A .

r5! {’5'}... .f. .:

74. 117.13.! C

[ULRJRHH .
jfi

an".

11:29.53
15...! ILA.» . \
If" u:

 

“1mm

/

(.

J
K‘ n
9- 9-02

This is to certify that the

dissertation entitled

Psychological Determinants of the Intention
to Support Watershed Best Management Practices

presented by

STEPHEN R. PENNINGTON

has been accepted towards fulfillment
of the requirements for

Doctoral Resource Development

degree in

 

 

/7

Maior professor

 

6-20-02
Date

 

MSU it an Affirmative Action/Equal Opportunity Institution 0- 12771

 

.UBRARY
Michigan State
University

 

 

 

 

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE 'D’ME DUE DATE DUE

 

SEEZ'i 2005;7- 5:1}? 23:4

I
44

 

DEC 1 4 2905

 

3 d

 

 

Away 1005

£53 2 52008

 

93,933,091

 

 

 

 

 

 

 

 

 

6/01 c:/CIRCIDaleDue.p65-p.15

 

PSYCHOLOGICAL DETERMINANTS OF THE INTENTION TO
SUPPORT WATERSHED BEST MANAGEMENT PRACTICES

By

Stephen R. Pennington

A DISSERTATION
Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Resource Development

2002

ABSTRACT

PSYCHOLOGICAL DETERMINANTS OF THE INTENTION TO SUPPORT
WATERSHED BEST MANAGEMENT PRACTICES

By

Stephen R. Pennington

There is a major effort by the United States Environmental Protection Agency to
convince municipalities to manage their water resources on a watershed basis.
Unfortunately, there is little information as to how citizens think about managing water
resources. This study uses two psychological instruments, the fairness heuristic and the
theory of planned behavior in order to better understand the watershed best management
practices (BMPs) peOple intend to support. Using a mail survey (n = 608) of property
owners there were two key attitudes found to be held by the sample population, one based
on the rights of the environment and the other reflecting people’s resistance to change. It
was found that these attitudes are mediated by people’s fairness evaluations of the
process in which policy is developed. Cluster analysis based on the two attitudes showed
respondents held two different worldviews: individualist and egalitarian. Determinants of
the intention to support BMP implementation for the individualist cluster were attitudes
and locus of control whereas for the egalitarian cluster it also included personal
responsibility and level of education. Knowledge was not a determinant for either cluster.
Findings suggest that worldviews are a stronger predictor of the intention to support BMP

implementation than knowledge.

Copyright by
Stephen R. Pennington
2002

I dedicate this dissertation to my wife, Kirsten. Not only did she provide support and
encouragement throughout, but together with our newly born daughter, Emilee, they have
taught me the true meaning of life.

iv

ACKNOWLEDGEMENTS

Sincere appreciation is extended to my Ph.D. advisor, Dr. Scott Witter, for his
support and advice. He was instrumental in the success of my Ph.D. program. His
administrative skills, thoughtfulness and commitment to his students are examples I will
try to follow.

I am truly grateful to Dr. Jo Ann Beckwith, for her ongoing help, support and
fiiendship. Her knowledge of the subject matter helped shape and strengthen this
dissertation.

I wish to express my gratitude to Dr. Ralph Levine, for his continuing help in the
completion of this manuscript. His command of statistics and research methods not only
challenged me, but also helped elevate this manuscript’s quality to the highest possible
level.

Special thanks are extended to Dr. Frank Fear, for his ongoing support and
encouragement from the moment I arrived at Michigan State University. Beginning early
on, I came to rely on his feedback to help focus my writing. His understanding of the
research process helped me sort out the wheat from the chaff and produce a product I can
be proud of.

My sincerest appreciation to my fi’iend and Master’s thesis advisor, Dr. John
FitzGibbon, for his encouragement from the very start. Without his suggestion to
continue my education, I wouldn’t be where I am today.

I want to thank and acknowledge both the Department of Environmental Quality

and Michigan State University Extension for their financial support of my research. The

financial support provided by each helped me stay on track and finish. For this my wife
and l are truly grateful.

I was fortunate to have two peOple proof reading this manuscript. My wife,
whose attention to detail saved me on more than one occasion. I am not only deeply
gratefiil to her for her love, support and encouragement but for her keen mind. My other
proofreader was my sister, Dr. Lori Pennington-Gray. The countless hours of discussion,
her pertinent insight, not to mention love and support were instrumental in getting me
over the inevitable hurdles one encounters along the way.

To my family, I express my sincerest gratitude for their understanding, patience
and encouragement throughout my education. Special thanks to Mom and Dad for
making the return to university a possibility back in 1993 and to my mother in-law, Linda
Ackley for the countless hours of babysitting that allowed me to complete the

dissertation.

vi

TABLE OF CONTENTS

 

 

 

 

 

 

 

LIST OF TABLES IX
LIST OF FIGURES XI
CHAPTER 1: INTRODUCTION AND STATEMENT OF PROBLEM - _ - 1
Theoretical Perspective .................................................................................................. 9
Purpose Statement ........................................................................................................ 13
Research Questions ...................................................................................................... l3
Delimitations ................................................................................................................ 15
Definitions ................................................................................................................... 15
Organization of Study ................................................................ - .................................. 17
CHAPTER II: REVIEW OF THE LITERATURE 18
PART I 19
Values .......................................................................................................................... 19
The Role of Values in Public Decision-Making ........................................................... 20
Value Orientations ....................................................................................................... 27
Value Orientations and the Social Construction of Worldviews .................................. 30
The Evolution of the Fairness Heuristic as an Attitudinal Measure ............................. 36
Basic Categories of Environmental Policy and Implied Justice Concerns ................... 41
PART II 45
The Theories of Reasoned Action (TRA) and Planned Behavior (TPB) ...................... 45
The Behavioral Intention - Behavior Relationship ....................................................... 48
Attitudes ....................................................................................................................... 48
Personal Responsibility ................................................................................................ 53
Locus of Control (LOC) ............................................................................................... 55
Knowledge ................................................................................................................... 57
Demographic Variables ................................................................................................ 61
Predicting and Explaining Intentions and Behavior Using TRA & TPB ..................... 62
Summary ...................................................................................................................... 63
CHAPTER III: METHODOLOGY - - - 66
Study Design ................................................................................................................ 66
Background .................................................................................................................. 66
Population and Sample ................................................................................................ 67
Instrumentation ............................................................................................................ 69
Data Analysis ............................................................................................................... 76

vii

CHAPTER IV: DATA ANALYSIS AND INTERPRETATION - - 79

 

Sample Profile and Biases ............................................................................................ 79
Demographic Characteristics ....................................................................................... 80
Constructing the Attitudinal Clusters ........................................................................... 84
Cluster Demographics .................................................................................................. 90
Characteristics .............................................................................................................. 91
Fairness Evaluations .................................................................................................... 93

Comparison Between Individualist and Egalitarian Clusters’ Fairness Evaluations 98
Relationships Between Knowledge, Locus of Control and Personal

 

Responsibility and the Individualist and Egalitarian Clusters .................................... 101
The Individualist and Egalitarian Clusters and the Intention to Act ........................... 104
Correlations Between Study Variables ....................................................................... 106
Variables Having the Greatest Influence on the Intention to Act ............................... 110
Non-Response Results ............................................................................................... 113
Study Limitations ....................................................................................................... 114
CHAPTER V: CONCLUSIONS, DISCUSSION AND RECOMMENDATIONS. 116
Conclusions and Discussion ....................................................................................... 116
Implications for the Management of Sycamore Creek Watershed ............................. 132
Recommendations for Future Research ..................................................................... 134
APPENDICIES - _ -- -- 136
Appendix A: Survey And Response Rate Data .......................................................... 137
Appendix B: Response Frequencies For Survey Questions ....................................... 154
Appendix C: Exploratory Factor Analysis Output ..................................................... 173
Appendix D: Cluster Analysis Output ....................................................................... 184
Appendix E: Fairness Heuristic Analysis Output ....................................................... 187
Appendix F: BMP Analysis .................................................................... 208
Appendix G: Regression Analysis Output ................................................................. 210
BIBLIOGRAPHY _ _ -_ _ - _ 217

 

viii

Table I:

Table 2:

Table 3:

Table 4:

Table 5:

Table 6:

Table 7:

Table 8:

Table 9:

Table 10:
Table 11:

Table 12:

Table 13:

Table 14:

Table 15:

Table 16:
Table 17:

Table 18:

LIST OF TABLES

Grid/Group Nomenclature and Worldviews .................................................. 35
Summary of Findings from Meta-Analyses of the TRA and TPB ................. 62
Examples of How Different Effect Size Can Give a Different Impression 63

Summary of Findings on the Strength of the Relationships Between
Psychological Variables ................................................................................ 65

Research Questions, Study Variables and the Items on the Questionnaire. . . .77

Demographic Characteristics of Respondents Compared to the Population
of Ingham County .......................................................................................... 80

Principle Component Factor Analysis Results of 20 Heuristic Statements 85

Correlation Matrix of the Three Factors ......................................................... 86
Attitudinal Factor Items ................................................................ 87
Means of Fairness Heuristic Factors by Clusters ........................................... 89
Demographic Profiles of Cluster Groups ....................................................... 91

Percentage Distribution, Mean Score and Standard Deviation of
Responses to Fairness Heuristic Statements .................................................. 94

Percentage of Respondents Who Agreed with and Ranked Preference for

Each Fairness Heuristic Statement .................................................. 96
Comparison of Study Variables by Attitudinal Groups ................................ 102
Percentage of Respondents Who Support and Ranked Preference for

Each of the Best Management Practice Options .......................................... 105
Pearson Correlations (N = 608) Between Study Variables .......................... 107
Partial Correlations Coefficients .................................................................. 108

Multiple Regression Estimates for the Intention to Support BMP
Implementation ............................................................................................ l 1 1

ix

Table 19: By Cluster Regression Estimates for the Intention to Support BMP
Implementation ............................................................................................ 1 l3

Figure 1:

Figure 2:

Figure 3:
Figure 4:
Figure 5:

Figure 6:

LIST OF FIGURES

Proposed Model of Responsible Environmental Behavioral Intentions
Based on Hines, Hungerford and Tomera (1986)

 

The Cognitive Hierarchy

 

Path Models for Theory of Reasoned Action (A) and Theory of

 

Planned Behavior (B)

Map of Sycamore Creek Watershed in Relation to
Ingham County, Michigan

 

Hines, Hungerford and Tomera’ s (1986) Proposed Model of
Responsible Environmental Behavior

 

Cluster Membership and Centers by Factors. ...............

 

xi

14, 71

20

-46

67

-- 70
“-89

CHAPTER I:

INTRODUCTION AND STATEMENT OF PROBLEM

Twenty-five years of water pollution regulation under the 1972 Clean Water Act
(CWA) has failed to achieve its stated goal to "restore and maintain the chemical,
physical, and biological integrity of the Nation's waters" [CWA, 1972] largely because of
the Environmental Protection Agency’s (EPA) inability to control nonpoint source
pollution. Nonpoint source (NPS) pollution or polluted runoff remains a leading cause of
water pollution in both agricultural and urban areas (USEPA, 1997) and therefore the
impacts are everywhere. Regardless of the large expenditures by industry and municipal
wastewater systems to reduce point source pollutions, forty percent of the United States’
waterways still do not meet the minimum federal guidelines (USEPA, 1997).

The Clean Water Act addresses water quality problems in two ways. The Act
requires states to identify waterbodies whose quality does not support their designated
beneficial uses. Total Maximum Daily Loads (TMDLs) must then be developed for each
of the listed segments based on the assimilation capacity of the system. In many cases,
structural best management practices (BMPs) have to be implemented to obtain the
required reduction in pollutant concentrations.

In addition, Phase I of the National Pollution Discharge Elimination System’s
(NPDES) stormwater program requires medium and large cities (100,000+ people) to
obtain permits for the discharge of stormwater runoff. Phase II of the stormwater program
targets municipalities under 100,000 pe0ple and requires them to have stormwater
management plans in place by March, 2003. These permits contain requirements for

implementation of BMPs mostly nonstructural, however, there is a growing awareness

that additional measures such as structural BMPs, may be necessary to achieve the
desired quality of storm water discharges. Many metropolitan and regional agencies also
require stormwater treatment in environmentally sensitive areas. Consequently, structural
BMPs designed to improve water quality are being installed at an unprecedented rate
(Barrett, 2000; Roesner, Bledsoe, & Brashear, 2001).

Conventional structural BMPs include extended detention basins, biofilters
(vegetated strips and swales), sand filters, infiltration devices and wet basins to name a
few. Nonstructural BMPs typically take the form of management guidelines, regulations
and information and education (I & E) programs. Under pressure from regulators,
environmentalists and other stakeholders, structural BMPs are often installed without
regard to the nature of the impairment of the receiving water nor the impact to the local
community (Barrett, 2000). The goal of this research is to understand how pe0ple think
about their local watershed and the BMP strategies for improving water quality they
would support.

Water-resource scientists have presented a convincing case over the years for
giving greater weight to the human dimensions of water management, but have identified
a number of problems in trying to accomplish this goal (Dunlap & Scarce, 1991; Harris,
1977; Kempton, Boster, & Harley, 1995; Satterfield & Gregory, 1998). Over 30 years
ago, Biawas and Durie (1971) argued that water resource decisions ought to be based
primarily on social criteria because the ultimate goal is to improve the quality of human
life. They suggested that planners not restrict themselves to technical and economic
measures in evaluating alternatives, but seek out and apply human dimension factors as

well. Since that time, many researchers, including Biawas (1973), have concluded that

incorporating social dimensions will make the planning process more relevant and
meaningful, but at the same time will render a complicated process even more complex
(Dunlap & Scarce, 1991; Harris, 1977; Satterfield & Gregory, 1998; Syme & Nancarrow,
1997b)

There has been surprisingly little research on whether there are consistent
dimensions in the ways in which people integrate their thoughts about water resources.
The studies available have been conducted from the perspective of a particular activity,
use or amenity associated with water. For example, Syme and Williams (1993c)
examined the structure of perceptions in the context of the aesthetics of drinking water.
Smith and colleagues (1991, 1992) have examined color and clarity in terms of
recreational use of water for bathing. Other authors have discussed the meaning of water
in cultural or spiritual terms (Woolmington & Burgess, 1983). Yet, there are a few
research studies that try to empirically substantiate the dimensions surrounding water
resources.

