much be true.

However a single response fails to capture the
,&!!

!

individual’s typical (or mean) response. That day may very well be a 1 or a 2; however, he may
very well answer 4 on most if not all other days.
Thus the very facet that makes ESM a powerful measure of behavior, values and
attitudes, i.e. its ability to capture instantaneous snapshots of each phenomena, also requires
those who use it as a method to be extremely patient. If single-administration surveys (SAS)
generalize too much, i.e. introduce errors by asking respondents to interpret broad swaths of their
behavior or beliefs, ESM may generalize too little, accidentally—or intentionally—measuring
behavioral or attitudinal oddities.

These oddities may also not necessarily be tied to the

individual, particularly in socio-environmental applications. For example, if the study takes
place in the winter, during periods of intense snow-fall, individuals across samples may be more
likely to answer in a particular direction because of this environmental reality. Again, this is not
a failing of ESM—it is valuable in providing a check on SAS research that may underreport such
anomalies and provide overly optimistic—or pessimistic—claims of PEB and pro-environmental
attitudes or values.
Before moving to Hypothesis 3 (and 3A), the relationships between the contributions,
education and purchases, deserve some attention. The data show that on average, individuals
who report a great effect from their education on their overall SWB make significantly less (R is
negative) pro-environmental transportation choices. And on average, individuals who report a
great effect from the purchases they make on their overall SWB make significantly more (R is
positive) pro-environmental transportation choices. The possible contextual or constraint-based
explanations for such antithetical results are too numerous to expound upon here. Yet what is
perhaps most interesting is that perhaps the context and constraints do not matter. Perhaps the
value-behavior relationships seen here dominate those external variables and the underlying

,'!!

!

values that guide each ‘contribution to satisfaction’ do predict pro-environmental and nonenvironmental behavioral decisions and attitudes. Obviously more research is necessary before
such claims can be made, yet it is perhaps only through ESM that such apparent anomalies can
be identified, and perhaps only through ESM can these relationships truly be understood.

3.3.3.3. Hypothesis 3: Values and SWB
H3 and H3A stated that a positive relationship exists between participants’ SWB and ST
value orientations, and a negative relationship exists between participants’ SWB and SE value
orientations.

To examine these relationships two types of questions were asked of the

participants, both of which have been described previously in Sections 3.3.3.1 (SWB) and 3.3.3.2
(Contributions to Satisfaction). As with Hypothesis 2, the relationships were examined both on
an individual participant basis by comparing SWB and the cumulated classification designators
ST and SE, and on a value basis by comparing SWB and the specific contribution to satisfaction
responses. A sample of the results is given below in Tables 6 and 7 (the full table of results is in
Appendix C).
Table 6. Individual Value Orientations and SWB Results

Variable 1
SE=1/ST=0

Variable
2
TOTAL
SWB
GEN
SWB
SNAP
SWB

Pearson
2
!
Critical
Value

Pearson
2
!
Asymp.
Statistic Signif.

Pearson's
R
(Correla.)

Approx.
Signif.

df

N

21

63 32.671

17.505

0.680

-0.113

0.377

9

64 16.919

6.429

0.696

-0.073

0.564

15

64 24.996

16.548

0.347

-0.101

0.426

,(!!

!

Table 7. Individual Contributions to Satisfactions and SWB Results

Variable 1

Variable
2

df N

Pearson
2
!
Critical
Value

Pearson
2
!
Statistic

Self-Enhancement Values (Extreme Effects Only)
Personal
AppearTOTAL
ance
SWB
23 70 35.172
37.568*
GEN
SWB
9
71 16.919
11.639
SNAP
SWB
15 71 24.996
20.399
Personal
Finances

TOTAL
SWB
GEN
SWB
SNAP
SWB

Weather

TOTAL
SWB
GEN
SWB
SNAP
SWB
TOTAL
SWB
GEN
SWB
SNAP
SWB

Pearson's
R
(Correla.)

0.028

0.370**

0.002

0.234

0.325**

0.006

0.157

0.288*

0.015

Approx.
Signif.

23

70

35.172

24.384

0.383

-0.224

0.063

9

71

16.919

8.180

0.516

-0.147

0.22

15

71

24.996

17.753

0.276

-0.160

0.183

0.612

0.157

0.197

0.368

0.029

0.812

0.778

0.212

0.078

Self-Transcendent Values (Extreme Effects Only)
Religious TOTAL
Events
SWB
23 69 35.172
20.487
GEN
SWB
9
70 16.919
9.789
SNAP
SWB
15 70 24.996
10.639
Relationships

Asymp.
Signif.

23

69

35.172

18.977

0.703

0.223

0.066

9

70

16.919

10.163

0.337

0.276*

0.021

15

70

24.996

11.577

0.711

0.125

0.302

23

67

35.172

22.383

0.497

0.238*

0.052

9

68

16.919

5.357

0.802

0.144

0.241

15

68

24.996

23.111

0.082

0.247*

0.042

* denotes 95% Significance
** denotes 99% Significance
,)!!

!

The data demonstrate that the cumulative effects of the ST and SE value orientations as
measured in this study are not predictive of SWB, i.e. there is no consistent or significant
positive or negative relationship between value orientations and SWB. Again, Ho3 cannot be
rejected. This lack of association is clarified by the value basis comparisons illustrated in Table
6.
There is a positively significant relationship between the effects of individuals’ personal
2

appearance on their SWB and their overall (TOTAL) SWB (! = 37.568, df= 23; R= .370). That
positive relationship is also consistent across their GEN SWB (R= .325) and SNAP SWB (R=
.288) (for clarification on these measures see Section 3.3.3.1). While such a relationship may
initially seem glaringly obvious, it is important to remember that the effect of one’s personal
appearance as measured is not only positive. Both negative and positive effects were measured
and contributed to the overall ‘personal appearance contribution to satisfaction.’ This data does
not support Kasser (2010, 2011) and Brown and Kasser’s (2005) supposition that self-enhancing
values such as image and status contribute negatively to SWB. Of course this dataset is based on
a small sample of iPhone owning students, who may collectively feel more positive about their
personal appearance than non-iPhone owning students.

And thus such data lacks the

generalizability necessary to refute such assertions. However, such results supported by and
instructed by Kasser’s call for research to move past self-reports and mitigate SDR and recall
error cannot and should not be ignored.
While the contribution to satisfaction category personal appearance does not support
Brown and Kasser’s supposition, the relationships and weather categories do provide initial
support (R= .276 and R= .238, respectively). These effects correlate significantly with GEN

,*!!

!

SWB for relationships and with TOTAL SWB and SNAP SWB for weather. While the results
for relationships may be robust, the results for weather may be ecologically invalid since the
examination periods all took place during the winter season. Replication of such results would
need to occur both within the same season and across all seasons to ensure such results hold.
Certainly the same limitations that were discussed in Section 3.3.3.2 regarding the
momentary, often non-replicated, and ungeneralizable nature of this study’s contribution to
satisfaction questions are in play here.

Yet as was mentioned previously, this does not

automatically invalidate the results but instead suggests that greater attention and more study,
particularly research that utilizes ESM, is necessary.

3.3.3.4. Hypothesis 4: Values and Pro-environmental Knowledge Campaigns
Hypothesis 4 stated that individuals will attribute greater expertise to a pro-environmental
knowledge campaign that relies on self-transcendent value orientations than a campaign that
relies on self-enhancement values. In order to examine this hypothesis each participant was first
asked a series of contribution to satisfaction questions (as described in Section 3.3.3.2.) so as to
be categorized as either a ST or SE individual. He or she was then shown three video-clips and
asked three questions regarding each clip at the video’s conclusion. All videos were sent to and
watched on the participants’ Apple iPhone.
The first video was designated as relying upon and advancing a self-transcendent value
orientation. The second video was designated as relying upon and advancing a self-enhancing
value orientation. The questions asked at the conclusion of each video asked the respondent to
gauge the video narrator’s expertise, the scientific content of each video, and the accuracy of the
narrator’s claims. Due to inconsistencies in the language of each narrator and the vagueness of

,+!!

!

certain pre-identified claims the accuracy questions were thrown out and not a part of this
study’s data analysis. The third video relied upon fear and insecurity, two causes of selfenhancing value orientation (Kasser, 2010), and were not a part of Hypothesis 4’s analysis (but
were analyzed for Hypothesis 5).
Paired Samples Tests were conducted on the data to determine which of the two videos
(STVID and SEVID) the participants attributed greater expertise and scientific content to as well
as to which video received a greater average of both (STCODE and SECODE). The results are
shown below in Tables 8 and 9.
Table 8. Environmental Knowledge Video Results I
Paired Samples Statistics
Entire Sample
Mean
Pair 1
Pair 2
Pair 3

ST Expertise
SE Expertise
ST Scientific
SE Scientific
STCODE
SECODE

N

Std. Deviation

Std. Error Mean

2.500
2.790
1.410
1.790
3.910
4.590

68
68
68
68
68
68

0.889
0.821
0.717
0.703
1.313
1.225

0.108
0.100
0.087
0.085
0.159
0.149

Paired Samples Correlations
N
Pair 1
Pair 2
Pair 3

ST & SE
Expertise
ST & SE
Scientific
ST & SE
CODE

Correlation

Sig.

68

0.245*

0.044

68

0.348**

0.004

68

0.376**

0.002

* denotes 95% Significance
** denotes 99% Significance

,,!!

!

Table 9. Environmental Knowledge Video Results II
Entire Sample

Paired Differences

Mean
ST & SE
Pair 1 Expertise
ST & SE
Pair 2 Scientific
ST & SE
Pair 3 CODE

Std.
Deviation

Std. Error
Mean

95% Confidence Interval of the
Difference
Sig. (2Lower Upper
tailed)

-0.294

1.052

0.128

-0.549

-0.04

0.024**

-0.382

0.811

0.098

-0.579

-0.186

0**

-0.676

1.419

0.172

-1.020

-0.333

0**

* denotes 95% Significance
** denotes 99% Significance
The results in Tables 8 and 9 show that the SE video and its narrator were thought to
show more expertise and greater scientific content than the ST video and its narrators. Certainly
there are underlying factors at play in each video with regards to the theme, content,
cinematography, setting, language used, and even the subject matter (i.e. the ST video was
advocating high-efficiency light-bulbs and the SE video was advocating sustainable consumer
products). Each video definitely has multiple influences on each respondent’s attitudes and
values. Thus it would certainly not be appropriate to extrapolate from these results that selftranscendent videos in general are thought to show less expertise or less scientific content than
self-enhancing videos. When one separates the sample by gender however, the results do
become more interesting—see Table 10 below.

$--!
!

!

Table 10. Environmental Knowledge Video Results by Gender
Paired Samples Statistics
Mean

N

Std.
Deviation

ST Expertise
SE Expertise
ST Scientific
SE Scientific
ST CODE
SE CODE

2.83
2.77
1.5
1.93
4.33
4.7

30
30
30
30
30
30

0.648
0.728
0.572
0.691
0.959
1.208

0.118
0.133
0.104
0.126
0.175
0.221

ST Expertise
SE Expertise
ST Scientific
SE Scientific
ST CODE
SE CODE

2.24
2.82
1.34
1.68
3.58
4.5

38
38
38
38
38
38

0.971
0.896
0.815
0.702
1.464
1.247

0.157
0.145
0.132
0.114
0.237
0.202

Gender
F

Pair 1
Pair 2
Pair 3

M

Pair 1
Pair 2
Pair 3

Std. Error
Mean

Here we see that females attribute greater expertise to the ST video than to the SE
video, however attribute greater scientific content to the SE video than do males. Males on
the other hand attribute much greater expertise (greater than the entire sample mean) to the
SE video and less scientific content to the SE video than the sample mean. Overall, females
attribute greater scientific content and expertise to both videos than do both males and the
overall sample means. Such relationships do not provide sufficient evidence to reject Ho4.

3.3.3.5. Hypothesis 5: Values and Pro-environmental Knowledge Campaigns II
Hypothesis 5 stated that individuals who demonstrate self-transcendent value orientations
attribute greater expertise to a pro-environmental knowledge campaign that relies on selftranscendent value orientations. In order to test this hypothesis the study examined the value
orientations and the video-response questions on an individual participant basis and on an

$-$!
!

!

individual value basis as seen previously in Section 3.3.3.3. Two analyses were conducted. The
first tested the association between individuals classified as SE or ST and the expertise, scientific
content and accuracy of statements made in each of the three videos (including the video that
used fear and insecurity, a cause of SE values).

4

The second tested the individual contribution to

satisfaction responses and the video-response variables listed above. Both analyses looked for
positive associations between SE individuals or values and the SE video, and ST individuals or
values and the ST video. A sample of the results is given below in Tables 11 and 12 (the full
table of results is in Appendix D).
Table 11. Individual Value Orientations and Environmental Knowledge Videos

Variable 1
SE=1/ST=0

Variable 2
ST
EXPERTISE
SE
EXPERTISE
SE-CAUSE
EXPERTISE
ST
SCI.CONTENT
SE
SCI.CONTENT
SE-CAUSE
SCI.
CONTENT
ST
ACCURACY
SE
ACCURACY

df N

Pearson
2
!
Critical
Value

Pearson
Pearson's
2
!
Asymp. R
Approx.
Statistic Signif. (Correla.) Signif.

4

63 9.488

11.541*

0.021

-0.138

0.282

3

61 7.815

4.668

0.198

-0.224

0.083

4

62 9.488

2.068

0.723

0.127

0.325

3

63 7.815

2.196

0.533

0.129

0.313

3

61 7.815

1.922

0.589

-0.082

0.532

3

62 7.815

2.585

0.460

0.139

0.283

3

61 7.815

1.300

0.729

-0.069

0.595

3

60 7.815

1.403

0.705

-0.110

0.402

* denotes 95% Significance
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
4

Questions regarding the accuracy of the narrators’ statement were
asked only for ST and SE videos, not the SE-CAUSE video.
$-%!
!

!

Of the eight video-response variables, the only one to show a significant positive
2

relationship to a value orientation, in this case to self-transcendence, is ST expertise (! = 11.541,
df= 4, and R= -.138). This alone supports Hypothesis 5 (H5), i.e. ST individuals attribute greater
expertise to ST videos than to SE videos. The other seven video-response variables however do
not show significant associations and thus H5 must be rejected.
Table 12. Individual Contributions to Satisfaction and Environmental Knowledge
Videos
Pearson
2
Pearson
!
Pearson's
2
!
Critical
Asymp. R
Approx.
Variable 1
Variable 2
df N Value
Statistic Signif. (Correla.) Signif.
Self-Enhancement Values (Extreme Effects Only)
Personal
ST
Finances
EXPERTISE
4 70 9.488
SE
EXPERTISE
3 68 7.815
SE-CAUSE
EXPERTISE
4 68 9.488
ST
SCI.CONTENT 3 70 7.815
SE SCI.CONT 3 68 7.815
SE-CAUSE
SCI.CONT
3 68 7.815
ST
ACCURACY
3 68 7.815
SE
ACCURACY
3 66 7.815
ST
Purchases
EXPERTISE
2 22 5.991
SE
EXPERTISE
2 21 5.991
How Others ST
Viewed You EXPERTISE
4 39 9.488
SE
EXPERTISE
3 37 7.815
SE-CAUSE
EXPERTISE
4 38 9.488
ST SCICONT
2 39 5.991

$-&!
!

0.853

0.931

-0.009

0.938

2.363

0.501

-0.094

0.446

5.354

0.253

0.211

0.084

2.937
3.778

0.401
0.286

0.118
0.129

0.332
0.294

3.899

0.273

0.208

0.09

7.741

0.052

0.192

0.116

1.288

0.732

0.134

1.081

1.689

0.430

-0.261

0.24

5.923

0.052

0.380

0.089

0.415

0.981

0.020

0.904

1.837

0.607

0.130

0.442

1.759
5.924

0.780
0.052

0.134
0.289

0.423
0.074

!

Table 12 (cont’d)
Self-Transcendent Values (Extreme Effects only)
Local
Environment ST
Quality
EXPERTISE
4 70 9.488
SE
EXPERTISE
3 68 7.815
SE-CAUSE
EXPERTISE
4 68 9.488
ST SCICONT
3 70 7.815
SE SCICONT
3 68 7.815
SE-CAUSE
SCICONT
3 68 7.815
ST
ACCURACY
3 68 7.815
SE
ACCURACY
3 66 7.815

5.125

0.275

-0.008

0.947

3.735

0.291

0.022

0.857

2.949
4.294
6.896

0.566
0.231
0.075

-0.156
0.111
0.302

0.205
0.36
0.012

1.344

0.719

0.061

0.62

3.181

0.365

0.200

0.101

3.561

0.313

0.223

0.071

Table 12 presents the results of the Chi-square tests of individual values and goals (as
represented by the contribution to satisfaction questions) and the video-response questions. The
nearly significant (94.8%) associations present are between the values personal finances and the
2

accuracy of the ST narrator’s statement (! = 7.741, df= 3 and R= .192), purchases and the SE
2

video’s expertise (! = 5.923, df= 2 and R= .380) and how others viewed you and the ST video’s
2

scientific content (! = 5.924, df= 2 and R= .289). While the second association supports
Hypothesis 5, the first and last associations do not.