In the context of water planning, Harris (1977) provided the first multi-
dimensional analysis of the conceptions of water in an attempt to find some underlying
dimensions that could assist planners within a multi-obj ective decision-making
framework. With a sample of three hundred respondents from disparate subgroups, he
used multi—dimensional scaling techniques to find five principal vectors relating to (1)
quality of drinking water, (2) allocation and conservation, (3) natural beauties of water,
(4) public involvement and (5) public access to water bodies. These vectors were

consistent between subgroups (e. g. social scientists and water engineers).

Nancarrow et al. (1996-97) found similar dimensions existed in their two studies.
The first study was conducted on 1080 residents of three Australian cities (Perth,
Canberra and Sydney) while the second, conducted two years later, was a socio-economic
stratified sample of the residents of Perth. Both used personal interview methodology for
administering the questionnaire. They found three common factors: aesthetics,
conservation and utility. When they tried to cluster respondents based upon these factors
a four-cluster solution was obtained. They named these: "Self-Interested", "Earthy",
"Environmentalists" and "Service Oriented" people. Interestingly, when the water rights
statement was included, the difference between the cluster means for the three factors
mentioned above was not evident. It was found that the water rights statement removed
the discriminating influence between clusters.

In recent years the concerns about increasing the efficiency and effectiveness of
water management have meant that issues such as pr0perty rights, the rights of the
environment and the social and economic bases underpinning management decisions
have demanded greater attention (Syme & Nancarrow, 1997a). The basic underlying
themes in most planning disputes relate to what is just, fair and equitable in terms of who
should benefit from planning, who should bear the costs and how should the decisions be
made (Syme & Nancarrow, 1997b). Government policies constantly state that resources
will be allocated equitably, yet the area that has received the least attention is the
definition of what is just, fair or equitable as seen by the range of stakeholders in
watershed planning decisions (Syme & Nancarrow, l997b).

Syme and Fenton (1993a) attempted to examine the structure of equity and

proportional preferences (Rasinski, 1987) for allocation decision-making for groundwater

in Perth, Australia. There was a partial replication of Razinski’s two-factor (equality and
prOportionality) structure of equity. Furthermore, there was a stated preference for
arbitration procedures in dispute settlement (Lind & Tyler, 1988).

Despite these encouraging results, the qualitative feedback from respondents
indicated that the two-factor notion of equity did not incorporate all the subtleties of
values that people consider to be important in water allocation planning. Also, the
preferences provided by the Razinski’s (1987) framework were thought by some to be
too simple for decision-making with multiple conflicts. As a result of these conclusions, 3
series of studies aimed at establishing universal fairness criteria were undertaken.

The second study tried to broaden the measures of the philosophical basis (Wenz,
1988) on which the community might derive fairness perceptions. Such concepts as
people’s attitudes towards short and long-term planning and the concept of procedural
justice were tested from a wider philosophical base (Syme, Nancarrow, & McCreddin,
1999). The 111 philosophical statements moved fi'om virtue theory (people who already
have the resource are inherently deserving), through ideas of the common good and
differing formulations of benefit/cost analysis. When the philosophical questions were
examined thematically, there was very clear support for the following positions: (1) water
is a common good and should be managed for the welfare of the community as a whole;
(2) more than market mechanisms are required for an adequate holistic allocation policy;
(3) efficiency of use is an important component when considering allocation; (4) there is
a moral obligation that human users affect others as little as possible; (5) water quality

should be considered as well as quantity; (6) there is an obligation for general public

information or involvement in allocation decisions; and (7) the environment has
allocation ‘rights’(Syme & Nancarrow, 1996).

Interestingly, there was only a modest correlation between long-term planning
(certainty) and agreement on allocation decisions being short-term and dependent on the
circumstances at the time (Syme & Nancarrow, 1996). The interaction between these
dimensions would seem to indicate the importance of framing fairness judgments in
particular situations. This led to the adoption of universal fairness and situational fairness
principles in later studies (Syme & Nancarrow, 1996). Universal fairness criteria are
guiding principles used to evaluate all decisions where as situational principles arise only
when the outcome is likely to impact the community or the individual. Also lending
support to this dichotomy was the fact that respondents tended to support prior rights in
their allocation priorities, but they did not support items measuring attitudes towards
prior rights as a universal principle. Therefore, the subsequent studies made the
conceptual distinction between fairness principles being applied differently at the
universal and situational levels (Syme & Nancarrow, 1996).

Several other studies served to refine and substantiate Syme and his colleagues
(1999) findings under a variety of circumstances. For instance, the third study refined and
applied the universal philosophies to actual decision-making systems and provided
specific criteria by which overall fairness could be evaluated. Studies four, five and six
which were a mixture of both qualitative and quantitative approaches, tested the fairness
heuristic under differing political and social circumstances was well as over time. As a
result, Syme et al., (1999) confidently make the following assertions with regard to water

allocation planning in Australia:

0 A large portion of pe0p1e believe in the rights of the environment and its
preservation for a range of uses for future generations.

0 Fair decision-making processes are of paramount importance to community
acceptance of water allocation decisions.

0 Water markets alone are not considered fair or acceptable processes for allocating
or re-allocating water.

0 Economic arguments are of a lesser importance to process considerations when
deciding how water should be allocated or re-allocated.

0 Efficiency of use is a major determinant of the fairness of water allocation
systems.

The lessons learned from studying water allocation in Australia are more relevant
in the United States today than ever before. This is because the Environmental Protection
Agency and its designated state agencies have adopted a decentralized approach to
watershed management (USEPA, 1997). Under this approach, public participation and
deliberative decision-making procedures allow for strategies to emerge as a result of a
process which brings to the table all relevant social, environmental and economic matters
(USEPA, 1997). The traditional institutional lines become blurred with new formal and
informal linkages being established depending upon the requirements of the intended
actions (Burroughs, 1999). The aim is to create local learning organizations that have the
ability to adapt to ongoing changes presented by their operating environment (Bawden,
1995)

Clearly if the EPA approach is to succeed, the process must balance the needs of
multiple users and uses in planning efforts and it should maximize acceptance of
decisions through perceptions of having been treated 'fairly'. Conflicts in values have
increased as utilitarian views are being replaced with a more environmental orientation

(Brown & Harris, 1992; Dunlap & Van Liere, 1978; Eckersley, 1992; Stem, Dietz, Kalof,

& Guagnano, 1995). Gallup data (Gallup, 1989) for example, show that 75% of the
population now claims to be environmentalist. Researchers attribute this shift in value
orientation to p0pulation growth (Manfrado & Zinn, 1996) and changes in the nation's
demographics (Steel, List, & Shindler, 1994). For example, younger, more educated,
urban dwellers tend to de-emphasize traditional uses of forests (e. g., logging, mining,
grazing) and place higher values on issues such as wildlife preservation. There is little
reason to assume that watershed issues would be any different.

Because these trends are likely to continue into the future, it is important for
managers to better understand the implication that diversely held values can have on
planning initiatives. Theory predicts that more general value orientations affect attitudes
regarding specific objects and/or situations and attitudes, in turn, influence behavior
(Ajzen & Fishbein, 1977). As was just presented, Syme’s (l993,l996,1999) work
suggests attitudes are moderated by fairness evaluations. Never the less, several recent
studies stress the importance of environmentally centered values (including fairness) in
guiding policy and management actions (Dunlap & Scarce, 1991; Kempton et al., 1995;
Satterfield & Gregory, 1998). It has been argued by Gregory, Keeney, and von
Winterfeldt (1992) that it is essential for environmental managers to listen closely to the
concerns, fears, wisdom and preferences voiced by the people who might potentially be
affected by their actions. Also, a substantial body of evidence supports the notion that
experts and laypersons often view the world quite differently (Slovic, 1987). The real
question then becomes how do we respond to these differences in both the planning

process and subsequent outcomes?

Although the importance of including human factors in water resources planning
has been discussed at length, a surprisingly limited amount of research has emerged
within the context of American watersheds. As a result, a critical need exists to define
and describe the human domain surrounding planning within American watersheds.
Iogically, little can be done towards incorporating human factors into the planning-
decision process without a clear delineation of the applicable factors, a description of
how the factors relate to each other and an assessment of the relative values with which

the factors are held.

Theoretical Perspective

This research uses an attitudinal model to guide the exploration of the
relationships between the antecedents variables for the intent to support the
implementation of BMPs. Based upon Hines, Hungerford, and Tomera's (1986)
hypothesized model the antecedents of ‘knowledge of environmental issues’, ‘knowledge
of action strategies’, ‘locus of control’, ‘attitude’ and ‘personal responsibility’ are used to
predict behavioral intentions towards proposed BMPs. This model is consistent with
Azjen’s "Theory of Planned Behavior" (Ajzen, 1985), one of the most cited behavioral
theories. According to his theory, intention to act has a direct effect on behavior and can
be predicted by attitude, subjective norms and perceived behavior control. Individual
values underlie both a person’s worldview and their attitudes. Therefore, people
possessing similar attitudes within a p0pulation (i.e., they have the same general attitude
towards an object) are ofien said to have the same worldview (Dake, 1991, 1992) and

would likely have many shared values.

Hines, Hungerford and Tomera's (1986) meta-analysis of variables related to
environmental behavior can be categorized into cognitive, affect and situational factors.
The cognitive factor, that is the awareness level about the object, is related to knowledge
of the environment including action skills and strategies. The affect variables, or feelings
and emotions associated with the objects, are generally defined by attitude, locus of
control and responsibility. Situational factors such as economic and social constraints
and/or pressures and opportunities to choose different actions either counter act or
strengthen the cognitive and affect factors.

Several researchers have suggested that an individual’s attitude towards the
management of a resource is moderated by their perception of the fairness of the process
as well as outcomes of the proposed decisions (Peterson, 1994; Syme & Nancarrow,
1997b; Syme etal., 1999). One research instrument designed to capture these interactions
is the fairness heuristic. The heuristic provides a range of fairness criteria that may
contribute to stakeholders overall fairness evaluations, making it possible to understand
the elements of a community's view on the appropriate basis for decision-making. The
fairness heuristic has been successfirlly used in the area of water allocation in Australia.
Its mix of attitudes towards water, planning attitudes and lay philosophies make it an
apprOpriate theoretical foundation on which to base research in American watersheds.

The concept of procedural justice and the demonstration of its significance gained
prominence in Lind and Tyler's seminal book - The Social Psychology of Procedural
Justice (Lind & Tyler, 1988). Procedural justice research focuses on the characteristics of
a decision-making process which make it seem ‘just’ to people vulnerable to the

consequences of a decision (Rasinski, 1987). General dimensions of procedural justice

10

such as voice or the feeling that one has had the opportunity to influence the process have
been demonstrated and replicated (Axelrod, 1994). The major hypothesis of procedural
justice is that if procedural justice is demonstrated in a decision-making process, the
outcome is more likely to be accepted (Axelrod, 1994; Lind & Tyler, 1988; Rasinski,
1987; Syme et al., 1999; Thibaut & Walker, 1975). Distributive justice as a concept
relates to the evaluation of whether an outcome was just in terms of the distribution of
resource between stakeholders. In this way, equity and distributive justice are closely
related. The dimensions of equity seem to be the bases on which individuals assess
whether or not distributive justice has been achieved.

One of the purposes of this study is to better understand the groups of respondents
holding different viewpoints about implementing BMPs within Sycamore Creek
watershed in Ingham County, Michigan. Consequently, there is a need to provide a
theoretical foundation for these groupings; several theories are available, but one based
on people’s worldviews about the environment seems appropriate. The relational pattern
of worldviews put forth by Karl Dake (1992) is rooted in cultural theory. He describes
five patterns of interpersonal relationships surrounding the perceptions of environmental
risk: hierarchical, individual, egalitarian, fatalist and autonomous. These relational forms
are hypothesized to engender shared representations of how the environment is viewed
(Douglas, 1985; Douglas & Wildavsky, 1982) and will be used to guide analysis.

The planning literature is replete with examples of framing the perceptions of a
problem differently (Messick, 1993). The framework ranges from the general to the more
specific where the outcome is likely to impact either the community or the individual.

These two terms have been differentiated in the literature by the terms universal fairness

ll

and situational fairness (Lind & Tyler, 1988; Syme & Nancarrow, 1996). The values that
contribute to these overall judgments are labeled fairness criteria.

The varied and diverse nature of BMPs and how they are implemented allow for
an exploration of different policy options. If it is assumed that policy falls into one of six
categories: 1) regulation, 2) taxation and other charges, 3) subsidies, 4) market
mechanisms, 5) human rights and 6) voluntary, then it is conceivable to create a BMP
research instrument depicting each of these. By first educating the respondent both
verbally and visually about BMPs, a frame of reference will be created. The idea of
flaming an approach to environmental problem solving was reviewed by Bardwell
(1991). She drew together concepts from cognitive psychology and conflict management
to focus upon the process of problem definition. Problem framing refers to “a concerted
effort to focus on one’s understanding of a problem” (Bardwell, 1991, p. 607). The
framing concepts in this study are defined the following way:

0 A frame of reference is an analytical model of values concerning a specific water
resource policy or management issue.

0 A personal frame of reference refers to the values expressed by an individual.

0 A common frame of reference refers to the distinctive pattern of values common to a
number of individuals.

Framing the policy choices for BMP implementation was chosen in an effort to
control for two of the three boundary conditions that Fishbein and Ajzen (1975; Aj zen &
Fishbein, 1977) identified as being able to affect the magnitude of the relationship
between attitudes and behavior. Particularly, the use of a frame of reference should

increase the degree to which the measure of intention and the behavioral criterion

12

correspond with respect to their levels of specificity and it can control for the degree to
which carrying out the intention is under the volitional control of the individual.
Previous research indicates that the principles at the situational level are not
constructed independently from those at the universal level (Ajzen & Fishbein, 1977;
Axelrod, 1994). Just how universal and situational fairness criteria change and interact
when people are framing their fairness judgments is not yet totally clear. However, it
may be that the more urgent the situation and the greater the need for the short-term
actions to achieve long-term sustainability, the more that situational fairness may

dominate (Nancarrow, Smith, & Syme, 1996-7).

Purpose Statement
The primary purpose of the study is to assess the relationships between the

antecedents of knowledge, attitudes, locus of control and sense of responsibility on the
intention to support the implementation of watershed best management practices. Also, it
will examine the role of individual fairness judgments towards developing watershed
policies. Based on these variables, the study population will be categorized and
described. Lastly, the study will determine which variables influence the intention to

support the implementation of watershed best management practices.

Research Questions

This study addresses the following research questions:

Q1: What attitudinal groups exist in the community surrounding the implementation
of BMPs in the Sycamore Creek watershed?

Q2: How do the demographic characteristics vary by attitudinal group?

Q3: What are the “fairness evaluations” regarding the implementation of BMPs for the
sample population and for each attitudinal group?

13

Q4:

Q5:

Q6:

Q7:
Q8:
Q9:

Q10:

What is the relationship between "personal responsibility" and the attitudinal
groups?

What is the relationship between "knowledge of the issues" and the attitudinal
groups?

What is the relationship between "knowledge of action strategies" and the
attitudinal groups?

What is the relationship between "locus of control" and the attitudinal groups?
Do the attitudinal groups differ in terms of intention to act?

What are the correlations between knowledge, personal responsibility, locus of
control and intent to act?

Which of the variables of knowledge, attitudes, personal responsibility and locus
of control have the greatest influence on one’s intention to act? (see Figure 1)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Knowledge of
Action Strategies
Knowledge
of Issues
Intention to
H Act
Attitudes
Social and
Locus 0f Personality
COHtI'OI Factors
Personal /
Responsibility

 

 

 

 

Figure 1: Proposed Model of Responsible Environmental Behavioral Intentions

Based on Hines, Hungerford and Tomera (1986)

14

 

Delimitations

The study was delimited to a sample of property owners within the Sycamore
Creek watershed, Ingham County, Michigan. The restriction of property ownership
further delimited the study to individuals eighteen years old and above possessing the

income and desire necessary to own property.

Definitions

The following terms are defined for the purpose of this study. For many of the
terms, a more operational explanation is detailed in Chapter III.

Attitude: describes the individual’s feelings, pro or con, favorable or unfavorable,
with regard to particular aspects of the environment (Hines et al., 1986; Newhouse,
1990). Ajzen & Fishbein (1980) suggested that attitude categories include attitudes
towards objects as well as more specific attitudes towards certain issues or attitudes
toward taking action.

Best management practices: can be defined as "schedules of activities, prohibitions of
practices, maintenance procedures, and structural and/or managerial practices, that when
used singly or in combination, prevent or reduce the release of pollutants to waters. .."
(MRSC, 2000).

Distributive Justice: distribution of outcomes on either an equal or equitable basis
(Lind & Tyler, 1988).

Equity: the value that people should receive retums appropriate to their contribution
(Lind & Tyler, 1988).

Equality: the value that all people should receive the same return (Lind & Tyler,

1988).

15

Intention to Act: behavioral intention is indicated by a person’s subjective perception
and report of the probability that s/he will perform the behavior in question (Parcels,
1984). Intended behavior is used as a substitute for actual behavior, but may not be as an
accurate predictor (Hwang, Kim, & Jeng, 2000).

Knowledge: environmental knowledge can be categorized into three levels, (1)
knowledge about the issues, (2) knowledge about the action strategies and (3) possessing

an action skill (Boerschig & DeYoung, 1993; Hines et al., 1986).

Locus of Control: is a construct that refers to an individual's beliefs about whether the
outcomes of his/her actions are dependent on what his/her do (internal control
orientation) or are determined by events outside his/her personal control (external control

orientation) (Rotter, 1966).

Personal Responsibility: means a personal obligation or sense of duty to implement
actions (Boerschig & DeYoung, 1993).

Procedural Justice: procedures consistent with personal values and that show dignity
and respect for participants (Lind & Tyler, 1988; Tyler, 1988).

Values: are standards or criteria that guide action as well as other psychological
phenomena such as attitudes, judgments and attributions (Rokeach, 1979). Values are
considered deeper and more stable than attitudes, representing standards of "oughts and
shoulds" and are viewed as determinants of attitudes (Rokeach, 1979, p. 272).

Value Orientations: clusters of interrelated fundamental values (Schwartz, 1992;

Stem et al., 1995).

I6

Worldviews: defined as shared beliefs and values that justify different ways of
behaving with corresponding cultural biases towards different patterns of social relations

(Dake, 1991,1992).

Organization of Study

The presentation of this research is organized into five chapters. Subsequent to
this introductory chapter is a review of the literature germane to the research. The chapter
is divided into two main sections. The first section reviews the theoretical underpinnings
while the second focuses on specifying the relationships between the model components
through a review of previous research. The following topics are discussed in the first
section: values and their role in decision-making; value orientations and worldviews; and
the development of the fairness heuristic as an attitudinal measure. In section II the topics
discussed are: the theories of reasoned action (TRA) and planned behavior (TPB), the
behavior — behavioral intention relationship, attitudes, personal responsibility, locus of
control, knowledge, the influence of demographic variables and predicting and explaining
intentions and behavior using TRA & TPB. The third chapter outlines the steps taken in
formulating the survey instrument, how the survey was conducted and the methods used
for the analysis. Chapter IV presents the attitudinal factors found in the sample
population, it profiles the attitudinal groups who poses similar worldviews, it explores the
fairness evaluations towards implementing BMPs, it explores the relationships between
each of the study variables and it determines which variables best predict behavioral
intentions. Finally, Chapter V contains the discussion and conclusions, the implications
of this research to Sycamore Creek watershed managers and recommendations for further

research.

17

CHAPTER II:

REVIEW OF THE LITERATURE

The purpose of this chapter is to review the literature on the theory of
environmental decision-making and the antecedents of knowledge, attitudes, locus of
control and personal responsibility thought to influence these decisions. Several
researchers (Peterson, 1994; Rasinski, 1987; Syme et al., 1999) have hypothesized that
fairness evaluations mediate the relationship between these antecedents and the intention
to act environmentally responsible. Therefore, this chapter also reviews the literature on
procedural and distributive justice in environmental decision-making and is divided into
two main sections. Part I reviews the psychological theory on which the research is based
and is broken down into five sections: (1) values, (2) the role of values in decision-
making, (3) value orientations, (4) value orientations and the social construction of
worldviews, (5) the evolution of the fairness heuristic as an attitudinal measure and (6)
basic categories of environmental policy and implied justice considerations. Part II
reviews previous research employing the same study variables so that a better
understanding of the relationships between the model variables can be understood. Part II
is divided into eight sections: (1) the theories of reasoned action (TRA) and planned
behavior (TPB), (2) the behavioral intention - behavior relationship, (3) attitudes, (4)
personal responsibility, (5) locus of control, (6) knowledge, (7) demographic variables
and (8) predicting and explaining intentions and behavior using TRA & TPB. The chapter

concludes with a summary.

18

PART I

Values

Although even a cursory review of the literature on human values yields a large
number of definitions there are five features that are common to most of the value
definitions (Maslow, 1959; Rokeach, 1973; Schwartz & Bilsky, 1987). According to the
literature, values are (3) concepts or beliefs, (b) about desired end states or behaviors, (c)
transcend specific situations, (d) guide selection or evaluation of behavior and events and
(e) are ordered by relative importance (Schwartz, 1992). Values, as defined by Rokeach
(1979), are standards or criteria that guide action as well as other psychological
phenomena such as attitudes, judgments and attributions. Values are considered deeper
and more stable than attitudes, representing standards of "oughts and shoulds" (Rokeach,
1979, p. 272) and are viewed as determinants of attitudes. Schwartz (1992) added "the
primary content aspect of a value is the type of goal or motivational concern that it
expresses" (Rokeach, 1979, p. 4). Some examples include efficiency and practicality,
achievement and success, democracy, freedom, and equity, to name a few. These
definitions and examples suggest that the values people embrace are responsible for
guiding their pursuits in life.

Theory suggests that an individual's view of the environment in which he or she
lives can be organized into a hierarchy consisting of values, value orientations (i.e.,
patterns of basic beliefs), attitudes/norms, behavioral intentions and behaviors (Ball-
Rokeach, Rokeach, & Grube, 1984; Fulton, Manfedo, & Lipscomb, 1996; Homer &
Kahle, 1988; Rokeach, 197 3; Rokeach, 1979). Each of these elements build upon one

another in what has been described as an inverted pyramid (Figure 2).

l9

 

 
   
 
   
   
 
 
   
    

Numerous

Faster to Change
Peripheral

Specific to Situations

Behaviors

Behavioral Intentions

Attitudes and Norms

  

Value Orientations
(Basic Belief Patterns)

Values

Few in Number
Slow to Change
Central to Beliefs
Transcend Situations

Figure 2: The Cognitive Hierarchy

Source: Fulton et al., 1996

Values tend to be widely shared by all members of a culture and as such are
unlikely to account for much of the variability in specific attitudes and behaviors. Rather,
values are reflected in attitudes via beliefs, value orientations and attitudes. For example,
basic beliefs serve to strengthen and give meaning to fundamental values and visa versa.

Patterns of these basic beliefs create value orientations (Fulton et al., 1996).

The Role of Values in Public Decision-Making

Relatively little is known about how people make political decisions under the
stress of conflict. However, three possible explanations can be found in the literature to
explain the apparent value conflict frequently encountered in decision-making. The first

explanation gives personal values a central role in reasoning about behavioral intentions

20

(value-driven models). An example is Tetlock's (1986) value pluralism model of
ideological reasoning that argues pe0ple reason in increasingly integrative complex ways
about an issue to the extent it invokes values that are both highly and equally cherished.
The second hypothesis is that personal values are mediated and therefore given a
secondary role. The example here is Lind's (1992) fairness heuristic model that argues an
assessment of ‘fairness’ mediates the relationship between values and support for policy
positions. The third set of explanations is the affect hypotheses that do not allow any role
for personal values in reasoning on highly emotional issues. These theories (Jackman,
1978; Schuman et al., 1985; Sears and Kinder, 1971) argue that people take positions

based on affect, not cognition like values. Let us look closer at each of these theories.

The Value Pluralism Approach

The value pluralism approach to explaining value conflict that may result in
cognitive dissonance (Festinger, 1962) is as follows: policy analysts frequently argue that
efforts to achieve one objective (6. g. equality) often require sacrificing or seriously
compromising other important objectives (e. g. merit). For example, policies designed to
manage natural resources often have the unwanted side effect of decreasing personal
income. And policies designed to increase economic growth and efficiency ofien
exacerbate income inequalities. In brief, making public policy choices requires making
value t1adeoffs(Sniderman, Brody, & Kuklinski, 1991; Sniderrnan & Tetlock, 1986b;
Tetlock, 1986). To support a particular policy means trading fulfillment of one cherished
value for another. For example, Tetlock (1986) found that when subjects ranked
‘equality’ higher than ‘a comfortable life’ as a personal value this was highly correlated (r

= .61) with whether the subjects supported paying higher taxes to assist the poor.

21

Subjects believed that to achieve equality for the poor meant forfeiting personal income
vis—a-vis higher taxes. Feather (1979) also found a significant correlation between value
rankings and public policy choices. Using the Rokeach (1979) value survey, he found
that subjects who supported politically conservative policies also supported particular
values (e.g., national security, cleanliness, obedience and salvation) over others (e. g.,
equality, freedom, love and pleasure). Value differences therefore underlie policy
choices.

In an effort to explain when people will acknowledge that important values are
indeed in conflict, Telock (1986) proposed a value pluralism model of ideological
reasoning. The model can be summarized in the three following propositions. First, all
ideologies have underlying core or terminal values (Lane, 1973; Rokeach, 1973, 1979)
that direct people's public policy preferences by specifying what the goals ought to be of
the public policy. Second, people differ in the degree to which they acknowledge their
core values are in conflict with one another. People with monastic ideologies believe their
values all point in one policy direction while people with pluralistic ideologies
acknowledge their core values sometimes point in conflicting directions. And thirdly,
people with more pluralistic ideologies display more integrative or "trade-ofi" reasoning
than those with monistic ideologies (Peterson, 1994). In short, value pluralism is
increased to the extent that core values are both more highly and equally prized, thus

setting the frame for decision-making (Barnberg, Kuhnel, & Schmidt, 1999).
Values are Mediated - The Fairness Heuristic

Lind et al., (1995) argue that people use subjective assessments of fairness, an

organizing principle that is derived in part from personal values (Rasinski, 1987), as an

22

organizing schema in social and organizational decision—making. In particular, they
hypothesized the Operation of a ‘fairness heuristic’ which suggests peOple form
impressions of the general fairness of an organization, authority or policy and use that as
a major criterion for support or opposition to the policy (Lind, 1992). People assess
fairness when encountering a new situation because they are suspicious of the intent or
result of a policy. Questions like “Do programs designed to improve a situation really
help the people whom they were intended to assist?” are ofien asked. Fairness here is a
judgment. The judgment is defined as what "is" rather than what "ought to be" as a value
judgment.

In the initial assessment stage, people are highly attuned to cues about a policy's
fairness including the assessment of how well a procedure shows respect and dignity for
the participants (Lind & Tyler, 1988; Tyler & Lind, 1992) and if it is congruent with
ones’ own personal values (Rasinski, 1987). Procedures that are consistent with personal
values and show dignity and respect for participants are considered fair and are likely to
be supported. Policies inconsistent with personal values and not showing dignity and
respect for those involved are likely to be opposed. Once an impression of fairness is
produced, it becomes extremely resistant to change because it provides a cognitively
available summary judgment (Peterson, 1994). People use their summary fairness
judgment in lieu of a more complicated policy analysis each time they are asked.
Therefore, the fairness heuristic posits that values play a role in developing policy
positions, if only a secondary one.

Both correlation and experimental evidence have supported the fairness heuristic

hypothesis. In a study assessing litigant reactions involving arbitration in Federal Court

23

cases, Lind et a1. (1995) found that decisions of whether to accept or reject an arbitrator's
award was most strongly related to people's procedural justice judgments. Results
revealed that judgments of procedural fairness mediated the effects of outcome
evaluations. People who thought the process was fair were more likely to accept the
mediation award, regardless of the outcome (Lind and Tyler, 1988; Tyler and Lind,
1992). Procedural justice judgments were better predictors of acceptance of a mediation
award than either subjective or objective measures of the award itself (Lind and Tyler,
1988; Tyler and Lind, 1992). Finally, results showed that use of the fairness heuristic was
not limited to individuals; corporate decision makers (some of the litigants represented
corporations as well) similarly indicated the use of a fairness heuristic.

In another field study of court-annexed arbitration hearings in New Jersey State
Courts, (MacCoun, Lind, Hensler, Bryant, & Ebener, 1988) found similar results.
Participants' assessments of fairness were strongly related to their perceptions of
procedural fairness (i.e. neutrality of the arbitrator and a process that grants full status to
those involved). Again, respondents who rated the process they experienced as more fair
also were more likely to accept awards of the arbiter and were more satisfied with the

final outcome of their case than those who rated their experiences as fair.