The lack of consistent significant

associations, either positive or negative, suggests either that the value orientations tested for in
this study are independent of the manner in which the individuals attributed expertise to the ST
and SE videos, or that limitations previously discussed regarding both the manner in which these
value orientations were reported and the videos’ content etc. sufficiently dominate those
associations.

Further research is required, perhaps a study that utilizes multiple pro-

$-'!
!

!

environmental knowledge campaigns, multiple videos, websites, advertisements, podcasts, etc.
each of which utilizes a predominantly ST or SE value-oriented focus.

3.4. Conclusions!
The current study was a conjoint application of ESM and the Apple iPhone on the value
orientation theory of pro-environmental behavior. The results of this application varied. While
the experiment’s design, implementation and analysis all proved that the iPhone is a capable
instrument and an excellent complement to ESM, it remains to be seen whether ESM and the
iPhone used conjointly produce accurate socio-psychological data—much less socioenvironmental data. This experiment in using the iPhone as a survey instrument certainly
implemented an ESM study at a significantly reduced cost. Financial, physical and human
resource requirements were all moderated considerably.
Though twelve iPhones were purchased, this proved unnecessary as only one was used in
the study, and that one exclusively by the researcher.

All preliminary contact with the

participants was done via email or through the participants’ own phone. No training of the
participant or the technology was necessary And only one physical meeting was necessary, this
to distribute the $20 incentive and receive the paper-survey that had been emailed to all
participants. Only one researcher was responsible, and necessary, to conduct the entire ESM
survey, and his responsibilities consisted primarily of emailing and texting the participants
survey questions, as well as receiving the responses and recording and coding the data. Such
data if recorded at the moment of response did not prove burdensome, and seemed comparatively
less than that of single-administration surveys. Furthermore, the recording, coding and analyzing
was and likely can be accomplished while the survey is being conducted, decreasing the

$-(!
!

!

monotony that may commonly accompany such efforts. While the project had a relatively low
number of participants (n=71), this study concludes that a single researcher could conduct a
similar investigation of up to two hundred participants.
The significance of such savings cannot and should not be overlooked. According to the
literature regarding ESM (e.g. Hektner et al, 2007), the number one barrier to using ESM and its
variants outside the field of psychology is its high cost. This study demonstrated that those costs
are only associated with previous research instruments and that adopting novel technologies,
such as the iPhone, eliminate such barriers, making ESM available to other disciplines and
arenas of investigation.
This study concludes that the adoption of ESM is a high-quality alternative to singleadministration surveys, focus groups and interviews. This is extremely important for socioenvironmental researchers. ESM has been shown previously to be adept at reducing recall errors
and biases (ibid), and can if used appropriately—particularly due to the randomness of the
signaling and response schedule—likely eliminate socially desirable responding, satisficing and
part-whole effects. This study was unable to conduct similar investigations using different
methods and instruments to verify these capabilities; however this study argues that such
investigations are not only worthy, but critical.
Brown and Kasser (2005) argued that their research especially needed to be replicated in
a manner that focused on mitigating SDR and recall error. They are not alone. ESM, when used
alongside the iPhone, can also reduce researchers’ reliance on self-reporting, a key source of
inaccuracy in socio-environmental surveys according to Corral-Verdugo (1997). This is possible
because the iPhone allows the researcher to directly monitor the respondents’ behavior through
the use of the device’s camera. This study used the iPhone’s camera to verify individuals’

$-)!
!

!

recycling behavior, their television and videogame behaviors and their food purchases. Such
phenomena, which this study urges is important to socio-environmental investigations, may be
inaccurately self-reported or even degraded through SDR or satisficing. Utilizing the iPhone’s
photographic capabilities provided this study with a possible way around such biases and errors.
The device also allows the respondent to access multimedia and the internet, allowing expanded
types of research, as demonstrated in this study’s use of pro-environmental YouTube videos.
Perhaps ESM’s two greatest advantages are its focus on ecological validity and
intrapersonal variation, and both were on display in this study.

Lab-based experiments,

computer-aided and telephone surveys, focus groups and interviews all extract the respondent
from his or her natural environment. Yet ESM and the iPhone allow for direct access—which is
of particular relevance in socio-environmental research, as most is typically aimed in at least
some way at measuring individuals’ relationship with their environment! The current study
allowed researchers to investigate that relationship while the respondent experienced, both
physically and cognitively, the environment in question.

The exact benefits of such

measurement were outside the purview of the current study, but should certainly be researched in
the future.
ESM is also able to penetrate and examine individuals’ values, beliefs and attitudes,
phenomena that are often time, space or context dependent. And the iPhone decreases the time
and distance between the action and its reporting thus increasing the ecological validity of the
result. In this study researchers were able to ask students about their feelings regarding school
on days the student was in class; it is even possible the students were in class at the very moment
they received the signal. This should increase the accuracy of such responses over paper-surveys
or computer-administered surveys that rely on retrospective reporting.

$-*!
!

!

As socio-environmental researchers become more and more interested in comparing
individuals’ values, worldviews and pro-environmental attitudes, ESM’s focus on intrapersonal
variation is key. It is no longer enough to measure general demographic trends. Instead, one
must now determine what creates transformation within the individual.
The second principal focus of this study was on individuals’ values and well-being, in an
attempt to discern a relationship between those two variables and one’s pro-environmental
behavior. This proved particularly difficult and complex, both because that relationship is
naturally complex and because ESM relies on snapshots of individuals’ well-being and behavior.
These snapshots, as opposed to what single-administration studies (SAS) generate, which are
typically generalized self-reports about one’s behavior averaged over a certain period of time, are
inherently more variable. This variation makes any type of relational analysis much more
difficult. A study such as this because of its limited scope and breadth was not able to rely on
the replication of particular questions and of samples to generate more robust results.
It is also important to note that this study was only able to control for certain variables.
While the sample population was fairly homogenous, i.e. of similar age, in the same class, and at
the same institution, other factors certainly play a role in their behavior, well-being and value
orientations. Certain analyses were run in which vehicle ownership, gender, employment status,
and whether or not students lived in a dorm or in an apartment, were controlled. These analyses
produced similar results to those presented in the tables above and in the appendices. Despite
these analyses, it is not possible to argue that this study’s results are ceteris paribus. For example
individuals’ families, their family’s income, their distance to school and work, their class
schedule, their clothes (personal appearance was one of the values examined) and their

$-+!
!

!

extracurricular activities all likely play a role in the variables and associations examined in this
study. Yet these variables were not controlled for in this study.
Thus, the apparent lack of relation between the variables examined in this study is either
a) truly a result of their independence, b) a result of individuals’ true variability, or c) a result of
variables that were unaccounted for in this examination. While the first two results would be
interesting and worthy of further investigation, the third muddies such a conclusion. Were we to
presume however for a moment that control variables were accounted for in this study’s results,
the inherent variability between and within individuals would be a convincing rationale to pursue
ESM-style investigations such as this. Furthermore, regardless of this study’s results, such
variability if it exists should make researchers think twice about the data that their SAS generate.
Such data may not only be rife with recall error and biases, but it may also only be capturing
individuals’ most typical behavior under certain normalized circumstances—not how that
individual actually behaves typically. It may also capture less tangible phenomena like values or
beliefs outside of their typical context. Such delineation is extremely important, and ESM
studies such as this regardless of their results make that delineation apparent.
Clearly the gulf that separates how individuals feel, believe and act in the moment and
how they feel, believe and act in general is wide. Yet focusing efforts solely on SAS forgets
about that opposite shore. This study does not present evidence that snapshots of well-being,
values and behavior are more robust or revealing. Instead, this study only suggests that such
snapshots may not always align with individuals’ generalized accounts and that more research
focusing on both snapshots and generalized accounts is necessary. Such research should focus
much more in-depth on each of the three phenomena investigated here—amongst many others.

$-,!
!

!

Only in conducting ESM-style investigations can researchers argue that they are
unnecessary. This study suggests the opposite, that they are indeed very necessary.

$$-!
!

!

APPENDICES

$$$!
!

!

APPENDIX A
ESM Survey Questions
(Contribution to Satisfaction Questions are marked with an asterisk)
1. On a scale of 1-7, 1 being not satisfied at all, and 7 being completely satisfied, what number best
describes how satisfied you are with your life these days?
2. On a scale of 1-7, 1 being not satisfied at all, and 7 being completely satisfied, what number best
describes how satisfied you are with how your day today has gone so far?
3. On a scale of 1-7, 1 being completely miserable, and 7 being extremely happy, what number best
describes your level of content so far today?
4. On a scale of 1-7, 1 being not at all, and 7 being very much, what number best describes how
much control you’ve had over your life today?
5. On a scale of 1-7, with 1 being very poorly and 7 being very well, please choose the number that
best describes how well you slept last night.
6. *On a scale of 1-7, with 1 being very poorly and 7 being very well, please choose the number
that best describes how you feel about your personal appearance today.
7. *On a scale of 1-7, with 1 being very poorly and 7 being very well, please choose the number
that best describes how you feel other people have viewed you today.
8. On a scale of 1-7, with 1 being none at all and 7 being very much, please choose the number that
best describes how much time you feel you have had for yourself today.
9. *On a scale of 1-7, with 1 being none at all and 7 being very much, please choose the number
that best describes how successful you’ve felt today.
10. *On a scale of 1-7, with 1 being none at all and 7 being very much, please choose the number
that best describes how often you thought about the quality of the environment around you today.
11. *On a scale of 1-7, 1 being a great negative effect, and 7 being a great positive effect, what
number best describes the effect that your personal finances have had on your level of
satisfaction today?
12. *On a scale of 1-7, 1 being a great negative effect, and 7 being a great positive effect, what
number best describes the effect that your health has had on your level of satisfaction today?
13. *On a scale of 1-7, 1 being a great negative effect, and 7 being a great positive effect, what
number best describes the effect that your relationships (e.g. family, friends, co-workers) has had
on your level of satisfaction today?

$$%!
!

!

14. *On a scale of 1-7, 1 being a great negative effect, and 7 being a great positive effect, what
number best describes the effect that purchases you have made today have had on your level of
satisfaction?
15. *On a scale of 1-7, 1 being a great negative effect, and 7 being a great positive effect, what
number best describes the effect that today’s weather has had on your level of satisfaction?
16. *On a scale of 1-7, 1 being a great negative effect, and 7 being a great positive effect, what
number best describes the effect that your education (or your experience at school today) has had
on your current level of satisfaction?
17. *On a scale of 1-7, 1 being a great negative effect, and 7 being a great positive effect, what
number best describes the effect that your religious beliefs or a religious event have had on your
level of satisfaction today?
18. *On a scale of 1-7, 1 being a great negative effect and 7 being a great positive effect, what
number best describes the effect that your local environment has had on your level of satisfaction
today?
19. If you could purchase any vehicle today, no matter the cost (the vehicle would cost you nothing),
what vehicle would you purchase? (Please be as specific as possible)
20. If you could move to any city in the world today, what city would it be?
21. If you could do anything in the world tonight, and money was not a concern, what would you
do? (Please be as specific as possible)
22. At your earliest convenience, please use your iPhone’s camera to take a picture(s) of the inside
and outside of your refrigerator.
23. At your earliest convenience, please use your iPhone’s camera to take a picture(s) of your
television set.
24. At your earliest convenience, please use your iPhone’s camera to take a picture(s) of something
outside your home that makes you feel pleasant or happy.
25. At your earliest convenience, please use your iPhone’s camera to take a picture(s) of where you
store your garbage can(s) and/or recycling container(s).
26. At your earliest convenience, please use your iPhone’s camera to take a picture of the odometer
in your automobile.
27. Please text back one, and only one, of the following as what you used for your primary mode of
transportation today so far: Automobile, Bus, Motorcycle, Scooter, Bicycle, Walking

$$&!
!

!

28. Please text back approximately how many miles you drove your car yesterday.
29. Please text back how many trips you made (or errands you ran) on your bicycle yesterday.
(Please count each roundtrip as only 1)
30. Please text back how many hours and/or minutes you have spent in front of a computer today?
31. Please text back how many hours and/or minutes you spent in front of a computer yesterday?
32. Please text back approximately how many minutes you spent on the telephone yesterday.
33. Please text back approximately how many minutes you have spent on the telephone today.
34. Please text back approximately how many hours and/or minutes you spent watching television or
movies yesterday.
35. Please text back approximately how many hours and/or minutes you have spent watching
television or movies today.
36. Please text back the number of meals you ate yesterday at home and the number you ate
yesterday at a restaurant. [Example: For 2 meals eaten at home and 1 at a restaurant, text “2-1”]
37. Please text back the number of meals you have eaten today at home and the number you ate at a
restaurant. [Example: For 2 meals eaten at home and 1 at a restaurant, text “2 home-1 away”]
38. Please text back approximately how many hours and/or minutes you spent online today.
39. Please watch the video “This Bulb” by clicking
http://www.youtube.com/watch?v=FvOBHMb6Cqc

on

the

following

hyperlink:!

40. Having watched the video “This Bulb,” what level of expertise do you feel it demonstrated?
i. None at all
ii. Very little
iii. Some
iv. Much
v. Very much
41. Having watched the video “This Bulb,” how would you describe the video’s content?
i. Not at all scientific
ii. A little scientific
iii. Somewhat scientific
iv. Very scientific

$$'!
!

!

42. In the video “This bulb,” Natalie Portman purports that “’this bulb’ will help reverse the effects
of global warming.” How accurate do you believe this statement to be?
i. Not at all accurate
ii. A little accurate
iii. Somewhat accurate
iv. Very accurate
43. Having watched the video “This Bulb,” how much do you feel you yourself have contributed to
climate change TODAY?
i. Much less than the average American
ii. Moderately less than the average American
iii. The same as the average American
iv. More than the average American
v. Much more than the average American
vi. I have not contributed to climate change today
44. Having watched the video “This Bulb,” how likely are you to replace any of the light bulbs in
your house or apartment with compact fluorescent bulbs (as seen in the video)?
i. Very unlikely,
ii. Unlikely,
iii. Likely,
iv. Very likely,
v. Already use compact fluorescent bulbs
45. Please watch the video “Unchained Goddess” by clicking on the following hyperlink:
http://www.youtube.com/watch?v=qF9WdV8pUPk
46. Having watched “Unchained Goddess,” a movie clip produced in 1958, what level of expertise
do you feel its narrator Dr. Frank Baxter demonstrated?
i. None at all
ii. Very little
iii. Some
iv. Much
v. Very much
47. Having watched “Unchained Goddess,” a movie clip produced in 1958, how would you describe
the video’s content?
i. Not at all scientific
ii. A little scientific
iii. Somewhat scientific
iv. Very scientific

$$(!
!

!

48. What effect does the fact that “Unchained Goddess” was filmed in 1958 have on your
willingness to believe its narrator Dr. Baxter?
i. Because of the video’s age, I am less willing to believe Dr. Baxter
ii. Because of the video’s age, I am more willing to believe Dr. Baxter
iii. I am equally willing to believe Dr. Baxter regardless of the video’s age
iv. Regardless of the video’s age, I do not believe Dr. Baxter
49. Having watched “Unchained Goddess,” what level of responsibility do you feel personally for
climate change?
i. None at all
ii. Very little
iii. Some
iv. Much
v. Very much
50. Having watched “Unchained Goddess,” who do you feel is most responsible for climate change?
i. No human being is responsible
ii. My parents and grandparents’ generations are most responsible
iii. My generation is most responsible
iv. Future generations are most responsible
51. Having watched “Unchained Goddess,” who do you feel is most responsible for taking action to
prevent or mitigate (slow down) climate change?
i. No human being is responsible
ii. My parents and grandparents’ generations are most responsible
iii. My generation is most responsible
iv. Future generations are most responsible
52. Please watch the video “Andrew Winston at Sustain 08” by clicking on the following hyperlink:
http://www.youtube.com/watch?v=xtpHzVplr1U
53. Having watched the video, “Andrew Winston at Sustain 08,” what level of expertise do you feel
its narrator Andrew Winston demonstrated?
i. None at all
ii. Very little
iii. Some
iv. Much
v. Very much
54. Having watched the video “Andrew Winston at Sustain 08,” how would you describe the video’s
content?
i. Not at all scientific
ii. A little scientific
iii. Somewhat scientific
iv. Very scientific

$$)!
!