Affect Explanations - No Role for Values

The insincerity and minimalism theories are both affect explanations for the role
of values in decision-making. These two theories posit a strikingly different role for
values in people's support (or lack there of) for public policy. In particular, researchers
(Jackman, 1978; Schumann, Stech, & Bobo, 1985; Sears & Kinder, 1971; Sidanius &

Pratto, 1993) have noted the paradox between support for racial equality in principle and

24

opposition to government programs such as affirmative action to achieve those ends.
Such ‘slippage’ between espoused values and public policy preferences is known as the
“principle-policy puzzle” (Sniderrnan & Tetlock, 1986a) and is a common finding in
surveys of political issues. In several arenas people's policy preferences are out of step
with at least one of their expressed values. Affect theorists argue that people's espoused
values are often inconsistent with policy preferences because policy preferences are
dictated by deep-seated, self-centered feelings and not rational thought (Peterson, 1994).

Proponents of the insincerity theory (Jackman, 1978) argue that regardless of
what they say, people are not committed to the liberal values they advocate. It is argued
that many Americans, particularly those well educated and most likely to purport liberal
values, provide lip service to egalitarian values because those values are socially
desirable when in fact, they support traditional, self-interest driven public policies.
McConahay (1986) goes so far as to say that those who support traditional or
conservative policies are using that as a cover for racist values. This line of argument
usually reduces to an impression management explanation where peOple firmish
egalitarian responses to value questions primarily to impress others.

Alternatively, the minimalist argument theorize there are only loose linkages
between values and policy positions which create inconsistency between values and
policy positions (Converse, 1964). People have only a very hazy notion of how their
values should translate into their support/opposition of policy. As a result, where strong
affect is aroused, it can overwhelm the role of personal values and dominate policy
preferences. Again using the example of affirmative action, many white Americans

experience discomfort in the presence of Afiican Americans (Crosby, Bromley, & Saxe,

25

1980; Katz, 1976). Therefore, when reasoning about race issues if someone feels
favorable towards African Americans, they favor affirmative action, but because many
whites are uncomfortable around African Americans they would oppose affirmative
action.

The insincerity and minimalism theories have quite different implications for
understanding attitudes about policy support. Belief in minimalism leads one to the
conclusion that people need to be educated about the implications of the values they hold.
Specifically in the case of affirmative action, it would lead in the direction of clarifying
the implications of the egalitarian values to the mass public. Conversely, the insincerity
hypothesis contends that education is a primary cause of the paradox of simultaneously
supporting egalitarian values and opposing affirmative action. White, Anglo-Saxon
public policy preferences are really driven by a desire to retain a privileged position in
society (Sidanius and Pratto, 1993). Unfortunately, it is quite difficult to distinguish
whether people are engaging in impression management or in intra-psychic conflict
because in most situations they are empirically indistinguishable (Peterson, 1994).
Empirical support for one of these theories can be interpreted as support for the other as
well.

The structure of human values is inherently interesting to many researchers
(Bazerman, Messick, Tenbrunsel, & Wade - Benzoni, 1997; Feather, 1979; Rasinski,
1987). Schwartz and Bilsky (1987) see three main benefits to focusing on values for
research purposes. First, the impacts of values as independent variables on both attitudes
and behaviors can be predicted, identified and interpreted more effectively and reliably

by using indices of the value orientations as opposed to single values (Ajzen & Fishbein,

26

1980; Rokeach, 1979; Schwartz & Bilsky, 1987). Second, the effects of social structural
variables (i.e. economic, political, religious, ethnic) on values as dependent variables can
be predicted, identified and interpreted more effectively by using value orientations as
opposed to single values (Ajzen & Fishbein, 1980; Rokeach, 1979; Schwartz & Bilsky,
1987). And lastly, cross-cultural studies seem to indicate that comparisons of value
importance are more comprehensive if value orientations are used because the
orientations will ideally cover all the significant types of value content whose meanings
are shared, where as research not guided by a concept of value structure must rely on
single values arbitrarily chosen by the researcher (Ajzen & Fishbein, 1980; Rokeach,

1979; Schwartz & Bilsky, 1987).

Value Orientations

Much of what has been written about incorporating values into environmental
management decision-making comes from the literature on economics. Contingent
valuation (CV) and its variants have strived since the early 1980’s to capture how people
value environmental goods and attach dollar amounts to these values. The prevailing
practice is to take maximum willingness to pay (WTP) as the measure of the value of a
good for an individual. This purchase model is the theoretical foundation for the CV
method and the value of a public good to the public is estimated by surveys in which
respondents state their willingness to pay for the good (Bazerman et al., 1997; Carson,
Louviere, & Anderson, 1994). This first value orientation relies on the tenants of
economic equity (Walster, Walster, & Berscheid, 1978) and utility theory (Bazerrnan et
al., 1997) and refers primarily to goals such as economic security or achievement,

material rewards and/or avoidance of economic, material or time costs. It parallels the

27

sustenance needs identified by Schwartz and Bilsky (1987) and Maslow (1959), although
it more accurately reflects the value placed on economic and material desires regardless
of one's actual need situation (Bazerman et al., 1997). This is an important distinction
because the pursuit of economic gain appears to motivate behavior well beyond a time
when physical needs are met (Bazerman et al., 1997).

Discrepancies between peoples willingness to accept payment and willingness to
pay for a public good (Bazerman et al., 1997), the high incidence of protest bids
(Jorgensen & Syme, 2000) and the possibility that people view payments as contributions
towards preservation (Bazerman et al., 1997) have all been cited as arguments that the '
contingent valuation method fails to capture the complexities surrounding how people
value the environment. Consequently, it can be surmised that economic valuation is only
one value orientation people use when making environmental decisions. Stern et al.
(1993) propose two other value orientations besides economic: (a) social requirements
and (b) universal.

The second value orientation represents the social aspects of life. It specifies
desires regarding social consequences from one's actions and includes both
belongingness and conformity drivers as well as aspects of social altruism (Schwartz &
Bilsky, 1987) and benevolence (Schwartz, 1992) motives. It postulates that the
motivation to seek belongingness and acceptance from others is a central guiding force in
decisions to act (Axelrod, 1994). Virtually all discussions on value theories have at least
one domain that involves social needs (Axelrod, 1994). It is assumed that conformity and
belongingness values, as well as part of the benevolent values, are rooted in relationship

needs and desires. Thus actions in accordance with these desires would, theoretically,

28

lead people to pursue goals such as the welfare of others (especially those close to you) as
one means of maintaining and/or enhancing one's feelings of connection with others.

Social values and desires are frequently noted in discussions of environmental
behavior. For instance, the desire for belongingness can induce people to act in a manner
consistent with valued others and the value placed on benevolence may prompt socially-
oriented people (people who place social values at the top of their hierarchy) to act
environmentally protective when they believe their actions can help minimize the plight
of other people (Axelrod, 1994). In addition, the social value orientation can be seen as
consistent with the cooperative orientation identified by Messick and McClintock (1968),
although a COOperative orientation may also overlap with the universal value domain.
Numerous studies have shown that players involved in experimental game paradigms
who have a cooperative orientation pursue outcomes that maximize gains for all players
and not just for themselves (Kramer, McClintock, & Messick, 1986).

The third motivational orientation, universal, is the most consistent with
Schwartz's (1992) universalism domain. The motivational content of this value type
involves the pursuit of self-respect garnered from making a contribution to the betterment
of the world, especially as it pertains to pursuing and attaining outcomes that correspond
with universal-type goals (e.g., equity, environmental preservation). Pursuing these goals
may in fact involve certain social or economic costs, which universally-oriented
individuals are willing to incur (Axelrod, 1994). For example, protesting the harvesting
of a section of forest may mean a loss of jobs and have no perceived social benefits, but
pe0ple may do it in response to their desire to improve environmental conditions - an

outcome consistent with a desire to act in a universalistic manner. This motivational

29

domain is most strongly reflected in those people willing to violate laws and court
injunctions and risk substantial fines in order to achieve a certain goal.

A cautionary consideration regarding the construction of this value taxonomy
should be noted. Although each domain is presumed to be an independent source of
motivation, overlapping values among the domains may occur. To illustrate, universal
goals may involve pursuing outcomes such as social justice - a goal that blurs the line
between the social and universal orientations. This confusion can be addressed by
recalling the basic motivations associated with each domain. People with a social
orientation are most concerned with maintaining and enhancing connections with others.
When applied to social issues, this goal is considered consistent with pursuing outcomes
that are believed to benefit a majority of people. People adhering to a universal value
domain embrace a contributory ethic - one which emphasizes the pursuit of a personal

conception of universal goals.

Value Orientations and the Social Construction of Worldviews

The literature on the social construction of risk can lend insight on the nexus
between value orientations and worldviews. Douglas (1975) and her colleagues argue that
conflicts over risk are best understood in terms of plural social constructions of meaning.
Competing cultures confer different meanings on situations, events, objects and
especially relationships. Their assertion is that risk perception is everywhere and always
biased by legitimized social groupings in the form of institutions embodied in everyday,
ordinary social interactions with family, friends and colleagues (Douglas, 1990; Douglas,
1986; Douglas, 1985; Douglas & Wildavsky, I982a; Douglas, 1982b; Schwartz &

Thompson, 1990). Indeed, this is true for more than just risk situations and the same

30

reasoning can be extended to most decision-making circumstances. Their reasoning
provides a useful explanation for people’s decision-making reasoning and therefore can
provide a construct for understanding them, especially as it pertains to value orientations.

In making the claim that risk perceptions are socially constructed, Douglas
(1975), Dake (1992) and others pr0pose a functional explanation why risk taking and risk
ignoring are products of the various involvements that individuals have in their social life
(Thompson, Ellis, & Wildavsky, 1990). Thus, cultural theory is a functional
interpretation of the myths of nature because risk is explained in terms of the contribution
a person’s perceptions have for maintaining a particular way of life (Dake, 1992).
Cultural theory accounts for the social construction of the environment in terms of three
linked domains that constitute a way of life: cultural biases, social relations and
behavioral strategies. Worldviews are defined as shared beliefs and values that justify
different ways of behaving with corresponding cultural biases towards different patterns
of social relations (Dake, 1991, 1992). When environmentalists blame the system for
environmental damage or when corporations proclaim a comUCOpia view of nature and
call for market controls (e.g., carbon taxes) or when bureaucratic organizations call for a
top-down management of technological hazards, these behaviors are functional because
they justify and maintain the pattern of social relations from which they arise (Dake,
1992; Thompson et al., 1990, p.104).

The social construction approach hypothesizes that identity is mediated by an
individual’s relationship to others (i.e., social and universal value orientations).
Individuals who identify with collectives that make decisions binding on all members

"will see themselves very differently than those who have weaker ties with others and

31

therefore tend to make decisions that only bind themselves" (Schwartz & Thompson,
1990, p.6). Cultural theory also maintains that identity is shaped by a second factor,
namely the extent that social prescriptions constrain a person’s behavior (Dake, 1992).
According to this ‘grid/group’ nomenclature, as cultural theory is sometimes called,
social prescriptions (the grid dimensions) and group identity (the group dimension) give
rise to distinctive myths of nature and specific types of rationality. This taxonomy
elegantly captures the value domains previously described and orders them into a
manageable number of ways of life. Specifically, decision-making strategies and how
people manage themselves are reduced to five basic patterns of life: hierarchical,
individualist, egalitarian, fatalist and autonomous (Rayner, 1986; Thompson, 1988;
Thompson & James, 1989; Wynne, 1989). These relational forms, together with the
cultural biases that justify them, are each hypothesized to engender shared representations
of what does and does not constitute a management perspective (Douglas, 1975). Put
another way, "adherence to a certain pattern of social relationships generates a distinctive
way of looking at the world; adherence to a certain worldview Iegitimizes a
corresponding type of social relations" (Thompson et al., 1990, p.1). Among all possible
choices, those selected for consideration or dismissal serve (often intentionally) to
strengthen one of these cultures and weaken the others.

Hierarchically arranged groups are those stemming from high levels of stratified
prescriptions (high grid) and strong group boundaries (high group) are hypothesized to
foster the myth that nature is ‘perverse or tolerant’ (Dake, 1991). This myth holds nature
to be robust, but only up to a point. Sustainable development is the rational

environmental strategy in a hierarchical culture because this policy takes advantage of the

32

perceived resilience of nature, but respects the known limits (UNESCO, 1991). In this
worldview, the limits of the ecosystem and hence appropriate resource conservation and
development strategies, can only be proposed by certified decision makers or experts.
The analogy here is that resource management is like a traditional family life where
compliance to regulations is supposed to flow up the ranks of long-lasting institutions just
as commands flow down (Dake, 1992).

Egalitarian groups are those with strong ingroup/outgroup boundaries (high
group) but with prescriptions that do not vary by rank and station (low grid) and believe
the myth that nature is ‘fragile’ (Dake, 1991). Just as the experts-know-best approach to
resource management justifies hierarchical social relations, so the egalitarian view that
nature is ephemeral justifies the precautionary approach to management. Egalitarian
groups are critical of the procedural rationality associated with hierarchy because they
prefer approaches to management policies that foster equality of outcomes (Rayner,
1988b). Egalitarians are hypothesized to flame natural resource issues in ethical terms
because this allows them to focus on the social and political dimensions and to criticize
the institutions responsible for natural resource management (Dake, 1992). In its extreme
form, egalitarianism calls for strict preservation of the environment (Wildavsky, 1991).

The collective community that unites both the hierarchical and egalitarian defense
and protection of the environment is the antithesis of the arguments put forth when
individualist forms of social relations prevail. lndividualists are hypothesized to hold the
myth of nature as benign, so that if people are released from artificial constraints
(regulation and enforcement) there will be few limits to the abundance for all and this

will more than compensate for any hazards created in the process. Deregulation is the

33

rational management strategy in low-grid/low-group cultures because individualists value
decisions stemming from personal judgment rather than collective control (Rayner,
1988a). The term individuals in this context are social beings generating and stabilizing a
form of social relations and institutions that defend their freedom to bid and bargain in
self-regulated networks with few prescriptions (Thompson, 1992).

Cultures of fatalism are those with high levels of prescription and with minimal
collective participation and are hypothesized to hold the myth of nature as capricious.
Fatalists may be those who have been excluded from the other ways of organizing social
life; those who cannot compete successfully in markets, who cannot meet the minimum
social standards of bounded and stratified groups and who cannot assemble the time,
energy or resources required for political participation (Thompson et al., 1990). Equally
plausible, fatalists may be those who simply want to be free from the disempowerment of
influence from well-wishers (Dake, 1992). Either way, fatalists are hypothesized to
construct a cultural bias that rationalizes isolation and resignation to stringent controls on
their behavior (Mars, 1982). “Why bother?” is the rational for resource management
strategies in this high-grid/low-group culture. Fatalists are thought to view life as a lottery
in which no particular management strategy is best. Theirs’ is a capricious world, where
they desire to be left alone and stay out of harms way.

The fifth cultural pattern is autonomy - a largely asocial way of life and is not
relevant to this research. Table 1 presents the four relevant worldviews.

The literature on risk and cultural theory have empirical findings that are useful in
understanding natural resource management decision-making. One finding is that those

who hold an egalitarian bias (who value equity and lessen the distinctions between

34

Table 1: Grid/Group Nomenclature and Worldviews

 

 

 

 

 

 

Social Prescriptions
High Grid Low Grid
§~ .3 High Group Hierarchical Egalitarian
8 5 Nature is Tolerant Nature is Fragile
0 2 Low Group Fatalists Individualist
Nature is Capricious Nature is Benign

 

 

 

 

Source: Dake 1992

people based on wealth, race, gender, authority, etc.), have been found to perceive the
dangers associated with most technologies as great, and their attendant benefits as small
(Dake, 1992). Among the largest empirical correlates of egalitarianism are concerns
about environmental pollution, the dangers associated with nuclear energy and the threat
of nuclear war (Dake, 1992). To clarify, it is not that cultural theory conceives that
individualistic or hierarchical oriented people do not perceive environmental concerns or
risks, just that they disagree with egalitarians about how these concerns should be ranked.
For example, hierarchically oriented people correlate higher with concerns about the loss
of respect for authority and other forms of insubordination, while individualism is more
highly correlated with economic issues such as lack of a stable investment climate.
Furthermore, this research on worldviews replicated previous findings that measures of a
person’s worldviews are related to traditionally assessed personality traits and personal
values as well as to social attitudes and policy preferences (Buss, Craik, & Dake, 1985;
Buss, Criak, & Dake, 1986). For instance, in a. California sample, hierarchy was related
to a cautious, moderate, unassuming personality style and to a highly conservative
political orientation. Conversely, egalitarianism was associated with a less inhibited,
more expressive and assertive way of behaving and with a more liberal political

orientation (Dake, 1991).

35

The Evolution of the Fairness Heuristic as an Attitudinal Measure

Some research suggests worldviews regarding natural resources become clearer
when viewed through the lens of environmental ethics, (Wenz, 1988) and procedural and
distributive justice (Syme & Nancarrow, 1997a; Syme & Nancarrow, 1997b; Syme et al.,
1999). General worldviews become elevated to the level of an attitude when they focus
on a specific attitudinal object and are viewed through the lenses of ethics and justice
(Peterson, 1994; Syme et al., 1999). It has been shown that attitudes provide a much
better predictor of behavior than values (Ajzen & F ishbein, 1980; Vaske & Donnelly,
1999) and in an effort understand the fairness heuristic it is necessary to explore the
principles it is founded on.

Environmental ethics have received considerable academic attention, particularly
in the discipline of philosophy. Ethics can be defined as the “study or discipline which
concerns itself with judgments of approval and disapproval, the tightness or wrongness,
goodness or badness, virtue or vice, desirability or wisdom of actions, disposition, ends,
objects, or states of affairs” (Runes, 1983, p.113). Environmental ethics deal more
specifically with human conduct towards the natural environment. It is inevitable that
humans interact with the natural environment. But “What ideas govern or structure this
interaction?” and “What is the appropriate relationship between humans and nature?”
For purposes of this study, environmental ethics are defined as the diversity of ideas
driving human relationships with the natural environment (Wenz, 1988). Examples
include stewardship of nature as a religious duty and the intrinsic rights of nature. As
used in this study, environmental ethics are more focused constructs than values as they
apply to human-environmental relationships generally rather than on specific natural

objects. There is a rich literature in history, philosophy and other environmentally related

36

fields of study regarding environmental ethics. Much of this literature is reviewed in
contemporary texts and this study relies on Wenz's “Environmental Justice” (Wenz,
1988) for structuring the attitudinal independent variable.
Justice is a value judgment about the moral rightness of a person's fate (Furby,
1986). Treatment by other people or non-human forces (i.e. policies) is judged to be just
if it meets the appropriate standards of what is morally correct. These standards are
defined by supportive values. Justice supporting values differ within and across persons
and context (Seligman, Syme, & Gilchrist, 1994) and are learned from cultural relations
(Dake, 1992; Fiske & Taylor, 1990). Equity, the value that people should receive returns
appropriate to their contribution and equality, the value that all people should receive the
same return are two values that have received extensive research attention and indeed
provided the foundation for the early research in water allocation. The value of fair
. procedures has also received a share of research attention (e. g. Lind and Tyler, 1988).
The literature on equity theory and research provides four key propositions about
social behavior that are reflected in the fairness heuristic. They are:
Proposition 1: Individuals will try to maximize their outcomes. The corollary to this is
that if individuals perceive they can maximize their outcomes by behaving equitably,
they will do so. Should they perceive they can maximize their outcomes by behaving
inequitably, they will do so.
Proposition [1: Groups can maximize collective rewards by evolving accepted systems
for equitably apportioning resources (or costs) among members. Furthermore, groups will
generally reward members who treat others equitably and generally punish (increase the
costs for) members who treat others inequitably.
Proposition [11: When individuals find themselves participating in inequitable

relationships, they become distressed. The more inequitable the relationships, the more
distressed individuals feel.

37

Proposition IV: Individuals who discover they are in an inequitable relationship will
attempt to eliminate their distress by restoring equity. The greater the inequity that exists,

the more distress they feel and the harder they will try to restore equity.
(Source: Walster et al., 1978)

Imbedded in the fairness heuristic questions are these four propositions. The intention is
to elicit from the public how they feel equitable policies can be formulated while causing
the least amount of distress.

It seems appropriate at this point to review the history of ethics and justice
considerations as they have been applied to water management. Initially in the area of
water allocation, Pierce ( 1979) attempted to relate people's central values to their
priorities for water allocation for the environment. This author found that adherence to
Rokeach's (197 3) value of a ‘world of beauty’ resulted in a higher priority for water for
environmental preservation. Conversely, support for a ‘comfortable life’ tended to be
negatively associated with priorities for allocations for conservation. The relationships
were modest however, and the values very general and perhaps too broad for an
evaluation of specific allocation systems (Syme & Nancarrow, 1997a). Nevertheless,
Pierce (1979) demonstrated that it was possible to empirically assess community values
and their relationship to priorities for water allocation.

In an initial attempt to apply more specific equity and procedural justice
constructs to water allocation issues, Syme and Fenton (1993a) addressed community
perceptions of equity and procedural justice in the context of groundwater allocation in
Australia. The allocation issue included environmental as well as human uses. As an
initial step, they attempted to replicate the measurement of Rasinki's (1987)

proportionality and egalitarianism equity factors derived in the context of social welfare

38

policy. They then attempted to relate these to preferred allocation structures derived from
Thibaut and Walker's (1975) early work on procedural justice.

The results showed that Rasinski's (1987) general community factor of
egalitarianism was able to be replicated and was strongly supported by the community
sample. The concept of proportionality (or allocation based on returns appropriate to
peOple's contribution) was less well defined and supported.

Syme and Fenton's ( 1993a) work was successful in extrapolating some of the
existing equity and justice theories in water allocation, but it was evident that the
questions were too general to accurately reflect participants' views and therefore limited
in its predictive ability. There was also a need for more precise questioning in the context
of specific decision-making systems (Syme & Nancarrow, 1997a) than simply value
orientations.

Recall that Lind and Tyler’s (1988) results revealed that judgments of procedural
fairness mediated the effects of outcome evaluations. Syme and Fenton's (1993) results
would seem to support this finding and indeed concluded that while justice considerations
provided some insights in relation to water allocation, they were far from sufficient for
explaining the ethic or culture towards water management planning. Research in other
disciplines has shown that quite sophisticated lay theories Operate in a variety of domains,
such as economics and education (Fumham, 1988). Similarity, a range of philosophies
and values may exist which influence attitudes towards the allocation of natural
resources, for example the notion that nature is fragile (Bengston, 1994).

The purpose of Syme et al.’s (1999) work in allocation was to establish the

relationship, if any, between expectations of the planning process and perceptions of

39

ethics and equity as well as what philosophical questions the lay public were comfortable
with (Syme & Nancarrow, 1992). Due to the failure to adequately capture the
proportionality construct in the previously mentioned study, the philosophical questions
posed to stakeholders were widened. This approach was warranted not only because the
data seemed to indicate this need, but also because post survey interviews from the first
study also suggested that two dimensions did not adequately represent the variety of
Opinions on how to allocate water. Therefore, a wide variety of philosophical statements
(initially 1 11) were posed to stakeholders to try and better capture how people thought
about water allocation. These pertained to egalitarianism and proportionality from the
first study and statements derived from Wenz's (1988) work "Environmental Justice".
While some philosophies were harder to express in simple statements than others, the
philosophies addressed included:

Virtue Theory

Water as a Common Good

Water as an Economic Good

Free Market Philosophy: Libertarianism
Efficiency Principles

Human Rights (Kant's Categorical Imperative, Positive Human Rights)
Animal & Environmental Rights
Utilitarianism as Hedonism

Distributive Justice

Certainty and Forecasting

Cost Benefit (Willingness to Pay)

Cost Benefit (Kaldor/Hicks Formulation)
Procedural Justice (Rawls)

Hare's Cost Benefit Philosophy

Singer's Philosophy of Animal Rights

Over subsequent studies in a variety of water allocation scenarios with a range of
stakeholders, the number of statements was reduced to a manageable 25 to 30 assertions

reflecting the ways people think about water management (Syme et al., 1999). The

40

fairness heuristic recognizes that separating procedural justice aspects of decision-making
from those associated with the outcome and even subjective feelings of enjoyment from
participation are quite difficult (Folger, 1996). Therefore, this version of the fairness
heuristic reflects a better range of philosophical principles about water management that
incorporates both procedural and moral concerns as well as those matters associated with
outcomes of distributive justice. As a result Of these six studies, Syme and his fellow
researchers (1999) confidently say the following:

0 Large portions of people believe in the ‘rights of the environment’ and its
preservation for the range of uses for firture generations;

0 Fair decision-making processes are of paramount importance to community
acceptance of water allocation decisions;

0 Water markets alone are not considered fair or acceptable processes for allocating or
relocating water;

0 Economic arguments are of lesser importance to process considerations when
deciding how water should be allocated or re-allocated; and

o Efficiency of use is a major determinant of fairness of water allocation systems.

Basic Categories of Environmental Policy and Implied Justice Concerns

The justice challenge is to create policies that relate the benefits gained by
polluting activities to the disadvantages or risks caused by them. One basic principle in
ecological justice could be: equal proportions of benefits and costs from abuse of the
common resources (Montada & Kals, 2000). This principle is certainly violated in cases
of environmental racism, but more generally by every extemalization of costs caused by
poflufion.

Environmental costs are notoriously extemalized, which means that others, not

those who cause them, bear the environmental damages and their costs: single citizens

41

within the community, the community as a whole, the state and other political
jurisdictions (Montada & Kals, 2000). Externalized costs pose a serious justice problem.
Those who cause the costs receive benefits without being charged with the costs.

Inter-jurisdictional justice problems are raised when costs cross jurisdictional
borders. Legal liability norms are a means to re-intemalize extemalized costs.
Consequently, inter-jurisdictional liability norms are needed. However, application of
these norms can be problematic as long as the cause of the damage is hard to validly
identify. Furthermore, the amount and valuation of damage frequently remains open to
question.

There are six basic categories of environmental policy: (1) legal regulations, (2)
regulation by tax, (3) subsidies, (4) pollution trading, (5) the establishment of human
rights and (6) appeals to responsible actors. legal regulations aimed at the reduction of
damages and risks, if strictly applied, are an effective means of environmental protection.
At the same time they may reduce injustices in the distribution of costs and risk within
and between groups (Montada & Kals, 2000; Tenbrunsel, Wade-Benzoni, Messick, &
Bazerman, 1997). Since they are generally considered valid by society, they guarantee
more equity (Montada & Kals, 2000). However, they also have costs and there may be
losers (e. g. corporations that go bankrupt). One-sided strict jurisdictional anti-pollution
norms may interfere with inter-jurisdictional competitiveness Of industries. Moreover,
they can lead to restrictions of civil rights, which are only justifiable as a means to
prevent dangerous risks or gross injustices (Montada & Kals, 2000). A simple example of
the restriction of individual civil rights occurs when in efforts to protect water resources

municipalities dictate to residents where they may and may not walk their pets.

42

Regulations by taxes and other charges, which are revenues for communities and
the state, make production and consumption of goods that create pollution more
expensive. This is the application of the ‘polluter pays principle’. Taxes and charges are
meant to reduce the injustices of the associated benefits and costs. States or community
revenues could be used for compensation of unjust ecological disadvantages as well as
for preventative aims, such as subsidizing coo-friendly technologies or traffic systems,
etc. (Tenbrunsel et al., 1997). Still, taxes and charges do not solve every justice problem.
Those who have the resources are Often able to afford the higher taxes as in the case of
gasoline. A sharp rise in taxes may mean economic ruin for corporations as well as
private households.

The subsidizing of eco-friendly alternatives for production, traffic, air
conditioning and consumption is the third policy instrument. This supply-oriented policy
is desirable because it does not restrict freedom rights, but it is negatively evaluated
because it allocates costs to the community instead of those who have caused the
damages and risks. Therefore, a combination of the first three policy instruments may be
most effective and might at the same time prevent economic hazards (Montada & Kals,
2000).

Another market-oriented policy is the allocation of emission rights. An example is
the Clean Air Act 1990 that allows corporations a specific amount of risky emissions.
The EPA is strongly promoting the use of effluent trading to achieve water quality
objectives and standards (USEPA, 1996). Essential to this concept is the definition of
emission rights as valuable and tradable goods (Montada & Kals, 2000). These types of

policies may have a limiting as well as incentive functions to avoid emissions.

43

Unfortunately, the just allocation of emission rights is problematic. What will be the basis
for allocation? Will it be the reduced emission rate through the use of best available
technologies? or the mean emission rights for a particular business sector? or the number
of employees within a corporation? or perhaps something else? The questions are endless
as the problem becomes more complex if allocations are required between business
sectors, generations and/or jurisdictions.

The allocation of legal or constitutional rights to the natural environment might be
a powerful measure to correct and prevent ecological damages and unjust distributions
(Syme & Nancarrow, 1992; Syme & Nancarrow, 1997b) but it would also cause some
problems. One is that the various constitutional rights are not consistent but partly
conflict with each other. Ecological rights may interfere with freedom rights, with
property rights or with the right to free gainfirl economic activities (Montada & Kals,
2000). Every newly established right such as the right to a safe ecology may restrict
already established rights. Therefore, allocation of ecological rights does not provide
consensual solutions. Rather, they legitimize claims that for the present may come into
conflict not only with self-interests of other people, but also with their constitutional or
legal rights.