!

55. In the video “Andrew Winston at Sustain 08,” Andrew Winston describes the “forces driving
sustainability” as “natural resource pressures, resource constraints, how much water we have,
and how much stable climate we have.” How accurate do you believe this statement to be?
i. Not at all accurate
ii. A little accurate
iii. Somewhat accurate
iv. Very accurate
56. Having watched the video, “Andrew Winston at Sustain 08,” how would you describe yourself?
i. Willing to pay significantly more (prices 20% or higher) for “green” products
ii. Willing to pay moderately more (prices 10-19% higher) for “green” products
iii. Willing to pay slightly more (prices 1-9% higher) for “green” products
iv. Not willing to pay more for “green” products, but would choose similarly priced “green”
products over “non-green” products
v. Not willing to purchase “green” products at any price

$$*!
!

!

APPENDIX B
Values and Pro-Environmental Behavior Data
Table 13. Values and Pro-Environmental Behavior

Variable 1

Variable 2

SE=1/ST=0 REC
SE=1/ST=0 TRANS
SE=1/ST=0 PEB

df N

Pearson
2
!
Critical
Value

Pearson
2
!
Statistic

Asymp.
Signif.

Pearson's
R
(Correla.)

1
2
4

3.841
5.991
9.488

0.184
4.237
1.973

0.668
0.120
0.741

0.060
-0.192
-0.064

0.675
0.128
0.621

2.263
1.115
4.145

0.132
0.573
0.387

-0.199
-0.113
-0.165

0.137
0.346
0.172

0.524
0.344
4.966

0.469
0.842
0.291

0.096
0.038
0.088

0.478
0.753
0.47

0.007
0.624
4.056

0.932
0.732
0.399

-0.011
0.078
0.056

0.934
0.519
0.647

51
64
63

Self-Enhancement Values (Extreme Effects Only)
Personal
Appearance REC
1 57 3.841
TRANS
2 71 5.991
PEB
4 70 9.488
Personal
Finances
REC
1 57 3.841
TRANS
2 71 5.991
PEB
4 70 9.488
Feeling of
Success
REC
1 56 3.841
TRANS
2 70 5.991
PEB
4 69 9.488
Purchases
Others
Viewed

Approx.
Signif.

REC
TRANS
PEB

1
2
4

17
22
21

3.841
5.991
9.488

0.052
1.257
1.125

0.819
0.533
0.890

0.056
-0.238
-0.171

0.832
0.285
0.458

REC
TRANS
PEB

1
2
4

31
40
39

3.841
5.991
9.488

0.040
1.887
2.372

0.842
0.389
0.668

0.036
-0.016
0.070

0.849
0.924
0.671

1
2
4

57
70
69

3.841
5.991
9.488

0.118
0.537
1.170

0.731
0.764
0.883

0.045
-0.067
-0.042

0.737
0.582
0.731

1
2
4

57
70
69

3.841
5.991
9.488

0.427
1.908
0.786

0.514
0.385
0.940

-0.087
-0.018
-0.085

0.522
0.882
0.488

Self-Transcendent Values
Religious
Events
REC
TRANS
PEB
Relationships
REC
TRANS
PEB

$$+!
!

!

Table 13 (cont’d)

Variable 2

df N

Pearson
2
!
Critical
Value

REC
TRANS
PEB

1
2
4

57
71
70

3.841
5.991
9.488

0.470
3.101
4.518

0.493
0.212
0.340

-0.091
-0.053
-0.106

0.502
0.658
0.381

REC
TRANS
PEB

1
2
4

56
68
67

3.841
5.991
9.488

0.383
2.942
2.610

0.536
0.230
0.625

-0.083
0.065
0.036

0.545
0.6
0.772

Value-free?
Education/
Class
REC
TRANS
PEB

1
2
4

48
61
60

3.841
5.991
9.488

2.350
3.119
3.039

0.125
0.210
0.551

-0.221
0.225
0.029

0.131
0.081
0.826

1
2
4

56
70
69

3.841
5.991
9.488

0.394
0.824
3.661

0.530
0.662
0.454

-0.084
-0.051
-0.099

0.539
0.673
0.416

0.508
0.140
0.331

-0.088
-0.143
-0.157

0.517
0.235
0.194

0.886
0.234
0.245

-0.019
-0.047
-0.079

0.889
0.699
0.514

0.952
0.811
0.857

-0.008
-0.075
-0.096

0.953
0.538
0.43

0.453
0.048
0.556

0.182
-0.454*
-0.213

0.485
0.034
0.353

Variable 1
Environment
Quality

Weather

Health

REC
TRANS
PEB

Pearson
2
!
Statistic

Self-Enhancement Values (All Effects (including minimal))
Personal
Appearance REC
1 57 3.841
0.438
TRANS
2 71 5.991
3.932
PEB
4 70 9.488
4.597
Personal
Finances
REC
1 57 3.841
0.020
TRANS
2 71 5.991
2.901
PEB
4 70 9.488
5.440
Feeling of
Success
REC
1 56 3.841
0.004
TRANS
2 70 5.991
0.419
PEB
4 69 9.488
1.328
Purchases

REC
TRANS
PEB

1
2
4

17
22
21

3.841
5.991
9.488

$$,!
!

0.562
6.079*
3.012

Asymp.
Signif.

Pearson's
R
(Correl.)

Approx.
Signif.

!

Table 13 (cont’d)

Variable 2

df N

Pearson
2
!
Critical
Value

REC
TRANS
PEB

1
2
4

31
40
39

3.841
5.991
9.488

0.040
2.364
1.803

0.841
0.307
0.772

0.036
-0.148
-0.037

0.847
0.361
0.823

1
2
4

57
70
69

3.841
5.991
9.488

0.001
2.119
3.137

0.977
0.347
0.535

0.004
-0.148
-0.114

0.978
0.222
0.352

1
2
4

57
70
69

3.841
5.991
9.488

2.144
1.504
0.595

0.143
0.472
0.964

-0.194
0.049
-0.051

0.148
0.688
0.679

1
2
4

57
71
70

3.841
5.991
9.488

0.994
1.804
4.494

0.319
0.406
0.343

-0.132
-0.147
-0.173

0.327
0.222
0.153

REC
TRANS
PEB

1
2
4

56
68
67

3.841
5.991
9.488

0.000
0.266
1.808

0.983
0.875
0.771

-0.003
-0.006
-0.026

0.983
0.964
0.837

Value-free?
Education/
Class
REC
TRANS
PEB

1
2
4

48
61
60

3.841
5.991
9.488

0.032
6.053*
2.453

0.858
0.048
0.653

0.026
0.315
0.170

0.862
0.014
0.195

1
2
4

56
70
69

3.841
5.991
9.488

0.004
0.970
4.696

0.952
0.616
0.320

-0.008
0.003
-0.010

0.953
0.979
0.934

Variable 1
How
Others
Viewed

Self-Transcendent Values
Religious
Events
REC
TRANS
PEB
Relationships
REC
TRANS
PEB
Environment
Quality
REC
TRANS
PEB
Weather

Health

REC
TRANS
PEB

Pearson
2
!
Statistic

Asymp.
Signif.

Pearson's
R
(Correl.)

Approx.
Signif.

* denotes 95% Significance
** denotes 99% Significance

$%-!
!

!

APPENDIX C
Values and Subjective Well-Being Data
Table 14. Values and Subjective Well-Being

Variable 1
SE=1/ST=0

Variable
2
TOTAL
SWB
GEN
SWB
SNAP
SWB

Pearson
2
!
Critical
Value

Pearson
2
!
Statistic

Asymp.
Sig.

Pearson's
R
Approx.
(Correl.) Signif.

21 63

32.671

17.505

0.680

-0.113

0.377

9

64

16.919

6.429

0.696

-0.073

0.564

15 64

24.996

16.548

0.347

-0.101

0.426

0.028

0.370**

0.002

0.234

0.325**

0.006

0.157

0.288*

0.015

0.383

-0.224

0.063

0.516

-0.147

0.22

0.276

-0.160

0.183

0.564

0.084

0.492

0.256

0.139

0.253

0.446

0.051

0.674

0.336

0.163

0.479

0.512

-0.073

0.748

0.388

0.173

0.44

df

N

Self-Enhancement Values (Extreme Effects Only)
Personal
TOTAL
Appearance SWB
23 70 35.172
37.568**
GEN
SWB
9 71 16.919
11.639
SNAP
SWB
15 71 24.996
20.399
Personal
TOTAL
Finances
SWB
23 70 35.172
24.384
GEN
SWB
9 71 16.919
8.180
SNAP
SWB
15 71 24.996
17.753
Feeling of
TOTAL
Success
SWB
23 69 35.172
21.274
GEN
SWB
9 70 16.919
11.295
SNAP
SWB
15 70 24.996
15.080
TOTAL
Purchases
SWB
11 21 19.675
12.375
GEN
SWB
6 22 12.592
5.254
SNAP
SWB
10 22 18.307
10.620

$%$!
!

!

Table 14 (cont’d)

Variable 1 Variable 2

df

Self-Transcendent Values
Religious
TOTAL
Events
SWB
23
GEN
SWB
9
SNAP
SWB
15
RelationTOTAL
ships
SWB
23
GEN
SWB
9
SNAP
SWB
15
Environmental
TOTAL
Quality
SWB
23
GEN
SWB
9
SNAP
SWB
15
TOTAL
Weather
SWB
23
GEN
SWB
9
SNAP
SWB
15
Value-free?
Education/ TOTAL
Class
SWB
21
GEN
SWB
9
SNAP
SWB
14

N

Pearson
2
!
Critical
Value

Pearson
2
!
Statistic

Asymp.
Sig.

Pearson's
R
Approx.
(Correl.) Signif.

69

35.172

20.487

0.612

0.157

70

16.919

9.789

0.368

0.029

0.812

70

24.996

10.639

0.778

0.212

0.078

69

35.172

18.977

0.703

0.223

0.066

70

16.919

10.163

0.337

0.276**

0.021

70

24.996

11.577

0.711

0.125

0.302

70

35.172

30.675

0.131

-0.009

0.939

71

16.919

7.381

0.598

0.042

0.729

71

24.996

14.809

0.465

-0.065

0.589

67

35.172

22.383

0.497

0.238*

0.052

68

16.919

5.357

0.802

0.144

0.241

68

24.996

23.111

0.082

0.247*

0.042

60

32.671

26.451

0.190

0.123

0.347

61

16.919

7.649

0.570

0.088

0.499

61

23.685

12.752

0.546

0.124

0.343

$%%!
!

0.197

!

Table 14 (cont’d)

Variable 1

Variable
2

df

N

Pearson
2
!
Critical
Value

Pearson
2
!
Statistic

Self-Enhancement Values (All Effects (Including minimal))
Personal
TOTAL
Appearance SWB
23 70 35.172
14.276
GEN
SWB
9 71 16.919
6.120
SNAP
SWB
15 71 24.996
17.853
Personal
TOTAL
Finances
SWB
23 70 35.172
15.108
GEN
SWB
9 71 16.919
3.477
SNAP
SWB
15 71 24.996
20.945
Feeling of
TOTAL
Success
SWB
23 69 35.172
25.408
GEN
SWB
9 70 16.919
5.256
SNAP
SWB
15 70 24.996
17.956
TOTAL
SWB
Purchases
11 21 19.675
12.296
GEN
SWB
6 22 12.592
5.018
SNAP
SWB
10 22 18.307
15.207
How Others TOTAL
Viewed
SWB
11 39 19.675
8.580
GEN
SWB
9 40 16.919
13.810
SNAP
SWB
8 40 15.507
4.352
Self-Transcendent Values
Religious
TOTAL
Events
SWB
23 69
GEN
SWB
9 70
SNAP
SWB
15 70

Asymp.
Sig.

Pearson's
R
(Correl.)

Approx.
Signif.

0.919

0.158

0.191

0.728

0.030

0.803

0.270

0.225

0.059

0.891

-0.001

0.995

0.942

-0.097

0.421

0.139

0.104

0.39

0.330

-0.061

0.619

0.811

-0.033

0.784

0.265

-0.042

0.729

0.342

0.354

0.115

0.542

-0.010

0.946

0.125

0.305

0.167

0.661

0.084

0.611

0.129

0.087

0.595

0.824

0.059

0.719

35.172

21.320

0.567

-0.016

0.896

16.919

7.702

0.564

-0.171

0.156

24.996

16.095

0.376

0.135

0.265

$%&!
!

!

Table 14 (cont’d)

Variable 1
Relationships

Environmental
Quality

Weather

Value-free?
Education/
Class

Health

Variable
2
TOTAL
SWB
GEN
SWB
SNAP
SWB
TOTAL
SWB
GEN
SWB
SNAP
SWB
TOTAL
SWB
GEN
SWB
SNAP
SWB
TOTAL
SWB
GEN
SWB
SNAP
SWB
TOTAL
SWB
GEN
SWB
SNAP
SWB

Pearson
2
!
Critical
Value

Pearson
2
!
Statistic

Asymp.
Sig.

Pearson's
R
Approx.
(Correl.) Signif.

23 69

35.172

16.802

0.819

-0.005

0.965

9

70

16.919

7.036

0.633

0.136

0.263

15 70

24.996

11.576

0.711

-0.112

0.355

23 70

35.172

18.427

0.734

-0.004

0.972

9

71

16.919

10.939

0.280

0.109

0.367

15 71

24.996

8.159

0.917

-0.103

0.393

23 67

35.172

11.772

0.974

0.005

0.969

9

68

16.919

6.309

0.709

-0.005

0.967

15 68

24.996

17.166

0.309

0.032

0.799

21 60

32.671

26.683

0.182

0.128

0.33

9

61

16.919

12.621

0.181

0.207

0.109

14 61

23.685

15.924

0.318

0.013

0.92

22 69

33.924

22.501

0.430

0.054

0.659

9

70

16.919

3.947

0.915

0.018

0.884

14 70

23.685

11.234

0.668

0.037

0.759

df

N

* denotes 95% Significance
** denotes 99% Significance

$%'!
!

!

APPENDIX D
Values and Environmental Knowledge Videos Data
Table 15. Values and Environmental Knowledge Videos

Variable 1

Variable 2

ST
SE=1/ST=0 EXPERTISE
SE
EXPERTISE
SE-CAUSE
EXPERT
ST SCICONT
SE SCICONT
SE-CAUSE
SCICONT
ST
ACCURACY
SE
ACCURACY

df N

Pearson
2
!
Critical
Value

ST
EXPERTISE
SE
EXPERTISE

Asymp.
Sig.

Pearson's
R
Approx.
(Correl.) Sig.

4

63 9.488

11.541*

0.021

-0.138

0.282

3

61 7.815

4.668

0.198

-0.224

0.083

4
3
3

62 9.488
63 7.815
61 7.815

2.068
2.196
1.922

0.723
0.533
0.589

0.127
0.129
-0.082

0.325
0.313
0.532

3

62 7.815

2.585

0.460

0.139

0.283

3

61 7.815

1.300

0.729

-0.069

0.595

3

60 7.815

1.403

0.705

-0.110

0.402

1.300

0.861

-0.029

0.814

1.269

0.737

0.080

0.515

3.499
3.916
2.234

0.478
0.271
0.525

0.176
0.196
-0.037

0.151
0.104
0.763

2.302

0.512

0.032

0.796

5.699

0.127

-0.171

0.164

2.792

0.425

-0.016

0.9

Self-Enhancement Values (Extreme Effects Only)
Personal
ST
Appearance EXPERTISE 4 70 9.488
SE
EXPERTISE 3 68 7.815
SE-CAUSE
EXPERTISE 4 68 9.488
ST SCICONT 3 70 7.815
SE SCICONT 3 68 7.815
SE-CAUSE
SCICONT
3 68 7.815
ST
ACCURACY 3 68 7.815
SE
ACCURACY 3 66 7.815
Personal
Finances

Pearson
2
!
Statistic

4

70 9.448

0.853

0.931

-0.009

0.938

3

68 7.815

2.363

0.501

-0.094

0.446

$%(!
!

!

Table 15 (cont’d)

Variable 1

Variable 2

Pearson
2
!
Critical
df N Value

Pers. Finan.

SE-CAUSE
EXPERTISE
ST SCICONT

4 68 9.448
3 70 7.815

5.354
2.937

0.253
0.401

0.211
0.118

0.084
0.332

3 68 7.815

3.778

0.286

0.129

0.294

3 68 7.815

3.899

0.273

0.208

0.09

3 68 7.815

7.741

0.052

0.192

0.116

3 66 7.815

1.288

0.732

0.134

1.081

4 69 9.488

2.424

0.658

-0.005

0.966

3 67 7.815

4.390

0.222

0.039

0.753

4 67 9.488
3 69 7.815
3 67 7.815

1.515
1.564
2.202

0.824
0.668
0.531

0.022
-0.050
0.012

0.859
0.681
0.922

3 67 7.815

0.707

0.871

-0.080

0.522

3 67 7.815

3.121

0.373

-0.207

0.092

3 65 7.815

2.035

0.565

0.171

0.173

2 22 5.991

1.689

0.430

-0.261

0.24

2 21 5.991

5.923

0.052

0.380

0.089

4 21 9.488
3 22 7.815
2 21 5.991

2.250
3.634
1.081

0.690
0.304
0.582

-0.166
-0.040
-0.123

0.472
0.861
0.597

3 21 7.815
2 21 5.991

3.900
0.915

0.272
0.633

-0.046
0.162

0.843
0.483

2 20 5.991

0.899

0.638

0.212

0.371

SE SCICONT
SE-CAUSE
SCICONT
ST
ACCURACY
SE
ACCURACY
Feeling of
Success

Purchases

ST
EXPERTISE
SE
EXPERTISE
SE-CAUSE
EXPERTISE
ST SCICONT
SE SCICONT
SE-CAUSE
SCICONT
ST
ACCURACY
SE
ACCURACY
ST
EXPERTISE
SE
EXPERTISE
SE-CAUSE
EXPERTISE
ST SCICONT
SE SCICONT
SE-CAUSE
SCICONT
ST ACCUR.
SE
ACCURACY

$%)!
!

Pearson
2
!
Statistic

Asymp.
Sig.

Pearson's
R
Approx.
(Correl.) Sig.

!

Table 15 (cont’d)

Variable 1

Variable 2

How Others
Viewed

ST
EXPERTISE
SE
EXPERTISE
SE-CAUSE
EXPERTISE
ST SCICONT
SE SCICONT
SE-CAUSE
SCICONT
ST
ACCURACY
SE
ACCURACY

Self-Transcendent Values
Religious
ST
Events
EXPERTISE
SE
EXPERTISE
SE-CAUSE
EXPERTISE
ST SCICONT
SE SCICONT
SE-CAUSE
SCICONT
ST
ACCURACY
SE
ACCURACY
Local
Environment ST
Quality
EXPERTISE
SE
EXPERTISE
SE-CAUSE
EXPERTISE
ST SCICONT
SE SCICONT

Pearson
2
!
Critical
df N Value

Pearson
2
!
Statistic

Asymp.
Sig.

Pearson's
R
Approx.
(Correl.) Sig.

4 39 9.488

0.415

0.981

0.020

0.904

3 37 7.815

1.837

0.607

0.130

0.442

4 38 9.488
2 39 5.991
3 37 7.815

1.759
5.924
5.069

0.780
0.052
0.167

0.134
0.289
0.166

0.423
0.074
0.328

3 38 7.815

4.479

0.214

0.042

0.804

3 37 7.815

5.964

0.113

0.122

0.473

3 36 7.815

4.655

0.199

0.316

0.061

4 70 9.488

3.905

0.419

0.050

0.679

3 68 7.815

1.748

0.626

0.039

0.75

4 67 9.488
3 70 7.815
3 68 7.815

1.133
3.111
3.134

0.889
0.375
0.371

0.044
0.088
0.098

0.722
0.469
0.426

3 67 7.815

1.367

0.713

0.019

0.877

3 68 7.815

0.058

0.996

0.009

0.939

3 66 7.815

2.100

0.552

0.040

0.749

4 70 9.488

5.125

0.275

-0.008

0.947

3 68 7.815

3.735

0.291

0.022

0.857

4 68 9.488
3 70 7.815
3 68 7.815

2.949
4.294
6.896

0.566
0.231
0.075

-0.156
0.111
0.302

0.205
0.36
0.012

$%*!
!

!

Table 15 (cont’d)

Variable 1
Env. Quality

Weather

Education/
Class

Variable 2
SE-CAUSE
SCICONT
ST
ACCURACY
SE
ACCURACY
ST
EXPERTISE
SE
EXPERTISE
SE-CAUSE
EXPERTISE
ST SCICONT
SE SCICONT
SE-CAUSE
SCICONT
ST
ACCURACY
SE
ACCURACY
ST
EXPERTISE
SE
EXPERTISE
SE-CAUSE
EXPERTISE
ST SCICONT
SE SCICONT
SE-CAUSE
SCICONT
ST
ACCURACY
SE
ACCURACY

Pearson
2
!
Critical
df N Value

Pearson
2
!
Statistic

Asymp.
Sig.

Pearson's
R
Approx.
(Correl.) Sig.

3 68 7.815

1.344

0.719

0.061

0.62

3 68 7.815

3.181

0.365

0.200

0.101

3 66 7.815

3.561

0.313

0.223

0.071

4 68 9.488

3.819

0.431

-0.010

0.937

3 67 7.815

3.491

0.322

-0.103

0.406

4 65 9.488
3 68 7.815
3 67 7.815

1.451
0.812
1.927

0.835
0.847
0.588

-0.037
-0.106
-0.016

0.772
0.39
0.901

3 65 7.815

0.954

0.812

-0.105

0.405

3 66 7.815

0.595

0.898

0.008

0.946

3 65 7.815

2.204

0.531

0.131

0.298

4 60 9.488

3.218

0.522

-0.197

0.132

3 58 7.815

1.669

0.644

-0.153

0.253

4 58 9.488
3 60 7.815
3 58 7.815

4.606
3.956
3.456

0.330
0.266
0.327

0.170
0.038
0.027

0.202
0.776
0.838

3 58 7.815

0.751

0.861

0.054

0.688

3 58 7.815

2.877

0.411

0.043

0.746

3 56 7.815

0.655

0.884

0.075

0.583

$%+!
!

!

Table 15 (cont’d)

Variable 1
Health

Variable 2
ST
EXPERTISE
SE
EXPERTISE
SE-CAUSE
EXPERTISE
ST SCICONT
SE SCICONT
SE-CAUSE
SCICONT
ST
ACCURACY
SE
ACCURACY

Pearson
2
!
Critical
df N Value

Pearson
2
!
Statistic

Asymp.
Sig.

Pearson's
R
Approx.
(Correl.) Sig.

4 69 9.488

3.780

0.437

0.007

0.953

3 67 7.815

0.597

0.897

0.068

0.583

4 68 9.488
3 69 7.815
3 67 7.815

0.871
2.767
7.463

0.929
0.429
0.059

0.068
0.016
-0.049

0.58
0.895
0.691

3 68 7.815

5.636

0.131

-0.227

0.063

3 67 7.815

1.349

0.718

0.048

0.701

3 65 7.815

0.272

0.965

0.024

0.849

* denotes 95% Significance
** denotes 99% Significance

$%,!
!

!

REFERENCES

$&-!
!

!

REFERENCES
./0123"456!786!92:#";<;6!=86!>!7"??"01@6!A8!B$,,'C8!D21E"F"FG!H;#;0?;/!0F/!IJ0J;/!
K<;L;<;F4;@!A;JM2/@!L2<!N0?:"FG!OF#"<2F1;FJ0?!.1;F"J";@8!!"#$%&'(")(
*%+,$"%-.%/&'(*0"%"-,01(&%2(3&%&4.-.%/5(676!%*$P%,%8!
.M:#"06!.86!>!72FG6!Q8!R8!B%--%C8!K;<@2F0?"JS!0F/!#0?:;@!E0@;/!10J;<"0?"@1T!UM;"<!
<;?0J"2F@M"V!0F/!2<"G"F@8!!"#$%&'(")(8"%1#-.$(91:0;"'"4:5(<66!&+,P'-%8!
.W5;F6!X8!B$,,$C8!UM;!UM;2<S!2L!K?0FF;/!Y;M0#"2<8!=$4&%,>&/,"%&'(?.;&+,"$(&%2(@#-&%(
A.0,1,"%(9$"0.11.15(BC6!$*,P%$$8!
.W5;F6!X86!>!Z"@ME;"F6!A8!B$,+-C8!D%2.$1/&%2,%4(E//,/#2.1(&%2(9$.2,0/,%4(F"0,&'(?.;&+,"$8!
OFG?;322/!D?"LL@6!Q=T!K<;FJ"4;![0??8!
.??;F6!=8!Y86!>!Z;<<0F/6!=8!98!B$,,,C8!OF#"<2F1;FJ0?!?24:@!2L!42FJ<2?6!@S1V0JMS6!0F/!
V<2;F#"<2F1;FJ0?!E;M0#"2<T!.!J;@J!2L!\;??;<]@!04J"#;?S!40<"FG!MSV2JM;@"@8!
*%+,$"%-.%/(&%2(?.;&+,"$5(G<6!&&+P&(&8!
.1;?0FG6!A86!>!Y0<J:@@;^6!_8!B$,*-C8!`FJ;<@:4M:FG;F!5:<!N0?"/"J0J!;"F;<!F;:;F!9:G;FP
I^0?0!aN0?"/"JS!@J:/";@!2L!0!F;3!?S"FG!@40?;b8!A,&4%"1/,0&5(<76!$-&P$%%8!
Y0G255"6!H8!K8!B$,,%C8!UM;!I;?LPH;G:?0J"2F!2L!.JJ"J:/;@6!XFJ;FJ"2F@6!0F/!Y;M0#"2<8!F"0,&'(
91:0;"'"4:(H#&$/.$':5(BBB%C6!$*+P%-'8!
Y0"?;S6!c8!_8!B$,,'C8!3./;"21(")(F"0,&'(I.1.&$0;!BZ2:<JM!O/"J"2F!;/8C8!Q;3!R2<^T!UM;!Z<;;!
K<;@@8!
Y0F/:<06!.8!B$,**C8!I;?LPOLL"404ST!U320</!0!`F"LS"FG!UM;2<S!2L!Y;M0#"2<0?!DM0FG;8!
91:0;"'"4,0&'(I.+,.J5(KL6!$,$P%$(8!
Y0F/:<06!.8!B$,+%C8!I;?LPOLL"404S!A;4M0F"@1!"F![:10F!.G;F4S8!E-.$,0&%(91:0;"'"4,1/5(GM6!
$%%P$'*8!
Y0F/:<06!.86!./01@6!Q8!O86!>!Y;S;<6!=8!B$,**C8!D2GF"J"#;!V<24;@@;@!1;/"0J"FG!E;M0#"2<0?!
4M0FG;8!!"#$%&'(")(9.$1"%&',/:(&%2(F"0,&'(91:0;"'"4:5(GB6!$%(P$&,8!

$&$!
!

!

Y0F/:<06!.86!./01@6!Q8!O86![0</S6!.8!Y86!>![23;??@6!\8!Q8!B$,+-C8!U;@J@!2L!JM;!G;F;<0?"JS!2L!
@;?LP;LL"404S!JM;2<S8!8"4%,/,+.(N;.$&O:(&%2(I.1.&$0;5(L6!&,P))8!
Y0F;<W;;6!H86!>!_"JJ10<6![8!B%--+C8!XF/"#"/:0?!/"LL;<;F4;@!"F!4M"?/<;F]@!10J;<"0?"@1T!UM;!<2?;!
2L!V;;<!<;?0J"2F@M"V@8!9.$1"%&',/:(&%2(F"0,&'(91:0;"'"4:(?#''./,%5(GL6!$*P&$8!
Y0<J?;JJ6!Z8!D8!B$,&%C8!I.-.-P.$,%48!D01E<"/G;T!D01E<"/G;!`F"#;<@"JS!K<;@@8!
Y0J";6!I8!B%--+C8!7"4^;/!K<2E?;1@!0F/!.VV?";/!O42F21"4@8!E-.$,0&%(!"#$%&'(")(E4$,0#'/#$&'(
*0"%"-,015(B6!$$*)P$$,$8!
YM0JJ04M0<SS06!\8!c86!>!=2MF@2F6!H8!B$,**C8!F/&/,1/,0&'(8"%0.O/1(&%2(3./;"218!Q;3!R2<^T!=2MF!
7"?;S!>!I2F@8!
Y"4^10F6!98!B$,*%C8!OF#"<2F1;FJ0?!0JJ"J:/;@!0F/!04J"2F@8!!"#$%&'(")(F"0,&'(91:0;"'"4:5(
KMB%C6!&%&P&%'8!
Y?0^;6!=8!B$,,,C8!d#;<421"FG!JM;!]N0?:;P.4J"2F!\0V]!"F!OF#"<2F1;FJ0?!K2?"4ST!J;F@"2F@!
E;J3;;F!F0"J2F0?!V2?"4S!0F/!?240?!;eV;<";F4;8!Q"0&'(*%+,$"%-.%/5(LB&C6!%(*P%*+8!
Y?0^;6!=86!>!_0#"@6!c8!B$,)'C8!Q2<1@6!N0?:;@!0F/!I0F4J"2F@8!XF!H8!98!Z0<"@!BO/8C6!@&%2P""R(")(
3"2.$%(F"0,"'"4:!BVV8!'()P'+'C8!DM"40G2T!H0F/PA4Q0??S8!
Y2/;FM0:@;F6!\8!N86!>!7S;<6!H8!I8!B$,+*C8!I24"0?!D2GF"J"2F!0F/!I24"0?!H;0?"JST!XFL2<10J"2F!
.4f:"@"J"2F!0F/!`@;!"F!JM;!90E2<0J2<S!0F/!JM;!H;0?!72<?/8!XF![8!=8!["VV?;<6!Q8!
I4M30<5!>!I8!I:/10F!BO/@8C6!F"0,&'(S%)"$-&/,"%(9$"0.11,%4(&%2(F#$+.:(3./;"2"'"4:!
BVV8!)P'$C8!Q;3!R2<^T!IV<"FG;<PN;<?0G8!
Y2</;F6!_86!>!Z<0F4"@6!=8!98!B$,*+C8!7M2!40<;@!0E2:J!;42?2GSg!K;<@2F0?"JS!0F/!@;e!/"LL;<;F4;!
"F!;F#"<2F1;FJ0?!42F4;<F8!!"#$%&'(")(9.$1"%&',/:5(L76!$,-P%-&8!
Y<;M16!=8!78!B$,))C8!E(/;."$:(")(O1:0;"'"4,0&'($.&0/&%0.8!Q;3!R2<^T!.40/;1"4!K<;@@8!
Y<23F6!c8!786!>!c0@@;<6!U8!B%--(C8!.<;!K@S4M2?2G"40?!0F/!O42?2G"40?!7;??PY;"FG!D21V0J"E?;g!!
UM;!H2?;!2L!N0?:;@6!A"F/L:?F;@@6!0F/!9"L;@JS?;8!F"0,&'(S%2,0&/"$1(I.1.&$0;5(ML6!&',P
&)+8!

$&%!
!

!

Y<23F6!A86!>!9;2F6!78!B$,,,C8!N;.(8"%1#-.$T1(U#,2.(/"(*)).0/,+.(*%+,$"%-.%/&'(8;",0.18!
Q;3!R2<^T!UM<;;!H"#;<@8!
Y<23F@J2F;6!_86!Y:F4M6!_8!I86!>!U<0"F6!c8!B%---C8!=2"FJ!1"e;/!?2G"J!12/;?@!2L!@J0J;/!0F/!
<;#;0?;/!V<;L;<;F4;@!L2<!0?J;<F0J"#;PL:;?!#;M"4?;@8!N$&%1O"$/&/,"%(I.1.&$0;(9&$/(?5(
GL6!&$(P&&+8!
Y<:F@3"^6!O8!B$,',C8!F:1/.-&/,0(&%2(I.O$.1.%/&/,+.(A.1,4%(")(91:0;"'"4,0&'(*VO.$,-.%/18!
Y;<^;?;S6!D.T!`F"#;<@"JS!2L!D0?"L2<F"0!K<;@@8!
D2M;F6!K86!>!D2M;F6!=8!B$,,)C8!Q,).(+&'#.1(&%2(&2"'.10.%/(-.%/&'(;.&'/;8!A0M30S6!Q=T!
O<?E0:18!
D2<<0?PN;</:G26!N8!B$,,*C8!_:0?!h<;0?"J";@h!2L!42F@;<#0J"2F!E;M0#"2<T!I;?LP<;V2<J!#@8!
2E@;<#0J"2F@!2L!<;P:@;!0F/!<;4S4?"FG!E;M0#"2<8!!"#$%&'(")(*%+,$"%-.%/&'(91:0;"'"4:5(
<M6!$&(P$'(8!
D2J;6!I86!>!A2@^23"J56!_8!I8!B$,,+C8!dF!JM;!/SF01"4!42#0<"0J"2F!E;J3;;F!"FJ;<V;<@2F0?!
E;M0#"2<!0F/!0LL;4JT!K<;/"4J"2F!L<21!F;:<2J"4"@16!;eJ<2#;<@"2F6!0F/!0G<;;0E?;F;@@8!
!"#$%&'(")(9.$1"%&',/:(&%2(F"0,&'(91:0;"'"4:5(MB6!$-&%P$-')8!
D<23F;6!_8!K86!>!A0<?23;6!_8!B$,)-C8!.!F;3!@40?;!2L!@24"0?!/;@"<0E"?"JS!"F/;V;F/;FJ!2L!
V@S4M2V0JM2?2GS8!!"#$%&'(")(8"%1#'/,%4(91:0;"'"4:5(6L6!&',P&('8!
D<23F;6!_8!K86!>!A0<?23;6!_8!B$,)'C8!N;.(&OO$"+&'(-"/,+.8!Q;3!R2<^T!7"?;S8!
D@"^@5;FJ1"M0?S"6!A8!B$,++C8!UM;!L?23!;eV;<";F4;!0F/!"J@!@"GF"L"40F4;!L2<!M:10F!
V@S4M2?2GS8!XF!X8!I8!D@"^@5;FJ1"M0?S"!>!A8!D@"^@5;FJ1"M0?S"!BO/@8C6!=O/,-&'(
*VO.$,.%0.W(91:0;"'"4,0&'(F/#2,.1(")(X'"J(,%(8"%10,"#1%.118!Q;3!R2<^T!D01E<"/G;!
`F"#;<@"JS!K<;@@8!
D@"^@5;FJ1"M0?S"6!A8!B%---C8!?.:"%2(P"$.2"-(&%2(&%V,./:!B%F/!;/8C8!I0F!Z<0F4"@42T!=2@@;SP
Y0@@8!
D@"^@5;FJ1"M0?S"6!A86!>!\<0;L6!H8!B$,+-C8!UM;!;eV;<";F4;!2L!L<;;/21!"F!;#;<S/0S!?"L;8!
E-.$,0&%(!"#$%&'(")(8"--#%,/:(91:0;"'"4:5(<K6!'-%P'$'8!