The last environmental policy category is appeals to responsible actors. This is
probably the most common environmental policy in watershed planning at the present
time. Appeals do not restrict freedom rights but neither do they prevent unjust
distribution of benefits and costs. The segmentation of the community that complies with
the appeals carries the burden and even contributes to the benefits of free-riding segments

that do not comply (Montada & Kals, 2000). Therefore, those actors who are principally

44

willing to comply may feel unjustly disadvantaged in comparison to the free-riders and

are demotivated when they become aware of the facts.

PART II
The Theories of Reasoned Action (TRA) and Planned Behavior (TPB)

The theory of reasoned action (Ajzen & Fishbein, 1980; Fishbein & Ajzen, 1975)
posits that behavioral intentions, which are immediate antecedents to behavior, are a
function of salient information or beliefs about the likelihood that performing a particular
behavior will lead to a specific outcome (Madden, Scholder Ellen, & Ajzen, 1992).
Fishbein and Ajzen (1975) divide the beliefs antecedent of behavioral intentions into two
conceptually distinct sets: behavioral and normative (Figure 2A). The behavioral beliefs
are postulated to be the underlying influence on an individual’s attitude toward
performing the behavior, whereas the normative beliefs influence the individual’s
subjective norm about performing that behavior (Madden et al., 1992). Hence,
information or salient beliefs affect intentions and subsequent behavior through attitudes
and/or through subjective norms (Madden et al., 1992). As noted by Fishbein and Ajzen
(1975), variables external to the model are assumed to influence intentions only to the
extent that they affect either attitudes or subjective norms.

The theory Of planned behavior (Ajzen, 1985) extends the boundary conditions of
pure volitional control specified by the theory of reasoned action. This is accomplished
by including beliefs regarding the possession of requisite resources and Opportunities for
performing a given behavior (Madden et al., 1992). The more resources and opportunities

individuals think they possess, the greater should be their perceived control over the

45

 

 

  
  

 
 
 

Behavioral
Intentions

   
 
    
 
 

Behavior

Subjective
Norms

 

 

 

(A)

 

     
 

   
   
 

 

       
   
 

  

Subjective Behavioral Behavior
Norms Intentions
;
if
erceived Actual
B eh av i 0 ral “..”-m”...- u._.‘..".......ououooou~ounusuouuuno-uu 9-00 B ehavi o ral
Control Control

 

 

 

 

(B)

 

 

 

 

Figure 3: Path Models for Theory of Reasoned Action (A) and Theory of
Planned Behavior (B)

 

Source: Ajzen, 2002

46

behavior (Ajzen, 1985). That is, when people have complete control over the behavior,
intentions alone should be sufficient to predict the behavior and perceived behavioral
control will make no significant contribution. In contrast, when a behavior is not under
complete volitional control, perceived behavioral control (to the extent that it is accurate)
provides important information that should add to the predictability of the model
(Madden et al., 1992).

Figure 23 presents the theory of planned behavior. Perceived behavioral control
is included as an exogenous variable that has both a direct effect on behavior and indirect
effect on behavior through intentions. The indirect effect is based on the assumption that
perceived behavioral control has motivational implications for behavioral intentions.
When people believe they have little control over performing the behavior because of a
lack of requisite resources, then their intentions to perform the behavior may be low even
if they have favorable attitudes and/or subjective norms concerning performance of the
behavior (Eden, 1993; Madden et al., 1992). Bandura, Adams, Hardy, and Howells
(1980) have provided empirical evidence that people’s behavior is strongly influenced by
the confidence in their ability to perform the behavior. The theory of planned behavior
specifies that for behaviors not completely under volitional control, perceived behavioral
control will add to the prediction of behavior over and above the effect of behavioral
intentions (Ajzen & Fishbein, 1977; Sutton, 1998).

The direct path from perceived behavioral control to behavior is assumed to
reflect the actual control an individual has over performing the behavior. Since the model
used for the presented research does not directly measure behavior, this link cannot be

tested and further discussion is therefore unwarranted.

47

The Behavioral Intention - Behavior Relationship

Sutton (1998) showed in his meta-analyses that when behavior was predicted
from intentions only, the product-moment correlation ranged between .44 and .62 (i.e.
explaining between 19% and 38% of the variance). In Cohn's (1992) terms, these would
all be described as medium or large effects.

There are several reasons that may cause poor predictive power; Sutton (1998)
cites nine possibilities, but of particular interest to this researcher is the principle of
correspondence (Ajzen & Fishbein, 197 7; Fishbein & Ajzen, 1975) since it can be
partially controlled for in the survey instrument. The principle of correspondence states
that in order to maximize predictive power, the predictor (attitude) and the criterion
(intentions) should be measured at the same level of specification or generality (Ajzen &
Fishbein, 1977; Ajzen & Fishbein, 1980; Sutton, 1998). The measure should be matched
with respect to four components, action, target, time and context. There is substantial
empirical evidence to support this idea (Ajzen, 1991; van den Putte, 1993) and
consequently this study uses theories surrounding environmental philosophies and justice
to guide the formulation of the independent variable based upon value orientations and

general attitudes (see environmental justice and the fairness heuristic section).

Attitudes
One of the first attempts at quantifying the strength of the relationship between

attitude and environmental behavior was Hines, Hungerford and Tomera’s (1986) meta-
analysis. For the purposes of their study, the attitudinal variable included those factors
that dealt with the individual’s feelings with regard to particular aspects of the

environment or objects related to the environment (Hines et al., 1986). Therefore, their

48

categorization of attitude included assessments of general attitude towards the
environment as well as more specific attitudes such as those towards the energy crisis,
unleaded gasoline and taking environmental action. NO distinction was made between
affective and cognitive components of attitudes.

F ifty-one outcome measures on the attitude—behavior relationship were coded.
Meta-analysis of the full set of these studies resulted in a corrected correlation coefficient
of .347 (SD = .224). Further examination of the data was conducted in an effort to
determine the nature of the attitudes under study. Consequently, forty-two of the attitude
studies coded dealt with attitudes towards the environment while nine studies were
concerned with attitudes towards taking action. A slightly stronger relationship was
detected between attitude towards action and behavior (r = .377, SD = .145) than was
observed between attitude towards the environment in general and behavior (r = .338, SD
= .243). Contrary to most studies when actual behaviors were assessed, the correlations
were higher than when behavior was self-reported (r = .427, SD = .290). Similarly when
the individuals in the studies had ties to environmental organizations, the correlations
were higher (r = .593, SD = .273).

One will note that the Hines, Hungerford and Tomera’s (1986) meta-analysis
reported on the attitude-behavior relationship and did not make the distinction between
behavioral intentions and behaviors. It is also somewhat dated. Therefore, it seems
prudent to conduct a review of more recent environmental studies and where possible to
examine issues surrounding water management.

Although used as a measure of environmental attitudes, beliefs, values and

worldviews, the New Environmental Paradigm Scale has been widely used over the past

49

two decades. It has been used most often with samples of the general public, but it has
also been used with samples of specific sectors such as farmers (Dunlap, Van Liere,
Mertig, & Jones, 2000). In general, the studies have found a relatively strong
endorsement of NEP beliefs across the various samples (i.e. correlations similar to the
Hines et al., 1986 study above). Both the ability to predict and identify groups from a
sample are forms of criterion validity. In studies of environmental interest groups, NEP
studies have consistently found that environmentalists score higher on the NEP scale than
the general public (Edgell & Nowell, 1989; Pierce, Steger, Steel, & Lovrich, 1992;
Widegren, 1998). Similarly, despite the difficulty of predicting behaviors from general
attitudes and beliefs, numerous studies have found significant relationships between the
NEP Scale and various types of behavioral intentions as well as both self-reported and
observed behaviors (Shultz & Oskamp, 1996; Stem et al., 1995). Therefore, the overall
evidence suggests that the NEP and other general attitudinal scales possess criterion
validity.

Turning toward environmental studies that focus on a specific environmental
attitudinal object (e.g., forests), it can be shown that the correlation between attitude and
intentions not only increases, but that the correlation between attitude and intention is
greater than that between attitude and measures of behavior. Also, as expected from the
general studies, these increases are especially significant if the study pOpulation has an
interest in the attitudinal Obj ect such as a hunter might have about deer. For example,
Fulton, Manfredo and Lipscomb (1996) found that attitudes mediated the relationship
between value orientations and behavioral intentions for 1,202 adult residents of

Colorado when asked about hunting/fishing and viewing of wildlife. They found the path

50

between hunting/fishing attitudes and intention to be significant (,6 = .79, t = 11.73, p _<_
.001) as was the path between wildlife viewing attitudes and intentions (,8 = .56, t = 8.84,
p _<_ .001). In both cases the paths between value orientations and intentions was not
significant.

In two other studies, one by Hrubes, Ajzen and Daigle (2001) on hunting attitudes
and behavior and one by Vaske and Donnelly (1999) on wildlands preservation, similar
results to the Fulton et al., (1996) study were found. Of the 727 people mailed a survey
from a list of those who purchased hunting licenses in Vermont in 1997, 395
predominantly white males (73%) showed a significant relationship between attitudes and
intentions (fl = .58, p 5_ .01). With regard to wildland preservation, 960 residences (53%
response rate) living along the Front Range region of Colorado also demonstrated a
significant relationship between attitudes and intentions by indicating a pro-wildland
preservation voting intention (fl = .94, p < .001, R2 = 88%).

There are just two studies on water management using the theory Of planned
behavior or theory of reasoned action. In a study on whether to adopt micro-irrigation
techniques on 44 strawberry farms in the State Of Florida, Lynne (1995) found several
interesting relations. First, it was found that perceived control was important in
explaining both the decision of whether or not to adOpt micro-irrigation techniques and
how much to invest in conservation technology. Second, it was found that the
significance of the subjective norm variable implied farmers are influenced by
community norms for water conserving behavior and that individuals who are more
influenced by the community will be more likely to adOpt and will adOpt more

intensively. The researchers maintain that the significance for perceived control, which is

51

another type of community influence, showed that coercive control (which would reduce
perceived control) could be counter productive. That is, it would not only slow the move
to becoming an adopter, but also reduce the intensity of investment in conserving
technology. They also found that actual control was significant. The fact that both
perceived behavioral control and actual control added significant explanatory power also
suggested that measuring only financial influence on investment is insufficient.

The second study by Luzar and Cosse (1998) was on Louisiana well owners’
willingness to pay for changes in state level water quality. They found that the attitudinal
variables significantly enhanced the explanatory power of their willingness to pay models
at both the state and individual levels. The state wide model improved from R2 = .08 to
R2 = .19 with the attitude variables and at the local level from R2 = .10 to R2 = .22 at the
individual level (both at a 90% level of confidence). Additionally, the subjective norm
variable was positive and significant in both models.

Several other variables were also found relevant to the Luzar and Cosse model.
Ownership of a private source of drinking water was positively associated with
willingness-tO-pay for improvements in water quality. The presence of young children in
the respondent’s household was positively and significantly associated with willingness-
to-pay for changes in water quality. Respondents with higher income were not only
willing to pay, but able to pay for changes in water quality. There was a non-linear
significant positive relationship between an individual’s age and education and
willingness-to—pay. Not statistically significant were explanatory variables indicating

occupation and gender.

52

Personal Responsibility

Geller (1995) has proposed that certain psychological states or expectancies affect
the propensity for individuals to actively care for the safety or health of others. The term
active caring refers to individuals who care enough about a particular problem or about
other people to implement an intervention process in attempts to make a beneficial
difference (Geller, 1995). By this definition, active caring is equivalent to the personal
responsibility variable in this study and is considered a social norm. Allen and Ferrand’s
(1999) study of college students’ tests Geller’s hypothesis that actively caring is an
important mediator leading to environmental concerns and action. They found no
evidence of a direct relation between personal control and environmentally friendly
behaviors. Instead, and consistent with Geller’s (1995) model, sympathy (i.e. the proxy
used for active caring) was found to mediate the relation between personal control and
total environmental behaviors. Specifically, the path from sympathy to total
environmentally friendly behavior was significant (,6 = .39, t(94) = 3.47, p = .25), but the
direct path between personal control and total environmentally friendly behavior was not
significant (,8 = .14, t(94) = 1.15, p . .25). While Allen and Ferrand’s (1999) results
support Geller’s hypothesis, Hwang et al.’s (2000) study on forest management found
only a small relationship between personal responsibility and intent to act (,8 = .20,
#436).

A parallel line of research examining the relation between altruistic social norms
and environmentally responsible behavior is also consistent with Geller’s (1995) views.
Specifically, Herberlein (1972) maintained that protecting the environment is perceived
as a moral and altruistic issue because environmental damage has negative consequences

for others. As a result, Herberlein (1972) suggested that Schwartz’s (1970; 1977) moral

53

norm activation model of altruism could be used to predict environmentally friendly
behaviors. Schwartz’s model predicts that altruistic behavior results when a moral norm
is activated. This activation occurs when an individual becomes aware that their behavior
has possible negative consequences for others and is willing to take personal
responsibility for the other’s well being. Therefore, Herberlein (1972) suggested that
individuals should act to protect the environment in situations similar to those that elicit
altruism: when an individual is both aware that their actions can have negative
consequences for the environment and feels personally responsible for these
consequences.

Several authors (e. g. Guagnano, 1995; Hooper & Neilsen, 1991; Van Liere &
Dunlap, 1978) have acted on Heberlein’s suggestion and have demonstrated that
Schwartz’s (1970, 1977) model can be used to predict environmentally fiiendly
behavior. For example, Van Liere and Dunlap (1978) found that individual rural residents
who were both aware of the negative environmental consequences of burning yard waste
and who also accepted responsibility for the environmental consequences of burning
were least likely to burn yard waste. Similarly, but more globally, Guagnano (1995)
found that residents’ awareness of the environmental consequences of their actions and
their willingness to take responsibility for these actions combined to predict willingness
to take action to protect the environment on a range of issues. These findings are
consistent with Geller’s model in that they demonstrate environmentally friendly

behaviors are at least partly a function of altruism.

54

Locus of Control

The locus of control (LOC) factor in the model represents one of the most
extensively researched variables. Locus of Control refers to a general belief regarding the
location of forces controlling an individual’s life (i.e. internal vs. external factors).
Persons with an internal locus of control believe they normally have personal control
over important life events as a result of their knowledge, skill and abilities (Geller, 1995).
In contrast, people with an external locus of control believe factors like luck, chance or
fate have significant influence in their lives (Rotter, 1966). Therefore, people with an
external LOC generally expect to have less personal control over the pleasant and
unpleasant consequences they experience than do people with an internal LOC. The use
of LOC for the proposed model of community action would appear justified since it
encompasses both external and internal components of an individual.