$&&!
!

!

D@"^@5;FJ1"M0?S"6!A86!>!90<@2F6!H8!78!B$,+'C8!?.,%4(&2"'.10.%/W(8"%)',0/(&%2(4$"J/;(,%(/;.(
/..%&4.(:.&$18!Q;3!R2<^T!Y0@"4!Y22^@8!
D@"^@5;FJ1"M0?S"6!A86!>!90<@2F6!H8!78!B$,+*C8!N0?"/"JS!0F/!<;?"0E"?"JS!2L!JM;!;eV;<";F4;!
@01V?"FG!1;JM2/8!!"#$%&'(")(Y.$+"#1(&%2(3.%/&'(A,1.&1.5(<MBB,C6!(%)P(&)8!
D@"^@5;FJ1"M0?S"6!A86!>!I4MF;"/;<6!Y8!B%---C8!?.0"-,%4(&2#'/W(@"J(/..%&4.$1(O$.O&$.()"$(/;.(
J"$'2(")(J"$R8!Q;3!R2<^T!Y0@"4!Y22^@8!
_0?S6![8!O8!B$,*&C8!N"J&$2(&(1/.&2:Z1/&/.(.0"%"-:8!I0F!Z<0F4"@42T!78![8!Z<;;10F8!
_;!R2:FG6!H8!B$,++P+,C8!OeV?2<"FG!JM;!/"LL;<;F4;!E;J3;;F!<;4S4?;<@!0F/!F2FP<;4S4?;<@T!UM;!
<2?;!2L!"FL2<10J"2F8!!"#$%&'(")(*%+,$"%-.%/&'(F:1/.-15(<K6!&'$P&($8!
_;!R2:FG6!H8!B$,,)C8!I21;!V@S4M2?2G"40?!0@V;4J@!2L!<;/:4;/!42F@:1VJ"2F!E;M0#"2<T!UM;!
<2?;!2L!"FJ<"F@"4!@0J"@L04J"2F!0F/!421V;J;F4;!12J"#0J"2F8!*%+,$"%-.%/(&%2(?.;&+,"$5(
6K6!&(+P'-,8!
_;!R2:FG6!H8!B%---C8!OeV0F/"FG!0F/!O#0?:0J"FG!A2J"#;@!L2<!OF#"<2F1;FJ0??S!H;@V2F@"E?;!
Y;M0#"2<8!!"#$%&'(")(F"0,&'(S11#.15(B7B&C6!(-,P(%)8!
_;4M;@F;6!A86!KS@5;5SF@^"6!U86!.<F/J6!=86!H0F@216!I86!IM;?/2F6!c8!A86!>!#0F!cF"VV;FE;<G6!.8!
B%--&C8!9"J;<0?!0F/!@S1E2?"4!"112<J0?"JST!UM;!;LL;4J!2L!;#"/;F4;!2L!?"J;<0?!
"112<J0?"JS!2F!@;?LP;@J;;1!@J<"#"FG!"F!<;@V2F@;!J2!12<J0?"JS!@0?";F4;8!!"#$%&'(")(
9.$1"%&',/:(&%2(F"0,&'(91:0;"'"4:5(KL6!*%%P*&*8!
_;??;!Z0#;6!.86!>!A0@@"1"F"6!Z8!B%--&C8!A0^"FG!/"@0E"?"JS!"FJ2!0!<;@2:<4;T!UM;!<2?;!2L!2VJ"10?!
;eV;<";F4;8!N;.(91:0;"'"4,1/5(<76!,P$-8!
_"4^"F@2F6!U8!986!>!i;??"FG;<6!K8!A8!B$,+-C8!.!421V0<"@2F!2L!JM;!E;M0#"2<0??S!0F4M2<;/!<0J"FG!
1"e;/!@J0F/0</!@40?;!L2<10J@8!!"#$%&'(")(EOO',.2(91:0;"'"4:5(7B6!$'*P$('8!
_";F;<6!O86!>!Z:W"J06!Z8!B$,,(C8!H;@2:<4;@6!V;<@2F0?!@J<"#"FG@6!0F/!@:EW;4J"#;!3;??PE;"FGT!.!
F212JM;J"4!0F/!"/"2G<0VM"4!0VV<204M8!!"#$%&'(")(9.$1"%&',/:(&%2(F"0,&'(91:0;"'"4:5(
7K6!,%)P,&(8!
_";J56!U86!IJ;<F6!K8!D86!>!\:0GF0F26!\8!.8!B$,,+C8!I24"0?!@J<:4J:<0?!0F/!@24"0?!V@S4M2?2G"40?!
E0@;@!2L!;F#"<2F1;FJ0?!42F4;<F8!*%+,$"%-.%/(&%2(?.;&+,"$5(GCB'C6!'(-P'*%8!
$&'!
!

!

_:F?0V6!H8!O86!>!9";<;6!c8!_8!N8!B$,**C8!90F/!OJM"4!2<!\2?/;F!H:?;T!D211;FJ!2F!h90F/!OJM"4!
H;0?"5;/h!ES!UM210@!.8![;E;<?;"F6!=IX6!%+B'C6!$,*%8!!"#$%&'(")(F"0,&'(S11#.15(GGB&C6!
%--P%$$8!
_:F?0V6!H8!O86!>!9";<;6!c8!_8!N8!B$,*+C8!UM;!hF;3!;F#"<2F1;FJ0?!V0<0/"G1hT!.!V<2V2@;/!
1;0@:<"FG!"F@J<:1;FJ!0F/!V<;?"1"F0<S!<;@:?J@8!!"#$%&'(")(*%+,$"%-.%/&'(*2#0&/,"%5(
[6!$-P$,8!
_:F?0V6!H8!O86!9";<;6!c8!_8!N86!A;<J"G6!.8!\86!>!=2F;@6!H8!O8!B%---C8!A;0@:<"FG!OF/2<@;1;FJ!2L!
JM;!Q;3!OF#"<2F1;FJ0?!K0<0/"G1T!.!H;#"@;/!QOK!I40?;8!!"#$%&'(")(F"0,&'(S11#.15(
B7B&C6!'%(P''%8!
_3S;<6!78!d86!9;;1"FG6!Z8!D86!D2E;<F6!A8!c86!K2<J;<6!Y8!O86!>!=04^@2F6!=8!A8!B$,,&C8!D<"J"40?!
<;#";3!2L!E;M0#"2<0?!"FJ;<#;FJ"2F@!J2!V<;@;<#;!JM;!;F#"<2F1;FJT!H;@;0<4M!@"F4;!
$,+-8!*%+,$"%-.%/(&%2(?.;&+,"$5(6B6!%*(P&%$8!
O/30</@6!.8!98!B$,(*C8!N;.(1"0,&'(2.1,$&P,',/:(+&$,&P'.(,%(O.$1"%&',/:(&11.11-.%/(&%2($.1.&$0;8!
Q;3!R2<^T!_<S/;F8!
O/30</@6!.8!98!B$,*-C8!N;.(-.&1#$.-.%/(")(O.$1"%&',/:(/$&,/1(P:(10&'.1(&%2(,%+.%/"$,.18!Q;3!
R2<^T![2?JPH"F;M0<JP7"F@J2F8!
O"GF;<6!I8!B%--$C8!UM;!<;?0J"2F@M"V!E;J3;;F!hV<2J;4J"FG!JM;!;F#"<2F1;FJh!0@!0!/21"F0FJ!
?"L;!G20?!0F/!@:EW;4J"#;!3;??PE;"FG8!XF!K8!I4M1:4^!>!c8!A8!IM;?/2F!BO/@8C6!Q,).(U"&'1(
&%2(\.''Z?.,%4W(N"J&$21(&(9"1,/,+.(91:0;"'"4:(")(@#-&%(F/$,+,%4!BVV8!$+%P%-$C8!
I;0JJ?;6!7.T![2G<;L;!0F/![:E;<8!
OS@;F4^6![8!=86!>!OS@;F4^6!I8!Y8!\8!B$,)'C8!N;.(-&%#&'(")(/;.(*:1.%0R(9.$1"%&',/:(,%+.%/"$:8!
92F/2FT!`!2L!92F/2F!K<;@@8!
Z;?/10F!Y0<<;JJ6!98!B%--'C8!Z;;?"FG@!2<!72</@g!`F/;<@J0F/"FG!JM;!42FJ;FJ!"F!@;?LP<;V2<J!
<0J"FG@!2L!;12J"2F0?!;eV;<";F4;8!!"#$%&'(")(9.$1"%&',/:(&%2(F"0,&'(91:0;"'"4:5(KMB%C6!
%))P%+$8!
Z;F"4M;?6!d8!B$,'(C8!N;.(91:0;"&%&':/,0(N;."$:(")(Y.#$"1,18!Q;3!R2<^T!78!78!Q2<J2F8!
Z;<<;<P"PD0<E2F;??6!.86!>!\23/S6!=8!A8!B%--*C8!OF#"<2F1;FJ0?!_;G<0/0J"2F!0F/![0VV"F;@@8!
*0"'"4,0&'(*0"%"-,015(7C6!(-,P($)8!

$&(!
!

!

Z"@ME;"F6!A8!B$,)*C8!.JJ"J:/;!0F/!JM;!V<;/"4J"2F!2L!E;M0#"2<8!XF!A8!Z"@ME;"F!BO/8C6!I.&2,%41(
,%(&//,/#2.5(/;."$:(&%2(-.&1#$.-.%/8!Q;3!R2<^T!7"?;S8!
Z"@ME;"F6!A86!>!.W5;F6!X8!B$,*(C8!?.',.)5(E//,/#2.5(S%/.%/,"%5(&%2(?.;&+,"$W(&%(,%/$"2#0/,"%(/"(
/;."$:(&%2($.1.&$0;8!H;0/"FG6!A.T!.//"@2FP7;@?;S8!
Z<0E?;6!_8!O8!I86!K?0JJ6!986!>![2;S6!I8!B$,,+C8!D2F4;0?0E?;!@J"G10@!0F/!V2@"J"#;!@;?LP
V;<4;VJ"2F@T!Z;;?"FG!E;JJ;<!0<2:F/!@"1"?0<!2JM;<@8!!"#$%&'(")(9.$1"%&',/:(&%2(F"0,&'(
91:0;"'"4:5(ML6!,-,P,%%8!
Z:FJ23"456!I86!>!H0#;J56!=8!B%--&C8!9"1/ZY"$-&'(F0,.%0.8!K0V;<!V<;@;FJ;/!0J!JM;!XFJ;<F0J"2F0?!
I24";JS!L2<!O42?2G"40?!O42F21"4@8!!
\0J;<@?;E;F6!Y86!A;0/23@6!=86!.E<0M01@;6!786!>!=04^@2F6!U8!B%--+C8!A0J;<"0?"@J"4!0F/!
;F#"<2F1;FJ0?!#0?:;@!2L!S2:FG!V;2V?;8!`FV:E?"@M;/!`FV:E?"@M;/!10F:@4<"VJ8!
`F"#;<@"JS!2L!I:<<;S8!
\;??;<6!O8!I8!B$,,(C8!.4J"#;?S!D0<"FG!L2<!JM;!OF#"<2F1;FJT!.F!XFJ;G<0J"2F!2L!Y;M0#"2<"@1!0F/!
[:10F"@18!*%+,$"%-.%/(&%2(?.;&+,"$5(6MB%C6!$+'P$,(8!
\F;;5S6!`86!>!H:@J"4M"F"6!.8!B%---C8!.!Z"F;!"@!0!K<"4;8!N;.(!"#$%&'(")(Q.4&'(F/#2,.15(
]]S]B=0F:0<SC6!$P$*8!
\2LL10F6!O8!B$,(,C8!N;.(O$.1.%/&/,"%(")(1.')(,%(.+.$:2&:(',).8!\0</;F!D"JS6!QRT!
_2:E?;/0Sj.F4M2<8!
\22/6!=8!B%--*C8!IM2V!]J"?!3;!/<2Vg!U;?;#"@"2F6!10J;<"0?"@1!0F/!0JJ"J:/;@!0E2:J!JM;!F0J:<0?!
;F#"<2F1;FJ8!3&11(8"--#%,0&/,"%(&%2(F"0,./:5(<C6!&)(P&+&8!
\<2:5;J6!Z8!A8!O86!c0@@;<6!U86!.M:#"06!.86!Z;F0F/;5P_2?@6!=8!A86!c"16!R86!>!90:6!I8!B%--(C8!UM;!
@J<:4J:<;!2L!G20?!42FJ;FJ@!04<2@@!$(!4:?J:<;@8!!"#$%&'(")(9.$1"%&',/:(&%2(F"0,&'(
91:0;"'"4:5(K[6!+--P+$)8!
\<2#;@6!H8!A8!B$,+*C8!H;@;0<4M!2F!I:<#;S!_0J0!k:0?"JS8!9#P',0(=O,%,"%(H#&$/.$':5(B<6!I$()P
I$*%8!

$&)!
!

!

\:0GF0F26!\8!.86!IJ;<F6!K8!.86!>!_";J56!U8!B$,,(C8!XFL?:;F4;@!2F!.JJ"J:/;PY;M0#"2<!
H;?0J"2F@M"V@T!.!Q0J:<0?!OeV;<"1;FJ!3"JM!D:<E@"/;!H;4S4?"FG8!*%+,$"%-.%/(&%2(
?.;&+,"$5(6MB(C6!),,P*$+8!
[0<J"G6!U86!c0"@;<6!Z8!\86!>!Y23?;<6!K8!.8!B%--$C8!K@S4M2?2G"40?!<;@J2<0J"2F!"F!F0J:<;!0@!0!
V2@"J"#;!12J"#0J"2F!L2<!;42?2G"40?!E;M0#"2<8!*%+,$"%-.%/(&%2(?.;&+,"$5(GG6!(,-P)-*8!
[;E;<?;"F6!U8!.8!B$,*%C8!UM;!90F/!OJM"4!H;0?"5;/T!I21;!I24"0?!K@S4M2?2G"40?!OeV?0F0J"2F@!
L2<!DM0FG"FG!OF#"<2F1;FJ0?!.JJ"J:/;@8!!"#$%&'(")(F"0,&'(S11#.15(6KB'C6!*,P+*8!
[;E;<?;"F6!U8!.8!B$,+$C8!OF#"<2F1;FJ0?!.JJ"J:/;@8!^.,/10;$,)/()#$(D-J.'/O"',/,R5(66!%'$P%*-8!
[;^JF;<6!=8!A86!I4M1"/J6!=8!.86!>!D@"^@5;FJ1"M0?S"6!A8!B%--*C8!*VO.$,.%0.(F&-O',%4(3./;"2W(
3.&1#$,%4(/;.(H#&',/:(")(*+.$:2&:(QS).8!UM2:@0F/!d0^@T!I0G;!K:E?"40J"2F@8!
["4^@6!=86!H010F0JM0F6!Q86!c"16!_86!A2F"E"6!A86!I;?@^S6!=86![0F@;F6!A86!;J!0?8!B%-$-C8!
.F/7;??F;@@T!.F!dV;F!A2E"?;!IS@J;1!L2<!.4J"#"JS!0F/!OeV;<";F4;!I01V?"FG8!
\,$.'.11(@.&'/;5(=0/"P.$(BZM6!$P$-8!
["F;@6!=8!A86![:FG;<L2</6![8!H86!>!U21;<06!.8!Q8!B$,+*C8!.F0?S@"@!0F/!@SFJM;@"@!2L!<;@;0<4M!2F!
<;@V2F@"E?;!;F#"<2F1;FJ0?!E;M0#"2<T!.!1;J0P0F0?S@"@8!!"#$%&'(")(*%+,$"%-.%/&'(
*2#0&/,"%5(<KB%C6!$P+8!
[F0J":^6!I8![8!B$,,$C8!OeV;<";F4;!@01V?"FG!3"JM!;?/;<?S!V;<@2F@T!.F!;eV?2<0J"2F!2L!JM;!
1;JM2/8!S%/.$%&/,"%&'(!"#$%&'(")(E4,%4(&%2(@#-&%(A.+.'"O-.%/5(GG6!'(P)'8!
[2VV;<6!=8!H8!B$,,$C8!H;4S4?"FG!0@!.?J<:"@J"4!Y;M0#"2<T!Q2<10J"#;!0F/!Y;M0#"2<0?!IJ<0J;G";@!
J2!OeV0F/!K0<J"4"V0J"2F!"F!0!D211:F"JS!H;4S4?"FG!K<2G<018!*%+,$"%-.%/(&%2(
?.;&+,"$5(6GB%C6!$,(P%%-8!
[2<1:JM6!I8!O8!B$,+)C8!UM;!@01V?"FG!2L!;eV;<";F4;@!"F!@"J:8!!"#$%&'(")(9.$1"%&',/:5(BL6!%)%P
%,&8!
[:LL2</6!A8!H86!IM";?/@6!.8!986!IM"LL10F6!I8!I86!K0JS6!=86!>!Y0?0E0F"@6!A8!B%--%C8!H;04J"#"JS!J2!
;42?2G"40?!121;FJ0<S!0@@;@@1;FJT!.F!;e01V?;!:@"FG!:F/;<G<0/:0J;!V<2E?;1!
/<"F^;<@8!91:0;"'"4:(")(E22,0/,+.(?.;&+,"$15(<76!%-(P%$$8!

$&*!
!

!

[:<?E:<J6!H8!U8!B$,,*C8!H0F/21?S!@01V?"FG!JM"F^"FG!"F!JM;!F0J:<0?!;F#"<2F1;FJ8!!"#$%&'(")(
8"%1#'/,%4(&%2(8',%,0&'(91:0;"'"4:5(7B6!,'$P,',8!
[:@@;<?6!O8!B$,)%C8!S2.&1W(U.%.$&'(S%/$"2#0/,"%(/"(9#$.(9;.%"-.%"'"4:!B78!H8!Y8!\"E@2F6!
U<0F@8C8!92F/2F6!Q;3!R2<^T!A041"??0F8!
=;F4^@6!D8!B$,,-C8!N0<";J";@!2L!0?J<:"@18!XF!=8!=8!A0F@E<"/G;!BO/8C6!?.:"%2(1.')Z,%/.$.1/!BVV8!(&P
)*C8!DM"40G2T!`F"#;<@"JS!2L!DM"40G2!K<;@@8!
=2MF@2F6!A86![0@MJ<2:/"6!I86!>!9"F/@0S6!_8!I8!B$,,&C8!I2:<4;!A2F"J2<"FG8!91:0;"'"4,0&'(
?#''./,%5(<<LB$C6!&P%+8!
c0MF;10F6!_86!>!c<:;G;<6!.8!Y8!B%--)C8!_;#;?2V1;FJ@!"F!JM;!A;0@:<;1;FJ!2L!I:EW;4J"#;!
7;??PY;"FG8!!"#$%&'(")(*0"%"-,0(9.$1O.0/,+.15(6CB$C6!&P%'8!
c0MF;10F6!_86!c<:;G;<6!.8!Y86!I4M^0/;6!_8!.86!I4M30<56!Q86!>!IJ2F;6!.8!.8!B%--'C8!.!@:<#;S!
1;JM2/!L2<!4M0<04J;<"5"FG!/0"?S!?"L;!;eV;<";F4;T!UM;!_0S!H;42F@J<:4J"2F!A;JM2/!
B_HAC8!F0,.%0.(GC7B(*-%C6!$**)P$*+-8!
c0"@;<6!Z8!\8!B$,,+C8!.!\;F;<0?!A;0@:<;!2L!O42?2G"40?!Y;M0#"2<8!!"#$%&'(")(EOO',.2(F"0,&'(
91:0;"'"4:5(6KB(C6!&,(P'%%8!
c0"@;<6!Z8!\86!H0FF;S6!A86![0<J"G6!U86!>!Y23?;<6!K8!.8!B$,,,C8!O42?2G"40?!E;M0#"2<6!
;F#"<2F1;FJ0?!0JJ"J:/;6!0F/!L;;?"FG@!2L!<;@V2F@"E"?"JS!L2<!JM;!;F#"<2F1;FJ8!*#$"O.&%(
91:0;"'"4,1/5(LB%C6!(,P*'8!
c0?@6!O86!I4M:104M;<6!_86!>!A2FJ0/06!98!B$,,,C8!O12J"2F0?!0LL"F"JS!J230</!F0J:<;!0@!0!
12J"#0J"2F0?!E0@"@!J2!V<2J;4J!F0J:<;8!*%+,$"%-.%/(&%2(?.;&+,"$5(G<6!$*+P%-%8!
c0V?0F6!I8!B%---C8![:10F!Q0J:<;!0F/!OF#"<2F1;FJ0??S!H;@V2F@"E?;!Y;M0#"2<8!!"#$%&'(")(
F"0,&'(S11#.15(B7B&C6!',$P(-+8!
c0V?0F6!I86!>!c0V?0F6!H8!B$,+,C8!UM;!#"@:0?!;F#"<2F1;FJT!K:E?"4!V0<J"4"V0J"2F!"F!/;@"GF!0F/!
V?0FF"FG8!!"#$%&'(")(F"0,&'(S11#.15(LB6!(,P+)8!
c0<0V0F2@6!O86!i"11;<10F6!=86!Z2<?"55"6!=86!>!A0<J;F@6!=8PY8!B%--,C8!D1.$(*VO.$,.%0.(=+.$(N,-.W(
E%(S%,/,&'(X$&-.J"$R8!K0V;<!V<;@;FJ;/!0J!JM;!D[X!%--,8!!

$&+!
!

!

c0@@;<6!U8!B%--%C8!N;.(;,4;(O$,0.(")(-&/.$,&',1-8!D01E<"/G;6!A.T!AXU!K<;@@8!
c0@@;<6!U8!B%--(C8!Z<:G0?"JS6!G;F;<2@"JS6!0F/!10J;<"0?"@1!"F!4M"?/<;F!0F/!0/2?;@4;FJ@8!XF!c8!
.8!A22<;!>!98![8!9"VV10F!BO/@8C6!\;&/(2"(0;,'2$.%(%..2(/"()'"#$,1;_(8"%0.O/#&',>,%4(
&%2(-.&1#$,%4(,%2,0&/"$1(")(O"1,/,+.(2.+.'"O-.%/!BVV8!&(*P&*&C8!Q;3!R2<^T!IV<"FG;<!
I4";F4;8!
c0@@;<6!U8!B%-$-C8!O42?2G"40?!DM0??;FG;@6!A0J;<"0?"@J"4!N0?:;@6!0F/!I24"0?!DM0FG;8!XF!H8!
Y"@30@P_";F;<!BO/8C6!9"1,/,+.(91:0;"'"4:(&1(F"0,&'(8;&%4.8!Q;3!R2<^T!IV<"FG;<8!
c0@@;<6!U8!B%-$$C8!@#-&%(S2.%/,/:(&%2(*%+,$"%-.%/&'(8;&''.%4.18!!
c0@@;<6!U8!B"F!V<;@@C8!D:?J:<0?!#0?:;@!0F/!JM;!3;??PE;"FG!2L!L:J:<;!G;F;<0J"2F@T!.!4<2@@P
F0J"2F0?!@J:/S8!!"#$%&'(")(8$"11Z8#'/#$&'(91:0;"'"4:8!
c0@@;<6!U86!>!HS0F6!H8!A8!B$,,&C8!.!/0<^!@"/;!2L!JM;!.1;<"40F!/<;01T!D2<<;?0J;@!2L!L"F0F4"0?!
@:44;@@!0@!0!4;FJ<0?!?"L;!0@V"<0J"2F8!!"#$%&'(")(9.$1"%&',/:(&%2(F"0,&'(91:0;"'"4:5(7B6!
'$-P'%%8!
c0@@;<6!U86!>!HS0F6!H8!A8!B$,,)C8!Z:<JM;<!;e01"F"FG!JM;!.1;<"40F!_<;01T!_"LL;<;FJ"0?!
42<<;?0J;@!2L!"FJ<"F@"4!0F/!;eJ<"F@"4!G20?@8!9.$1"%&',/:(&%2(F"0,&'(91:0;"'"4:(?#''./,%5(
666!%+-P%+*8!
c0@@;<6!U86!HS0F6!H8!A86!D2:4M10F6!D8!O86!>!IM;?/2F6!c8!A8!B%--'C8!A0J;<"0?"@J"4!#0?:;@T!UM;"<!
40:@;@!0F/!42F@;f:;F4;@8!XF!U8!c0@@;<!>!_8!c0FF;<!BO/@8C6!91:0;"'"4:(&%2(0"%1#-.$(
0#'/#$.W(N;.(1/$#44'.()"$(&(4""2(',).(,%(&(-&/.$,&',1/,0(J"$'2!BVV8!$$P%+C8!70@M"FGJ2F6!
_DT!.1;<"40F!K@S4M2?2G"40?!.@@24"0J"2F8!
c0@@;<6!U86!HS0F6!H8!A86!i0e6!A86!>!I01;<2LL6!.8!=8!B$,,(C8!UM;!<;?0J"2F@!2L!10J;<F0?!0F/!@24"0?!
;F#"<2F1;FJ@!J2!?0J;!0/2?;@4;FJ@]!10J;<"0?"@J"4!0F/!V<2@24"0?!#0?:;@8!A.+.'"O-.%/&'(
91:0;"'"4:5(G<6!,-*P,$'8!
c0@@;<6!U86!>!IM;?/2F6!c8!A8!B%---C8!dL!3;0?JM!0F/!/;0JMT!A0J;<"0?"@16!12<J0?"JS!@0?";F4;6!
0F/!42F@:1VJ"2F!E;M0#"2<8!91:0;"'"4,0&'(F0,.%0.5(<<6!&(%P&((8!
c0J5;#6!H8!_86!>!=2MF@2F6!U8!H8!B$,+*C8!9$"-"/,%4(.%.$4:(0"%1.$+&/,"%W(E%(&%&':1,1(")(
P.;&+,"$&'($.1.&$0;8!Y2:?/;<6!DdT!7;@J#";3!K<;@@8!

$&,!
!

!

c"??"FG@32<JM6!A8!.86!>!\"?E;<J6!_8!U8!B%-$-C8!.!70F/;<"FG!A"F/!X@!0F!`FM0VVS!A"F/8!F0,.%0.5(
GGC6!,&%8!
c?"FG;<6!O8!B$,*+P$,*,C8!_"1;F@"2F@!2L!JM2:GMJ!0F/!"10G;<S!"F!F2<10?!30^"FG!@J0J;@8!
!"#$%&'(")(E'/.$.2(F/&/.1(")(8"%10,"#1%.115(L6!,*P$$&8!
c<2@F"4^6!=8!.8!B$,,,C8!I:<#;S!H;@;0<4M8!E%%#&'(I.+,.J(")(91:0;"'"4:5(BC6!(&*P()*8!
c<2@F"4^6!=8!.86!>!Y;<;FJ6!A8!c8!B$,,&C8!D21V0<"@2F@!2L!V0<JS!"/;FJ"L"40J"2F!0F/!V2?"4S!
V<;L;<;F4;@T!JM;!"1V04J!2L!@:<#;S!f:;@J"2F!L2<10J8!E-.$,0&%(!"#$%&'(")(9"',/,0&'(
F0,.%0.5(GM6!,'$P,)'8!
c:E;S6!H86!90<@2F6!H8!786!>!D@"^@5;FJ1"M0?S"6!A8!B$,,)C8!OeV;<";F4;!@01V?"FG!1;JM2/!
0VV?"40J"2F@!J2!4211:F"40J"2F!<;@;0<4M!f:;@J"2F@8!!"#$%&'(")(8"--#%,0&/,"%5(L76!
,,P$%-8!
901";??6!=8!U8!B$,,*C8!XF/"#"/:0?@!0F/!JM;!/"LL;<;F4;@!E;J3;;F!JM;18!XF!H8![2G0F6!=8!.8!
=2MF@2F!>!I8!H8!Y<"GG@!BO/@8C6!@&%2P""R(")(9.$1"%&',/:(91:0;"'"4:!BVV8!$$*P$'$C8!I0F!
_";G26!D.T!.40/;1"4!K<;@@8!
90<@2F6!H8!786!A2F;J06!\86!H"4M0</@6!A8![86!>!7"?@2F6!I8!B%--%C8!D2FJ"F:"JS6!@J0E"?"JS6!0F/!
4M0FG;!"F!/0"?S!;12J"2F0?!;eV;<";F4;!04<2@@!0/2?;@4;F4;8!8;,'2(A.+.'"O-.%/5(MG6!
$$($P$$)(8!
90<@2F6!H8!786!>!H"4M0</@6!A8![8!B$,,'C8!A,+.$4.%/($.&',/,.1W(N;.(.-"/,"%&'(',+.1(")(-"/;.$15(
)&/;.$15(&%2(&2"'.10.%/18!Q;3!R2<^T!Y0@"4!Y22^@8!
903J2F6!A8!K8!B$,,,C8!A;JM2/@!0F/!42F4;VJ@!L2<!J"1;PE:/G;J!<;@;0<4M!2F!;?/;<@8!XF!78!O8!
K;FJ?0F/6!.8!I8![0<#;S6!A8!K8!903J2F!>!A8!.8!A4D2??!BO/@8C6!N,-.(#1.($.1.&$0;(,%(/;.(
1"0,&'(10,.%0.1!BVV8!$-*P$%(C8!Q;3!R2<^T!c?:3;<8!
9;;6!R8!I8!B%--(C8!A;0@:<"FG!JM;!G;F/;<!G0V!"F!M2:@;M2?/!?0E2<T!.44:<0J;?S!;@J"10J"FG!
3"#;@]!0F/!M:@E0F/@]!42FJ<"E:J"2F@8!XF!Y8!I4MF;"/;<!>!98!=8!70"J;!BO/@8C6!?.,%4(
/"4./;.$5(J"$R,%4(&O&$/W(A#&'Z0&$..$()&-,',.1(&%2(/;.(J"$RZ',).(P&'&%0.!BVV8!%%,P
%'*C8!Q;3!R2<^T!D01E<"/G;!`F"#;<@"JS!K<;@@8!
9;M10FF6!=8!B$,,,C8!?.)#%2.(.-O,$,10;.$(X"$10;#%4(>#(D-J.'/P,'2#%4(&%2(
D-J.'/P.J#11/1.,%T!dV?0/;F6!9;@^;!:F/!Y:<"4M8!

$'-!
!

!