There exists an intimate relationship between LOC and personal responsibility
(Eden, 1993). This is because individuals focus on actionable responsibility, not solely on
moral responsibility (Eden, 1993). To be actionable, an individual must perceive the
internal control they can maintain over the outcomes of their behavior. That is, they have
a strong belief in their own efficacy. Therefore, the identification of pro-environmental
behavior is not only based upon it being responsible, but also possible and efficacious.

Actionable responsibility is curtailed by its context in that behavior is only
morally required within its efficacious range, which is usually an immediate one. Rather
than suggesting mere self-interest, the immediacy of responsibility points to individuals
being necessarily preoccupied with immediate issues (O'Riordan & Rayner, 1991).
Beyond the immediate, efficacy is weakened which in turn weakens perceived

responsibility (Eden, 1993). It becomes pointless to ascribe responsibility to the self to

55

undertake behavior that has no effect. There is an implicit utilitarian ethos to this
reasoning in that responsibility is not limited because of morality alone, but functionally
because of its efficacy in curtailing environmental damage (Eden, 1993). This supports
Schwartz’s (1970) emphasis on awareness of the consequences (see below) of a person’s
actions as contributing towards responsible behavior. Therefore, context is influential in
that it affects how efficacy is perceived. Efficacy is reinforced (or not) by societal norms
underlying the need for setting individual pro-environmental behaviors within social
circumstances (Eden, 1993).

In recent years, there have been numerous studies aimed at exploring the
relationship between locus of control and environmentally responsible behavior (Hines et
al., 1986; Sia, Hungerford, & Tommera, 1985/86; Smith-Sebasto, 1992). As a result of
this work, it has been suggested that instruments created to measure LOC as it relates to a
specific situation or behavior should yield more reliable results than a generic scale
(Lefcourt, 1982; Phares, 1976; Rotter, 1975). This would suggest if researchers wish to
create a locus of control instrument that would allow for precise predictions of
environmentally responsible behavior, they should construct items with referents to
specific environmentally responsible behaviors or actions (Smith-Sebasto & Fortner,
1994). Therefore, the instrument designed for this research was adapted from the
Environmental Action Internal Control Index (EAIC) (Smith-Sebasto & Fortner, 1994) to
the context of BMP implementation for watershed management.

The EAICI was chosen because evaluation of previous instruments concluded that
their internal consistency and validity were questionable (Smith-Sebasto & Fortner,

1994). The development of the EAICI started with the identification of examples of pro-

56

environmental behaviors from numerous published sources in the popular press. Items
that eventually became part of the instrument were those that should have a high
behavior-attitude relationship (Ajzen & Fishbein, 1977, 1980). The EAICI instrument is
comprised of 28 items of which six have been adapted to the context of watershed
management for the purposes of this study.

The Smith-Sebasto and Fortner (1994) study of undergraduate university students
found the correlation between LOC (as measured by the EAICI) and environmentally
responsible behavior was modestly positive (r = .33) at the .01 level of significance
(Smith-Sebasto & Fortner, 1994). Through discriminant analysis they also found the
EAICI accurately classified individuals self-reported environmental behavior in almost
82% of the cases. Further discriminant analysis showed that the EAICI could also
accurately classify individuals on the basis of perceived knowledge or skill at the use of

environmental action strategies (Smith-Sebasto & Fortner, 1994).

Knowledge

In natural resource management, the level of factual knowledge has been
identified as an external variable that links values and attitudes (Tarrant, Bright, &
Cordell, 1997). Although the effect of knowledge on attitudes is not conclusive, there
have been numerous studies suggesting a link between the two. For instance, Bright &
Manfredo (1997) concluded that individuals with higher knowledge levels have more
positive attitudes than those with low levels of knowledge. Fortner and Lyon (1985)
found a similar result when they studied the effects of viewing a Cousteau television
special on attitudes, although the attitudinal change appeared temporary (lasting only

approximately 2 weeks). Meanwhile, Steel et al. (1990) found that policy-relevant

57

knowledge was associated with perceptions of risk in the Great lakes for Canadians but
not Americans. There are two possible explanations for this ambiguity between studies:
(1) the appropriate knowledge component for predicting attitude has not been identified
(Boerschig & DeYoung, 1993; McKenzie-Mohr, Nemiroff, Beers, & Desmarais, 1995)
and/or (2) that context matters (Ajzen & Fishbein, 1977). Despite these inconclusive
findings, knowledge enhances the association between general environmental value
orientations (and by extension worldviews) and environmental policy preferences by
increasing the explained variance (Pierce, Lovrich, Tsurutani, & Abe, 1989; Steel et al.,
1990; Tarrant et al., 1997). Thus, it has been included in the model.

Environmental knowledge can be categorized into three levels: (1) knowledge
about the issues, (2) knowledge about the action strategy and (3) action skill (Boerschig
& DeYoung, 1993; Hines et al., 1986). Interestingly, these different levels of knowledge
are considered to influence the responsible environmental behavior in different ways.
For example, Hungerford and Volk (1990) categorized environmental behavior related
variables into categories: entry-level, ownership-level and empowerment-level. They
suggested that knowledge about general concepts is considered an important variable for
entry-level, in depth knowledge about issues for the ownership-level and knowledge
about action strategies and skills for the empowerment level. This suggests that
depending upon the context of the research, different types of knowledge should be
elicited fi'om respondents depending on the study goals.

Traditionally, researchers in the field of environmental education have claimed
they can change behavior by making peOple more knowledgeable about environmental

issues. The reasoning is that if one has more knowledge about the environment, the

58

awareness level should be higher, thus producing a more favorable attitude (Hungerford
& Volk, 1990). Individuals with a more favorable attitude in turn will be more
predisposed toward environmentally friendly behavior (see attitude section). Yet, more
environmental knowledge does not necessarily mean an increase in environmentally
responsible actions. A major difficulty in imparting environmental literacy lies in the
simple fact that research has not yet satisfactorily identified the knowledge components
that are precursors to responsible environmental behavior (Sivek & Hungerford, 1989).
Nevertheless, many researchers have used the change of knowledge as an effective
variable for explaining the change in environmentally responsible behavior (Sivek &
Hungerford, 1989; Smith-Sebasto & Fortner, 1994).

The Smith-Sabasto and Fortner (1994) study of 853 undergraduate university
students found several interesting relationships between knowledge, locus of control
(LOC) and environmentally responsible behavior. A correlation of .23 (p > .01) between
perceived knowledge of and skill at using environmental action strategies and LOC was
found, indicating a low positive relationship. A discriminant analysis was performed
between the scores on knowledge and LOC in an effort to assess whether knowledge
could predict an individual’s LOC orientation. Sixty-one percent of the grouped cases
were correctly classified. When a discriminant analysis between LOC and knowledge
was performed in an effort to predict an individual’s knowledge level from their LOC
score, eighty percent of the grouped cases were correctly classified. Hwang et al. (2000)
found a more modest correlation between knowledge and LOC (13 = .02 ). They attributed

this modest relation to their using an instrument designed to test general environmental

59

knowledge as opposed to more specific questions on knowledge or skills in their subject
matter (i.e. forest use).

Sabasto-Smith and Fortner’s (1994) study also found a moderately positive
correlation (r = .483, .01 level of significance) between knowledge and environmentally
responsible behavior, while Hines et al. (1986) found a correlation of r =. 185 in their
meta-analysis. These findings are similar to several other studies (Sia et al., 1985/86;
Sivek & Hungerford, 1989). However, the correlation between scores does not
necessarily indicate a direct relationship. In fact it is assumed by many researchers that
knowledge will influence attitude, which in turn will affect behavior (Hungerford &
Volk, 1990; Newhouse, 1990). Results from the Hwang et al. (2000) study support this.
They found a small relationship between knowledge and attitude (,6 = .09) and no
significant relation between knowledge and intent to act. As was the case previously, the
magnitude of this relationship could also have been affected by the use of a general
knowledge instrument and that result should improve by testing other levels of
knowledge such as those surrounding action skills and strategies (Boerschig & DeYoung,
1993; Hines et al., 1986). Regardless of the magnitude of the relationship, this finding
still indicates that attitude is a mediator variable between knowledge and intention to act.

Hamilton (1986), Eden (1993) and others have suggested there exists a
relationship between environmental knowledge and personal responsibility. This
relationship is also intuitive; the more one knows about a subject the more likely they
will feel a sense of personal responsibility, especially if the subject directly impacts on
them. Beckwith and Rayl (2002) found a partial correlation coefficient of. 1744 (p=.005)

between general environmental knowledge and environmental responsibility among

60

college students. Also, Hwang, et al. (2000) found a small relationship between

knowledge and personal responsibility (r = .03) in their study of forest management.

Demographic Variables

The extensive literature on environmental attitudes, age, education, urban
residence and political ideology indicates that socio-demographic factors are found to
have a consistent, substantially significant association with environmental attitude
(Christianson & Arcury, 1992; Van Liere & Dunlap, 1980). The conclusions of these
factors indicates that younger, better educated, urban, liberal individuals are more
concerned about the environment and have more positive attitudes toward the
environmental movement. Other factors that have weak or inconsistent relationships to
environmental attitude include gender, income and occupational prestige (Christianson &
Arcury, 1992; Samdahl & Robertson, 1989).

Income, education, gender (particularly male), and environmental attitude have a
consistent positive association with public environmental knowledge (Arcury & Johnson,
1987; Arcury, Johnson, & Scollay, 1986; Lovrich, Pierce, Tsurutani, & Abe, 1986; Pierce
et al., 1989). Age, being liberal and exposure to sources of information such as television
news programs are less consistently correlated with environmental knowledge. Lovrich et
al. (1986) also found that the perceived seriousness of a water problem to be positively
associated with environmental knowledge and environmental knowledge to be positively

associated with support for protective water policy measures.

61

Predicting and Explaining Intentions and Behavior Using TRA & TPB

This section summarizes the results of meta-analyses on the theories of TRA/TPB
to determine the percentage of variance explained. These reviews vary greatly in terms of
the number and type of studies included and the sophistication of the meta-analytic
methods used. Table 2 presents this summary.

Table 2: Summary of Findings From Meta-Analyses of the TRA and TPB
Regression of attitudes and social norms or

 

 

attitudes, social norms and perceived behavioral Effect Size
control on behavioral intentions
Reviewer R R2
Farley etpal. (1981) .71 .50
Sheppard et al. (1988) .66 .44
Azjen (1991) .71 .50
van den Putte (1993) .68 .46
, Conner & Arrnitage (1998) 63 .40

 

Adapted from: Sutton, 1998, p. 1320

The findings for behavioral intentions show reasonable consistency with multiple
correlations ranging from between .63 to .71. This accounts for between 40% and 50% of
the variance. For intentions, the two theories are typically explaining no more than 50%
of the variance. This seems disappointing in view of the fact that in the vast majority of
the studies, intentions and its predictors are measured at the same time on the
questionnaire using similar items; conditions that should maximize predictive power
(Sutton, 1998).

There are a number of different standards of comparison that can be used in
evaluating the percentage of variance explained. Neither the TRA nor TPB fare well by
this standard. In practice, however, the maximum percentage of variance that can be
explained in a real application is often substantially less than 100. For example Table 3
shows five effect size measures that can be calculated on a simple randomized control

experiment on a smoking intervention program.

62

Table 3: Examples of How Different Effect Size Can Give a Different

 

 

 

Impression
Randomized control trial of a new treatment for smoking
Condition 1 N l Succeed | Fall
Intervention 100 70 30
Control 100 30 70

 

Difference in success rates = 70 —30 = 40
Odds ratio = (70 x 70)/(30 x 30) = 5.4
Relative success = 70/30 = 2.3
Product-moment r (phi-coefficient) = .40
Percentage of variance grplained = 16

 

 

 

 

 

 

MnfiWN-A

 

 

Adapted from: Sutton, 1998, p. 1323

The difference measure shows that the intervention improved the success rate by
40 percentage points. The odds of successfully quitting smoking were over five times
higher in the intervention condition, compared with the control condition. The relative
success rate shows that the intervention more than doubled the chances of successfully
quitting. All of these measures suggest that the new treatment had a substantial and
clinically useful effect. However, the percentage of variance explained in the
dichotomous independent variable was 16 %, which seems unimpressive. Therefore, it
can be concluded that the 40% - 50% of variance explained in the various meta-analyses

0f TBA/TPB are indeed large effect sizes (Cohn, 1992).

Summary

The beginning of this literature review proposed a theoretical foundation for the
StUdy of attitudes and their effect on intention to act. It began with presenting a hierarchy
0fPsychological variables beginning with values and explained three theories of how
values might operate in decision-making. Progressing up the hierarchy, the next section
focused on value-orientations and beliefs and showed how these were founded in

lndlvIdUal values and were relevant to worldviews. Worldviews were shown to be

63

general attitudes and represented value orientations within a context. Four relevant
worldviews were presented. Drawing from cultural theory, the worldviews were
hierarchical (high group/high grid), egalitarian, (high group, low grid), individualists
(low, group, low grid) and fatalists (high grid, low group). The last section in part one
dealt with attitudes and the fairness heuristic. This section discussed the philosophical
tenants underlying this research instrument and presented its evolutionary development.

Part II of the literature review began by presenting the theory of planned behavior,
which was the underling theory for the study design. The theory of planned behavior
defined the scope of the research and defined the study variables. Having defined the
variables, the remainder of the literature concerned itself with reviewing the relationship
between the study variables knowledge, locus of control, personal responsibility, attitudes
and intent to act. Table 4 summarizes these findings.

Clearly, from this summary table it is apparent that the psychological variables of
attitude, locus of control and personal responsibility are the most important variables in
influencing behavior and behavioral intentions. Knowledge is important, but indirectly
and its effects on behavior and intentions appear to be mediated by the other
psychological variables. lastly, demographic variables are poor predictors of behavior

and behavioral intentions.