9;2V2?/6!.8!B$,'+C8!E(F&%2(8"#%/:(E'-&%&08!Q;3!R2<^T!deL2</!`F"#;<@"JS!K<;@@8!
9;#"F6!\8!B$,,&6!.V<"?!$%C8!U22!G<;;F!L2<!JM;"<!23F!G22/8!E2+.$/,1,%4(E4.5(7LB%,C8!
9SFF6!K8!B%--$C8!UM;!"1V04J!2L!"F4;FJ"#;@!2F!<;@V2F@;!<0J;@!J2!V;<@2F0?!"FJ;<#";3!@:<#;S@T!
H2?;!0F/!V;<4;VJ"2F!2L!"FJ;<#";3;<@8!S%/.$%&/,"%&'(!"#$%&'(")(9#P',0(=O,%,"%(
I.1.&$0;5(<G6!&%)P&&)8!
A0FF;??6!H8!D86!i:50F;^6!=86!>!.<2F@2F6!H8!B%--(C8!S%/.$%./`8"-O#/.$(D1.(&%2(E2"'.10.%/(
Q.,1#$.(?.;&+,"$5(X'"J(*VO.$,.%0.1(&%2(91:0;"'"4,0&'(\.''Z?.,%4W(N;.(A,1O'&0.-.%/(
@:O"/;.1,18!K0V;<!V<;@;FJ;/!0J!JM;!O?;#;FJM!D0F0/"0F!D2FG<;@@!2F!9;"@:<;!
H;@;0<4M8!!
A0S;<6!Z8!I86!>!Z<0FJ56!D8!A8!B%--'C8!UM;!42FF;4J;/F;@@!J2!F0J:<;!@40?;T!0!1;0@:<;!2L!
"F/"#"/:0?@!L;;?"FG!"F!4211:F"JS!3"JM!F0J:<;8!!"#$%&'(")(*%+,$"%-.%/&'(91:0;"'"4:5(
6L6!(-&P($(8!
A4./01@6!_8!K8!B$,,(C8!7M0J!/2!3;!^F23!3M;F!3;!^F23!0!V;<@2Fg!!"#$%&'(")(9.$1"%&',/:5(
7G6!&)(P&,)8!
A4D0F4;6!.8!I8!B%-$-C8!*-"/,"%&'(Q&P"$(,%(S%/.$0#'/#$&'(F.$+,0.(*%0"#%/.$1W(E%(*VO.$,.%0.(
F&-O',%4(F/#2:a!`F"#;<@"JS!2L!X??"F2"@6!`<E0F0PDM01V0"GF8!
A;0/23@6!_8![86!A;0/23@6!_8!986!H0F/;<@6!=86!>!XXX6!78!78!Y8!B$,*%C8!N;.(Q,-,/1(/"(U$"J/;8!
Q;3!R2<^T!`F"#;<@;!Y22^@8!
A"?L2FJ6!U8!98!B%--,C8!UM;!;LL;4J@!2L!@24"0?!/;@"<0E"?"JS!2F!@;?LP<;V2<J;/!;F#"<2F1;FJ0?!
0JJ"J:/;@!0F/!;42?2G"40?!E;M0#"2<8!*%+,$"%-.%/&',1/5(6[6!%)&P%),8!
A"?L2FJ6!U8!986!>!_:4^"JJ6!=8!B%--'C8!UM;!@J<:4J:<;!2L!;F#"<2F1;FJ0?!0JJ"J:/;@T!0!L"<@JP!0F/!
@;42F/P2</;<!42FL"<10J2<S!L04J2<!0F0?S@"@8!!"#$%&'(")(*%+,$"%-.%/&'(91:0;"'"4:5(6L6!
%+,P&-&8!
A:??"G0F6!D8!Y86!I4MF;"/;<6!Y86!>!72?L;6!H8!B%---C8!U"1;!:@;!0F/!V2V:?0J"2F!<;V<;@;FJ0J"2F!"F!
JM;!I?20F!IJ:/S!2L!./2?;@4;FJ@8!.?L<;/!K8!I?20Fj`F"#;<@"JS!2L!DM"40G2!D;FJ;<!L2<!JM;!
IJ:/S!2L!72<^"FG!Z01"?";@8!

$'$!
!

!

Q;J;<6!=86!>!70^@E;<G6!=8!B$,)'C8!.!@J:/S!2L!<;@V2F@;!;<<2<@!"F!;eV;F/"J:<;!/0J0!L<21!
M2:@;M2?/!"FJ;<#";3@8!!"#$%&'(")(/;.(E-.$,0&%(F/&/,1/,0&'(E11"0,&/,"%5(B[B&-(C6!$+P((8!
Q23?"@6!_8!K86!>!D2M;F6!.8!R8!B$,)+C8!A22/P<;V2<J@!0F/!JM;!42??;G;!F0J:<0?!@;JJ"FGT!.!/0S!"F!
JM;!?"#;@!2L!JM<;;!<22110J;@!:F/;<!040/;1"4!V<;@@:<;8!91:0;"'"4,0&'(I.O"$/15(6G6!
(($P())8!
d"@M"6!I86!_";F;<6!O86!I42??2F6!D8!Q86!>!Y"@30@P_";F;<6!H8!B%--'C8!D<2@@PI"J:0J"2F0?!D2F@"@J;F4S!
2L!.LL;4J"#;!OeV;<";F4;@!.4<2@@!D:?J:<;@8!!"#$%&'(")(9.$1"%&',/:(&%2(F"0,&'(
91:0;"'"4:5(K7B&C6!')-P'*%8!
d12J26!.8!A86!>!IFS/;<6!A8!B$,,(C8!I:@J0"F;/!M;?V"FG!3"JM2:J!2E?"G0J"2FT!A2J"#0J"2F6!
?2FG;#"JS!2L!@;<#"4;6!0F/!V;<4;"#;/!0JJ"J:/;!4M0FG;!012FG!.X_I!#2?:FJ;;<@8!!"#$%&'(
")(9.$1"%&',/:(&%2(F"0,&'(91:0;"'"4:5(7K6!)*$P)+)8!
K0J26!D86!H2@6!A86!U<2442?"6!Y86!>!U010S26!.8!B%--'C8!*%+,$"%-.%/&'(P.',.)1W(&%(.-O,$,0&'(
+&',2&/,"%(J,/;(?$&>,',&%(1/#2.%/18!K0V;<!V<;@;FJ;/!0J!JM;!$+JM!D2FL;<;F4;!2L!JM;!
XFJ;<F0J"2F0?!.@@24"0J"2F!L2<!K;2V?;!OF#"<2F1;FJ!IJ:/";@8!!
K0:?M:@6!_8!98!B$,+'C8!U32P421V2F;FJ!12/;?@!2L!@24"0??S!/;@"<0E?;!<;@V2F/"FG8!!"#$%&'(")(
9.$1"%&',/:(&%2(F"0,&'(91:0;"'"4:5(L76!(,+P)-,8!
K0:?M:@6!_8!98!B%--%C8!I24"0??S!_;@"<0E?;!H;@V2F/"FGT!UM;!O#2?:J"2F!2L!0!D2F@J<:4J8!XF![8!X8!
Y<0:F6!_8!Q8!=04^@2F!>!_8!O8!7"?;S!BO/@8C6!N;.($"'.(")(0"%1/$#0/1(,%(O1:0;"'"4,0&'(&%2(
.2#0&/,"%&'(-.&1#$.-.%/!BVV8!',P),C8!A0M30M6!Q=T!O<?E0:18!
K0:?M:@6!_8!986!>!H;"/6!_8!Y8!B$,,$C8!OFM0F4;1;FJ!0F/!/;F"0?!"F!@24"0??S!/;@"<0E?;!
<;@V2F/"FG8!!"#$%&'(")(9.$1"%&',/:(&%2(F"0,&'(91:0;"'"4:5(7C6!&-*P&$*8!
K;?M016!Y8!78!B$,,&C8!UM;!"/2G<0VM"4!F0J:<;!2L!M:10F!V;<@2F0?"JST!Oe01V?;@!2L!JM;!
"/"2G<0VM"4!@;?LP42F4;VJ8!!"#$%&'(")(9.$1"%&',/:(&%2(F"0,&'(91:0;"'"4:5(7L6!))(P)**8!
K;J;<@6!_8!986!>!A4D2<1"4^6!O8!=8!B$,))C8!D21V0<0J"#;!<;?"0E"?"JS!2L!F:1;<"40??S!0F4M2<;/!
#;<@:@!W2EPJ0@^!0F4M2<;/!<0J"FG!@40?;@8!!"#$%&'(")(EOO',.2(91:0;"'"4:5(BC6!,%P,)8!
K"<0G;@6!_8!D86!>!OM<?"4M6!K8!H8!B$,*'C8!E$R(SSW(F"0,&'($.1O"%1.(/"(.%+,$"%-.%/&'(,-O.$&/,+.18!
I0F!Z<0F4"@42T!78![8!Z<;;10F8!

$'%!
!

!

K<:"JJ6!_8!\8!B$,*%C8!A;JM2/@!L2<!<;@2?#"FG!/"LL;<;F4;@!2L!"FJ;<;@JT!.!JM;2<;J"40?!0F0?S@"@8!
!"#$%&'(")(F"0,&'(S11#.15(6KB$C6!$&&P$('8!
H"4M"F@6!A8!986!>!_03@2F6!I8!B$,,%C8!.!42F@:1;<@!#0?:;@!2<";FJ0J"2F!L2<!10J;<"0?"@1!0F/!"J@!
1;0@:<;1;FJT!I40?;!/;#;?2V1;FJ!0F/!#0?"/0J"2F8!!"#$%&'(")(8"%1#-.$(I.1.&$0;5(<[6!
&-&P&$)8!
H"F/L?;"@4M6!.86!Y:<<2:GM@6!=8!O86!>!_;FJ2F6!Z8!B$,,*C8!Z01"?S!@J<:4J:<;6!10J;<"0?"@16!0F/!
421V:?@"#;!42F@:1VJ"2F8!!"#$%&'(")(8"%1#-.$(I.1.&$0;5(6G6!&$%P&%(8!
H"JJ;?6![8!786!>!7;EE;<6!A8!A8!B$,*&C8!_"?;110@!"F!0!\;F;<0?!UM;2<S!2L!K?0FF"FG8!9"',0:(
F0,.%0.15(L6!$((P$),8!
H2E"F@2F6!=8!K8!B$,+(C8!UM;!#0?"/"JS!0F/!<;?"0E"?"JS!2L!/"0<";@!#;<@:@!0?J;<F0J"#;!J"1;!:@;!
1;0@:<;@8!XF!Z8!U8!=:@J;<!>!Z8!K8!IJ0LL2</!BO/@8C6!N,-.5(4""215(&%2(J.''ZP.,%48!.FF!
.<E2<T!XF@J"J:J;!L2<!I24"0?!H;@;0<4M6!`F"#;<"@JS!2L!A"4M"G0F8!
H2^;04M6!A8!B$,)+C8!?.',.)15(&//,/#2.15(&%2(+&'#.18!I0F!Z<0F4"@42T!=2@@;SPY0@@8!
H2^;04M6!A8!B$,*&C8!N;.(%&/#$.(")(;#-&%(+&'#.18!Q;3!R2<^T!Z<;;!K<;@@8!
HS0F6!H8!A86!DM"<^2#6!N8!X86!9"JJ?;6!U8!_86!IM;?/2F6!c8!A86!U"12@M"F06!O86!>!_;4"6!O8!98!B$,,,C8!UM;!
.1;<"40F!/<;01!"F!H:@@"0T!OeJ<"F@"4!0@V"<0J"2F@!0F/!3;??PE;"FG!"F!J32!4:?J:<;@8!
9.$1"%&',/:(&%2(F"0,&'(91:0;"'"4:(?#''./,%5(6B6!$(-,P$(%'8!
I04^;"16![8!.86!>!\:<6!H8!D8!B$,*+C8!I;?LP/;4;VJ"2F6!2JM;<P/;4;VJ"2F!0F/!42F@4"2:@F;@@8!XF!\8!
O8!I4M30<J5!>!_8!IM0V"<2!BO/@8C6!8"%10,"#1%.11(&%2(1.')Z$.4#'&/,"%W(E2+&%0.1(,%(
$.1.&$0;!BN2?8!%6!VV8!$&,P$,*C8!Q;3!R2<^T!K?;F:1!K<;@@8!
I0:F/;<@6!I86!>!A:F<26!_8!B%---C8!UM;!42F@J<:4J"2F!0F/!#0?"/0J"2F!2L!0!42F@:1;<!
2<";FJ0J"2F!f:;@J"2FF0"<;!BIDdXC!/;@"GF;/!J2!1;0@:<;!Z<211]@!B$,((C!]10<^;J"FG!
4M0<04J;<]!"F!.:@J<0?"08!F"0,&'(?.;&+,"$(&%2(9.$1"%&',/:5(6K6!%$,P%'-8!
I4M0MF6!=8!B%--%C8!UM;!<2?;!2L!E;M0#"2<0?!42@J@!0F/!@24"0?!/;@"<0E"?"JS!0@!V<;/"4J2<@!2L!
;F#"<2F1;FJ0?!0JJ"J:/;@!0F/!42F@;<#0J"2F!E;M0#"2<T!0F!0F0?S@"@!2F!0GG<;G0J;!0F/!2F!
"F/"#"/:0?!/0J0!?;#;?8!^.,/10;$,)/()#$(A,)).$.%/,.''.(#%2(A,&4%"1/,10;.(91:0;"'"4,.5(6G6!
'(P('8!

$'&!
!

!

I4M0MF6!=86!>!Y2MF;<6!\8!B%--%C8!*+&'#&/,"%(I.1.&$0;(,%(*%+,$"%-.%/&'(F"0,"'"4:(&%2(,%(
*%+,$"%-.%/&'(91:0;"'"4:W(3./;"2"'"4,0&'(E1O.0/18!K0V;<!V<;@;FJ;/!0J!JM;!Y;<"4MJ!
0:@!/;1!K@S4M2?2G"@4M;F!XF@J"J:J8!!
I4M?;F^;<6!Y8!H86!>!7;"G2?/6!A8!Z8!B$,+,C8!\20?@!0F/!JM;!@;?LP"/;FJ"L"40J"2F!V<24;@@T!
D2F@J<:4J"FG!/;@"<;/!"/;FJ"J";@8!XF!98!.8!K;<#"F!BO/8C6!U"&'(0"%0.O/1(,%(O.$1"%&',/:(&%2(
1"0,&'(O1:0;"'"4:!BVV8!%'&P%,-C8![X??@/0?;6!Q=T!903<;F4;!O<?E0:1!.@@24"0J;@8!
I4M1:4^6!K86!c0@@;<6!U86!>!HS0F6!H8!A8!B%---C8!XFJ<"F@"4!0F/!;eJ<"F@"4!G20?@T!UM;"<!@J<:4J:<;!
0F/!<;?0J"2F@M"V!J2!3;??PE;"FG!"F!\;<10F!0F/!`8I8!42??;G;!@J:/;FJ@8!F"0,&'(S%2,0&/"$1(
I.1.&$0;5(BC6!%%(P%'$8!
I4M:?J56!K8!786!\2:#;"06!N8!N86!D01;<2F6!98!_86!U0F^M06!\86!I4M1:4^6!K86!>!Z<0F;^6!A8!B%--(C8!
N0?:;@!0F/!JM;"<!<;?0J"2F@M"V!J2!;F#"<2F1;FJ0?!42F4;<F!0F/!42F@;<#0J"2F!E;M0#"2<8!
!"#$%&'(")(8$"11Z8#'/#$&'(91:0;"'"4:5(G76!'(*P'*(8!
I4M:?J56!K8!786!i;?;5FS6!98!D86!>!_0?<S1V?;6!Q8!=8!B%---C8!.!1:?J"F0J"2F0?!V;<@V;4J"#;!2F!JM;!
<;?0J"2F!E;J3;;F!=:/;2PDM<"@J"0F!<;?"G"2:@!E;?";L@!0F/!0JJ"J:/;@!2L!;F#"<2F1;FJ0?!
42F4;<F8!*%+,$"%-.%/(&%2(?.;&+,"$5(G66!(*)P(,$8!
I4M:10F6![86!>!=2MF@2F6!A8!K8!B$,*)C8!.JJ"J:/;@!0F/!Y;M0#"2<8!E%%#&'(I.+,.J(")(F"0,"'"4:5(66!
$)$P%-*8!
I4M30<J56!I8![8!B$,**C8!Q2<10J"#;!"FL?:;F4;@!2F!0?J<:"@18!XF!98!Y;<^23"J5!BO/8C6!E2+&%0.1(,%(
*VO.$,-.%/&'(F"0,&'(91:0;"'"4:!BN2?8!$-6!VV8!%%$P%*,C8!Q;3!R2<^T!.40/;1"4!K<;@@8!
I4M30<J56!I8![8!B$,,%C8!`F"#;<@0?@!"F!JM;!42FJ;FJ!0F/!@J<:4J:<;!2L!#0?:;@T!UM;2<;J"40?!
0/#0F4;@!0F/!;1V"<"40?!J;@J@!"F!%-!42:FJ<";@8!XF!A8!K8!i0FF0!BO/8C6!E2+&%0.1(,%(
.VO.$,-.%/&'(1"0,&'(O1:0;"'"4:!BN2?8!%(6!VV8!$P)(C8!d<?0F/26!Z9T!.40/;1"48!
I4M30<J56!I8![8!B$,,'C8!.<;!UM;<;!`F"#;<@0?!.@V;4J@!"F!JM;!IJ<:4J:<;!0F/!D2FJ;FJ@!2L!
[:10F!N0?:;@g!!"#$%&'(")(F"0,&'(S11#.15(BCB'C6!$,P'(8!
I4M30<J56!I8![8!B%--)C8!9;@!#0?;:<@!/;!E0@;!/;!?0!V;<@2FF;T!JM;2<";6!1;0@:<;@!;J!
0VV?"40J"2F@8!I.+#.(X$&%0&,1.(2.(F"0,"'"4,.5(LMB'C6!,%,P,)+8!
I4M30<J56!I8![86!>!Y""@^S6!78!B$,+*C8!U230</!0!V@S4M2?2G"40?!@J<:4J:<;!2L!M:10F!#0?:;@8!
!"#$%&'(")(9.$1"%&',/:(&%2(F"0,&'(91:0;"'"4:5(BG6!((-P()%8!

$''!
!

!

I4M30<J56!I8![86!>![230</6!=8!.8!B$,+$C8!.!F2<10J"#;!/;4"@"2FP10^"FG!12/;?!2L!0?J<:"@18!XF!=8!
K8!H:@MJ2F!>!H8!A8!I2<<;FJ"F2!BO/@8C6!E'/$#,1-(&%2(;.'O,%4(P.;&+,"$W(F"0,&'5(
O.$1"%&',/:5(&%2(2.+.'"O-.%/&'(O.$1O.0/,+.18![X??@/0?;6!Q=T!903<;F4;!O<?E0:18!
I4M30<J56!I8![86!>!U;@@?;<6!H8!D8!B$,*%C8!.!J;@J!2L!12/;?!L2<!<;/:4"FG!1;0@:<;/!0JJ"J:/;P
E;M0#"2<!/"@4<;V0F4";@8!!"#$%&'(")(9.$1"%&',/:(&%2(F"0,&'(91:0;"'"4:5(6L6!%%(P%&)8!
I4M30<56!Q86!>!Y?;@@6![8!B$,,%C8!D2F@J<:4J"FG!<;0?"JS!0F/!"J@!0?J;<F0J"#;@T!.@@"1"?0J"2F!0F/!
42FJ<0@J!;LL;4J@!"F!@24"0?!W:/G1;FJ8!XF!98!98!A0<J"F!>!.8!U;@@;<!BO/@8C6!N;.(
0"%1/$#0/,"%(")(1"0,&'(b#24.-.%/1!BVV8!%$*P%'(C8!["??@/0?;6!Q=T!O<?E0:18!
I4M30<56!Q86!cF0:V;<6!Y86!["VV?;<6![8!=86!Q2;??;PQ;:10FF6!O86!>!D?0<^6!98!Z8!B$,,$C8!H0J"FG!
@40?;@T!Q:1;<"4!#0?:;@!10S!4M0FG;!JM;!1;0F"FG!2L!@40?;!?0E;?@8!9#P',0(=O,%,"%(
H#&$/.$':5(BB6!(*-P(+%8!
I4M30<56!Q86!IJ<04^6!Z86!>!A0"6![8PK8!B$,,$C8!.@@"1"?0J"2F!0F/!42FJ<0@J!;LL;4J@!"F!V0<JP3M2?;!
f:;@J"2F!@;f:;F4;@T!.!42F#;<@0J"2F0?!?2G"4!0F0?S@"@8!9#P',0(=O,%,"%(H#&$/.$':5(BBB$C6!
&P%&8!
I42??2F6!D8!Q86!c"1PK<";J26!D86!>!_";F;<6!O8!B%--&C8!OeV;<";F4;!I01V?"FGT!K<21"@;@!0F/!
K"JL0??@6!IJ<;FGJM@!0F/!7;0^F;@@;@8!!"#$%&'(")(@&OO,%.11(F/#2,.15(L6!(P&'8!
IM;?/2F6!c8!A86!>!c0@@;<6!U8!B$,,+C8!K:<@:"FG!V;<@2F0?!G20?@T!I^"??@!;F0E?;!V<2G<;@@!E:J!F2J!
0??!V<2G<;@@!"@!E;F;L"4"0?8!9.$1"%&',/:(&%2(F"0,&'(91:0;"'"4:(?#''./,%5(6L6!$&$,P$&&$8!
IM;?/2F6!c8!A86!>!A4\<;G2<6![8!B%---C8!OeJ<"F@"4!#0?:;!2<";FJ0J"2F!0F/!JM;!J<0G;/S!2L!JM;!
42112F@8!!"#$%&'(")(9.$1"%&',/:5(7K6!&+&P'$$8!
IM;?/2F6!c8!A86!HS0F6!H8!A86!_;4"6!O8!986!>!c0@@;<6!U8!B%--'C8!UM;!XF/;V;F/;FJ!OLL;4J@!2L!\20?!
D2FJ;FJ@!0F/!A2J"#;@!2F!7;??PY;"FGT!XJ]@!Y2JM!7M0J!R2:!K:<@:;!0F/!7MS!R2:!
K:<@:;!XJ8!9.$1"%&',/:(&%2(F"0,&'(91:0;"'"4:(?#''./,%5(GCB'C6!'*(P'+)8!
IM"LL10F6!I8!I86![:LL2</6!A86!["4^42e6!A86!K0JS6!=86!\FS@6!A86!>!c0@@;?6!=8!B$,,*C8!H;1;1E;<!
JM0Jg!0!421V0<"@2F!2L!<;0?PJ"1;!#;<@:@!<;J<2@V;4J"#;!<;40??!2L!@12^"FG!?0V@;@8!
!"#$%&'(")(8"%1#'/,%4(&%2(8',%,0&'(91:0;"'"4:5(7BB%C8!
IM2G<;F6!=8!Z86!>!U0S?2<6!98!d8!B%--+C8!dF!Y;M0#"2<0?POF#"<2F1;FJ0?!O42F21"4@8!I.+,.J(")(
*%+,$"%-.%/&'(*0"%"-,01(&%2(9"',0:5(6B$C6!%)P''8!

$'(!
!

!

I?2E2/06!=8!.86!d]Q;"??6!I8!.86!>!X#0?/"6!.8!B%--$C8!Z:F4J"2F@!2L!1:@"4!"F!;#;<S/0S!?"L;T!.F!
;eV?2<0J2<S!@J:/S!:@"FG!JM;!OeV;<";F4;!I01V?"FG!A;JM2/8!3#1,0&.(F0,.%/,&.5(c6!,P&%8!
I2M<6!I8!B%--$C8!O42P04J"#"@1!0F/!3;??PE;"FGT!Y;J3;;F!L?23!0F/!E:<F2:J8!XF!K8!I4M1:4^!>!c8!
A8!IM;?/2F!BO/@8C6!Q,).(U"&'1(&%2(\.''Z?.,%4W(N"J&$21(&(9"1,/,+.(91:0;"'"4:(")(@#-&%(
F/$,+,%4!BVV8!%-%P%$(C8!I;0JJ?;6!7.T![2G<;L;!0F/![:E;<8!
IJ;<F6!K8!.86!_";J56!U86!>!\:0GF0F26!\8!.8!B$,,(C8!UM;!Q;3!O42?2G"40?!K0<0/"G1!"F!I24"0?P
K@S4M2?2G"40?!D2FJ;eJ8!*%+,$"%-.%/(&%2(?.;&+,"$5(6MB)C6!*%&P*'&8!
IJ;<F6!K8!D8!B%---C8!U230</!0!D2M;<;FJ!UM;2<S!2L!OF#"<2F1;FJ0??S!I"GF"L"40FJ!Y;M0#"2<8!
!"#$%&'(")(F"0,&'(S11#.15(B7B&C6!'-*P'%'8!
IJ;<F6!K8!D86!>!_";J56!U8!B$,,,C8!.!N0?:;PY;?";LPQ2<1!UM;2<S!2L!I:VV2<J!L2<!I24"0?!
A2#;1;FJ@T!UM;!D0@;!2L!OF#"<2F1;FJ0?"@18!@#-&%(*0"'"4:(I.+,.J5(7B%C6!+$P,*8!
IJ;<F6!K8!D86!_";J56!U86!>!c0?2L6!98!B$,,&C8!N0?:;!d<";FJ0J"2F@6!\;F/;<6!0F/!OF#"<2F1;FJ0?!
D2F4;<F8!*%+,$"%-.%/(&%2(?.;&+,"$5(6BB&C6!&%%P&'+8!
IJ;<F6!K8!D86!>!d@^01V6!I8!B$,+*C8!A0F0G"FG!@40<4;!;F#"<2F1;FJ0?!<;@2:<4;@8!XF!_8!IJ2^2?@!>!
X8!.?J10F!BO/@8C6!@&%2P""R(")(.%+,$"%-.%/&'(O1:0;"'"4:!BN2?8!%6!VV8!$-'&P$-++C8!Q;3!
R2<^T!7"?;S8!
IJ"4^G2?/6!H86!A0?"06!.86!Z2@@;6!=8!.86!>![2E@2F6!=8!.8!B%---C8!Y<0"FP1"F/!@J0J;@T!X8!?2FG"J:/"F0?!
L";?/!@J:/S!2L!@?;;Vj30^;!L04J2<@!"FL?:;F4"FG!1;FJ0J"2F!<;V2<J"2F!?;FGJM8!F'..OW(
!"#$%&'(")(F'..O(&%2(F'..O(A,1"$2.$(I.1.&$0;5(6L6!$&,P%'%8!
IJ2F;6!.8!.86!IM"LL10F6!I8!I86!>!_;N<";@6!A8!78!B$,,,C8!O42?2G"40?!121;FJ0<S!0@@;@@1;FJ8!XF!
_8!c0MF;10F6!O8!_";F;<!>!Q8!I4M30<J5!BO/@8C6!\.''Z?.,%4W(X"#%2&/,"%1(")(&(@.2"%,0(
91:0;"'"4:!BVV8!%)P&,C8!Q;3!R2<^T!H:@@;??!I0G;!Z2:F/0J"2F8!
U0<<0FJ6!A8!.86!>!D2</;??6![8!c8!B$,,*C8!UM;!;LL;4J!2L!<;@V2F/;FJ!4M0<04J;<"@J"4@!2F!G;F;<0?!
;F#"<2F1;FJ0?!0JJ"J:/;PE;M0#"2<!42<<;@V2F/;F4;8!*%+,$"%-.%/(&%2(?.;&+,"$5(6[6!
)$+P)&*8!
U2:<0FG;0:6!H86!>!Y<0/E:<F6!Q8!A8!B%-$-C8!UM;!K@S4M2?2GS!2L!I:<#;S!H;@V2F@;8!XF!K8!N8!
A0<@/;F!>!=8!_8!7<"GMJ!BO/@8C6!@&%2P""R(")(F#$+.:(I.1.&$0;8!Y"FG?;S6!`cT!O1;<0?/!
\<2:V8!

$')!
!

!

U2:<0FG;0:6!H86!H"V@6!986!>!H0@"F@^"6!c8!B%---C8!N;.(O1:0;"'"4:(")(1#$+.:($.1O"%1.8!
D01E<"/G;T!D01E<"/G;!`F"#;<@"JS!K<;@@8!
U#;<@^S6!.86!>!c0MF;10F6!_8!B$,+$C8!UM;!Z<01"FG!2L!_;4"@"2F@!0F/!JM;!K@S4M2?2GS!2L!
DM2"4;8!F0,.%0.5(6<<B''+$C6!'(&P'(+8!
`S6!A8!.86!Z226!A8P_86!>!.G:"F"@6![8!B%-$-C8!`@"FG!OeV;<";F4;!I01V?"FG!A;JM2/2?2GS!J2!
./#0F4;!OFJ<;V;F;:<@M"V!UM;2<S!0F/!H;@;0<4M8!=$4&%,>&/,"%&'(I.1.&$0;(3./;"215(
<GB$C6!&$P('8!
N;</:SF6!K86!_;?#0:e6!O86!N0F!D2"??";6![86!U:;<?"F4^e6!Z86!>!A;4M;?;F6!X8!N8!B%--,C8!K<;/"4J"FG!
JM;!_:<0J"2F!2L!O12J"2F0?!OeV;<";F4;T!U32!OeV;<";F4;!I01V?"FG!IJ:/";@8!*-"/,"%5(
[B$C6!+&P,$8!
70"J;6!Y8!A86!D?0LL;S6!H86!>!["??E<0F/6!A8!B$,,+C8!_"LL;<;F4;@!E;J3;;F!#2?:FJ;;<@!0F/!
F2F#2?:FJ;;<@!"F!0!M"GMP/;10F/!@;?LP<;42</"FG!@J:/S8!91:0;"'"4,0&'(I.O"$/15(KG6!
$,,P%$-8!
70??04M6!A8!.86!>!70??04M6!98!B$,+&C8!91:0;"'"4:T1(1&%0/,"%()"$(1.'),1;%.11W(N;.(.$$"$(")(
.4",1-(,%(/;."$:(&%2(/;.$&O:8!I0F!Z<0F4"@42T!78![8!Z<;;10F8!
7;EE6!O8!=86!D01VE;??6!_8!U86!I4M30<J56!H8!_86!>!I;4M<;@J6!98!B$,))C8!D%"P/$#1,+.(3.&1#$.1W(
Y"%$.&0/,+.(I.1.&$0;,%(/;.(F"0,&'(F0,.%0.18!DM"40G2T!H0F/!A4Q0??S8!
7;E;<6!Y86![2?J6!.86!Q;:M0:@6!D86!U<0:JF;<6!K86!O?G;<6!D86!>!U;"4M;<J6!U8!B%--*C8!Q;:<0?!;#"/;F4;!
L2<!<;L;<;F4;P/;V;F/;F4;!"F!<;0?P10<^;J!J<0F@04J"2F@8!Y.#$"S-&4.5(GB6!''$P''*8!
7;"G;?6!H8![8!B$,+&C8!OF#"<2F1;FJ0?!0JJ"J:/;@!0F/!JM;!V<;/"4J"2F!2L!E;M0#"2<8!XF!Q8!H8!
Z;"1;<!>!O8!I8!\;??;<!BO/@8C6!*%+,"$%-.%/&'(91:0;"'"4:W(A,$.0/,"%1(&%2(9.$1O.0/,+.18!
Q;3!R2<^T!K<0;G;<8!
7;??@6!.8!B$,+(C8!c&$,&/,"%1(,%(1.')Z.1/..-(,%(/;.(2&,':(',).(")(-"/;.$1a!`F"#;<@"JS!2L!DM"40G28!
7M"J;!=<86!98!U8!B$,)*C8!UM;!M"@J2<"40?!<22J@!2L!2:<!;42?2G"40?!4<"@"@8!F0,.%0.5(<BB6!$%-&P$%-*8!
7"GG"F@6!=8!I8!B$,(,C8!XFJ;<<;?0J"2F@M"V@!012FG!AAKX!1;0@:<;@!2L!/"@@"1:?0J"2F!:F/;<!
@J0F/0</!0F/!@24"0?!/;@"<0E"?"JS!"F@J<:4J"2F@8!!"#$%&'(")(8"%1#'/,%4(91:0;"'"4:5(6G6!
'$,P'%*8!
$'*!
!

!

7"??"01@6!\8!D86!D2e6!O8!A86![;/E;<G6!N8!.86!>!_;4"6!O8!98!B%---C8!OeJ<"F@"4!?"L;!G20?@!0F/!M;0?JM!
<"@^!E;M0#"2<@!"F!0/2?;@4;FJ@8!!"#$%&'(")(EOO',.2(F"0,&'(91:0;"'"4:5(GC6!$*()P$**$8!
7"??"01@6!H8!B$,)+C8!UM;!D2F4;VJ!2L!Q2<1@8!XF!_8!I"??@!BO/8C6!S%/.$%&/,"%&'(*%0:0'"O.2,&(")(
/;.(F"0,&'(F0,.%0.18!Q;3!R2<^T!A041"??0F8!
7"@;10F6!A86!>!Y2GF;<6!Z8!l8!B%--&C8!.!M"GM;<P2</;<!12/;?!2L!;42?2G"40?!#0?:;@!0F/!"J@!
<;?0J"2F@M"V!J2!V;<@2F0?"JS8!9.$1"%&',/:(&%2(S%2,+,2#&'(A,)).$.%0.15(GL6!*+&P*,'8!
R0<1;S6!_8!B$,*,C8!N;.(O1:0;"'"4:(")(.:.J,/%.11(/.1/,-"%:8!Q;3!R2<^T!Z<;;!K<;@@8!
i:50F;^6!=8!B$,,,C8!*VO.$,.%0.(F&-O',%4(3./;"2W(8#$$.%/(&%2(O"/.%/,&'($.1.&$0;(
&OO',0&/,"%18!K0V;<!V<;@;FJ;/!0J!JM;!72<^@M2V!2F!U"1;P:@;!A;0@:<;1;FJ!0F/!
H;@;0<4M8!!
!

$'+!
!

!