64

Table 4: Summary of Findings on the Strength of the Relationships Between
Psychological Variables

Variable

Beta

 

Relationships correlation Weights Study
Intention - Behavior .44 - .62 Sutton (1998)
Attitude - Behavior .35 Hines et al. (1986) — meta analysis

17 Beckwith & Rayl (2002) -

' undergraduates

. . .56 - Fulton, et al. 1996) -

Ammdes “ Imemmns .79 hunting/wildlife viewing

.18 .09 Hwang et al. (2000) — forest

.58 Hrubes et al. (2001) - hunting
.94 Vaske & Donnely (1999) - wildland
.28 _ .47 Lazar & Cosse (1998) - water
quality

Attitude - Responsible .10 Hwang et al. (2000)

.43 Beckwith & Rayl (2002)
Responsible - Intention .10 Hwang et al. (2000)
Responsible - Behavior .39 Geller (1995) — safety & health

.33 Hines et al. (1986)

44 Allen & Ferrand (1999) —

° undergraduates

.36 Beckwith & Ray] (2002)
LOC -Responsible .17 .13 Hwang et al. (2000)
LOC - Behavior .31 Allen & Ferrand (1999)

33 Smith—Sabasto (1994) -

' undergraduates

.36 Hines et al. (1986)
LOC - Intent .20 Hwang et al. (2000)

.37 Hines et al. (1986)
LOC - Attitude .39 .39 Hwang et al. (2000)
Knowledge - Behavior .05 Hwang et al. (2000)

.30 Hines et al. (1986)

.17 Beckwith & Rayl (2002)
Knowledge - Attitude .09 .09 Hwang et al. (2000)

.18 Beckwith & Rayl (2002)
Knowledge -— LOC .02 .02 Hwang et al. (2000)

.23 Smith—Sabasto (1994)
Knowledge - .03 .02 Hwang et al. (2000)

.24 Beckwith & Rayl (2002)
Education - Behavior .18 Hines et al. (1986)

.48 Smith-Sabasto (1994)
Age - Behavior -. 15 Hines et al. (1986)
Income —— Behavior .16 Hines et a1. (1986)

 

65

CHAPTER III:
METHODOLOGY

Study Design

This study attempts to discover people’s worldviews, attitudes and the intent to
support the implementation of best management practices based upon selected
antecedents. The data collection method is a mail survey. A mail survey was chosen in
order to secure a sufficient sample size to be able to infer characteristics, attitudes, and
behaviors for the entire watershed population (Babbie, 1990). There was also a need to be

able to support policy Options quantitatively.

Background
The study was conducted in Sycamore Creek, a sub-watershed of the Red Cedar

River, which is part of the central Michigan portion of the Grand River watershed. The
watershed drainage area is approximately 67,740 acres and is located in the center of
Ingham County (NRCS, 1990). The primary land use in the southern half of the
watershed is agriculture with one major population center (the City of Mason). The
northern half of the watershed covers part of each of the cities of Holt and lansing. Also,
approximately 1,500 acres of Michigan State University farmland is located in the
northern part of the watershed (NRCS, 1990).

Several problems have been identified in the Sycamore Creek watershed. The
major types of pollutants to be controlled are sediment from soil erosion, phosphorous
fertilizers, nitrate fertilizers and agricultural pesticides (NRCS, 1990). These pollutants

cause sedimentation and turbidity problems, nuisance algae growth and

66

Ingham

County

 

 

 

 

 

 

 

 

L .
D 811*me * East .
Lansing
W %

(K L; ‘W’illiamston [:33] .

‘ Webherville
Sycamore ..p
Crook }
Watershed .

WK

5‘] * P
Mason Z [El

 

 

 

 

 

 

m.

D

 

Mchigan

 

 

 

 

Figure 4: Map of Sycamore Creek Watershed in Relation to Ingham County, MI

Population and Sample

groundwater contamination (N RCS, 1990). These types of problems are frequently

addressed by the instillation of best management practices by property owners.

The study sample was drawn from Ingham County’s 2001 tax database. There are

67

21,801 properties on the tax roll within Sycamore Creek watershed. Since the
independent variable of the study is individual attitudes, the research focused only on the

16,991 residential and agricultural properties. The categories of residential and

agricultural land were determined using a code assigned to each property indicating its
land use. The categories of commercial, industrial and developable were excluded. These
required codes are standardized for the State of Michigan.

Each of the residential prOperties was assigned a computer generated random
number. It was determined through the use of a standard sample size formula (Babbie,
1990) that there needed to be 376 respondents in order to achieve a 95% confidence level
and 639 respondents to achieve a 99% confidence level. An initial mailing of a letter of
introduction and explanation consisted of 1,750 property owners. The names of those
residents whose letters were not able to be delivered were removed from the mailing list
for the first survey. A total Of 1,650 prOperty owners were mailed questionnaires with the
hope there would be sufficient respondents to achieve the 99% confidence level. The
sample was comprised of 1576 randomly selected residential properties and all 74
agriculture properties. To be included in the sample, a property owner had to reside either
within the Sycamore Creek watershed or the greater Red Cedar watershed. Due to of the
small number Of agricultural properties, there was no attempt to separate them out as a
unit of analysis.

After removing those questionnaires that remained undeliverable or were either
refusals or blank returns, there remained 1,542 usable surveys. Of these 1,542
questionnaires, 608 usable surveys were returned for an overall response rate of 39%. It
should be noted that the response rate might have been higher had not the mailing
occurred shortly after the events of September 11, 2001 and also which pushed the

study’s mailing into the holiday season.

68

Instrumentation

The questionnaire was constructed from both original research material and
components from other research instruments. Dillman’s (2000) Tailored Design Method
was used to guide both construction and administration of the mail survey. The
questionnaire had seven sections: (1) Use and Thoughts About the Red Cedar River; (2)
Water Quality Concerns; (3) Best Management Practices; (4) BMP Combinations; (5)
Choice Experiment and Implementation Information; (6) Attitudes and Beliefs; and (7)
Demographics. Only part of section 1 and sections 2, 3, 5,6 and 7 were used for this
study.

Initially, three focus groups were conducted with local experts in an effort to
understand the issues surrounding BMPs and to establish content validity. The scientific
panel, as it came to be known, was drawn from organizations such as the Michigan
Department of Environmental Quality (MDEQ), the National Resource Conservation
Service (NRCS), Michigan State University Extension (MSUE), various departments at
Michigan State University and the Soil Conservation District. The information collected
in these focus groups helped determine survey content and structure in general and more
specifically, the policy choices presented in the questionnaire (section 5).

The questionnaire was designed using Hines, Hungerford and Tomera’s (1986)
proposed model of responsible environmental behavior (Figure 1 is presented below
again for convenience). This model was chosen for two reasons: (1) it contained all the
elements required by the Theory of Planned Behavior, but (2) it was not a path analytic
model. It was felt that the lack of previous research on attitudes toward water resource
management in a specific context would make it likely that any method reliant upon the

tight definition of the factors (as needed in a path analytic model) would fail. That is,

69

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Situational
Factors
Action Skills
Knowledge of }
Action Strategies
Knowledge 7 V
of Issues I t ti Responsible
. 11 en 0“ Environmental
Attitudes —’ t0 ACt Behavior
Locus of
Control \ Personality
Factors
Personal
Responsibility

 

 

 

 

 

Figure 5: Hines, Hungerford and Tomera’s (1986) Proposed Model of Responsible
Environmental Behavior

even if the factor scales were valid and reliable in a previous context, it does not
guarantee success in a new context due to the restructuring of the factors towards a new
attitudinal object by individual respondents.

This study concerns itself with the intention to act and its antecedents.
Furthermore, since there is not an action skill required by the respondent in order to
support (or not) the implementation of BMPs by municipalities, this factor is superfluous
and not included in the model. Specific situational factors were not able to be totally

captured by the survey instrument although there was considerable effort to establish

70

context. Consequently, the final model used in this study is presented in Figure 1. The
overall questionnaire design followed the pattern of eliciting general information from
the respondent, educating them about the current problems and possible solutions and

then asking them what they wish to do about the problem.

 

 

Knowledge of
Action Strategies

 

 

 

 

 

Knowledge
of Issues

 

 

 

 

 

 

Intention to
Act

 

 

 

Attitudes \

 

 

 

 

 

 

 

 

 

 

 

 

Social and
Locus of Personality
Control Factors
Personal { /
Responsibility

 

 

 

 

 

 

Figure 1: Proposed Model of Responsible Environmental Behavioral Intentions
Based on Hines, Hungerford and Tomera (1986)

Knowledge of the Issues and of Action Strategies

An index of Knowledge of the Issues was created through summing responses to
five dichotomous, yes (I) or no (0) questions (Q6-Q10). This section contained questions
such as whether the respondent knew about stormwater overflow plans and non-point
source pollution. A Knowledge of Action Strategies index was similarly created through

summing responses to six yes (1) no (2) or don’t know (0) questions (Q11-Ql6). This

71

section contained questions on whether they knew about each of six structural BMPs
used to control stormwater runoff. The use of a dichotomous variable is consistent with
other studies measuring knowledge (Reading, Clark, & Kellert, 1994, Fortner &

Lyon, 1985 #294; Steel et al., 1990) with the reasoning being that the respondent either
possesses the knowledge or they do not and the higher the score on the index, the greater
one knows. Those respondents indicating “Don’t Know” in the knowledge of action

strategies index were included in the “NO” category for the analysis.

Attitude
Construction of the Attitude factors was developed from items contained in the

fairness heuristic instrument as reported by Syme, Nancarrow and McCreddin (1999).
The attitude section consisted of twenty closed ended questions (out of thirty one possible
questions from the original instrument). A five-point Likert agree/disagree response
format was used. Questions range from inquiring whether people believed the

“. . .environment had the same right to water as peOple” to “. . .whether those people

upstream had a moral responsibility to look after those downstream”.

Locus of Control

Several locus of control (LOC) instruments have been developed to measure the
relationship between LOC and environmentally responsible behavior. In their 1994
article, Smith-Sabasto and Fortner demonstrated the internal consistency and validity of
these instruments were questionable. Their research compared the Environmental Action
Internal Control Index (EAICI) to other LOC instruments (e. g. Brown LOC, Index

Control Index and Need for Control Scale) and found that the EAICI was more highly

72

correlated with environmentally fiiendly behavior than other instruments. Furthermore,
the correlation coefficients between the various scales studied lent support to the EAICI’s
convergent and discriminant validity. Therefore, the EAICI was chosen to assess
individual levels of locus of control on managing the Sycamore Creek watershed (Smith-
Sebasto & Fortner, 1994). Internal consistency of the original scale is reported at .92 for
Cronbach’s alpha. Of the 28 questions used in the original instrument seven were adapted

for the current study with an alpha of .89.

Personal Responsibility

The four-item Personal Responsibility scale was constructed from one two-item
scale and two additional questions. The two-item scale is derived from the Ascription of
Responsibility factor within Schwartz’s Norm Activation Scale as reported by Guagnano,
(1995). The items were originally identified by face validity and checked by conducting a
factor analysis using principal-components techniques with oblimin rotation. Item factor
loadings were both greater than .75 with an eigenvalue of 1.07 and Cronbach’s alpha of
.65. The “I am partially responsible for the degraded state of our local rivers and streams”
question was used in the Hwang et al. (2000) study and found to correlate with other
variables in their study (which were similar the ones used in this research). The final item
was constructed by the researcher and appears to have face validity. Therefore, three of
the four items had either concurrent validity or construct validity as well as face validity.

Internal consistency of personal responsibility scale used in this research was .81.

73

Intention to Act

There were seven Intention to Act statements. Their design was based on
information gathered from the scientific panel (see above) and a review of the literature.
Each statement asked the respondent to indicate their level of support for each of the
following items: zoning of open space; subsidies to landowners for environmentally
friendly practices; stricter enforcement of current regulations; creation of new
regulations; public information and education programs; and voluntary programs.
Although the seven items were created to be independent of each other it was found they

had an alpha of .88 when scaled.

Questionnaire Construction and Design

The resulting thirty multiple part questions were pre-tested with two separate
groups of individuals. The first group of twenty individuals was drawn by reaching out to
the local community and requesting volunteers. Many of the names were provided by
MSU—Extension and Communications Department. This group was not considered to be
representative of the Sycamore Creek watershed, but rather was a convenience sample
necessitated by the availability of recording facilities. Each respondent was debriefed for
twenty to thirty minutes after they had completed the survey. Questions were further
revised based upon an item analysis of the pretest and comments made by the
respondents. Revisions at this point were mostly minor, consisting of some additional
wording and layout changes. The major change was the deletion of two questions on the
creation of markets for pollution trading. Most respondents were not familiar enough
with the concept to make a judgment and the researcher felt it was too difficult a concept

for the survey to try and convey in addition to the current content.

74

The revised questionnaire was then pre-tested a second time. This time an
intercept approach was taken to selecting respondents. People entering the Michigan
Secretary of State office in Mason seeking their automobile license renewal were asked
to participate while they waited. This was intended to be a convenience sample and was
not representative of the watershed population. The main purpose of this pre-test round
was to assess face and content validity aswell as user understandability. Ten individuals
agreed to take the survey and be debriefed afierwards. Very minor changes to wording
and layout were made based upon the respondent’s comments.

The complete instrument consisted of thirty-four Likert scale questions (five
choices from strongly disagree to strongly agree) on attitudes, personal responsibility and
locus of control; twenty questions on the frequency of current uses and the importance of
firture uses; one question on perceived water quality; five items on water quality
knowledge and six on knowledge of best management practices; three choice experiment
questions; seven on implementation information, one on preferred sources of
information; and eight demographic questions. The number and complexity of the
questions were limited by the time required to take the survey. It was felt that the
questionnaire needed to be able to be completed within a twenty to thirty minute time
frame (Dilhnan, 2000).

As mentioned previously, the study roughly followed Dillrnan’s protocol for mail
survey administration (Dilhnan, 2000). Five mailings were used to administer the survey.
Seventeen hundred and fifiy randomly selected peOple were initially mailed a pre-letter
introducing the study and informing them they would be receiving a questionnaire in the

mail within the next week. The undeliverable pre-letters were deleted from the mailing

75

list for the first mailing of the survey and the list was further reduced to arrive at a total of
1,650 potential respondents. The questionnaire was mailed to potential respondents
together with a postage-paid return envelope and a cover letter again explaining the study
and reminding them that participation in the study indicated their informed consent.
Reminder post cards were sent 14 days later to individuals who had not yet returned
completed questionnaires. Twenty-five days after the first mailing, new cover letters and
replacement surveys were sent to those whose original questionnaires were still
outstanding. A final reminder post card was mailed 10 days later to individuals who had

not yet returned a survey.

Data Analysis

There were 608 usable questionnaires returned plus 36 undeliverable, 23 return to
sender and 19 blank surveys. Response bias was determined by telephoning 20 non-
respondents to establish why they chose not to participate.

To help the reader keep track of the research questions, variables and the items on
the questionnaire they are summarized in Table 5.

The data were analyzed using the statistical package, SPSS 10.1 (2001).
Questions 3b and 22g were reverse coded so they were directionally consistent with each
of the other variable questions. Then the FREQUENCIES procedure was applied to
obtain a “picture” of the data in the form of raw frequencies and percentage of occurrence
for the sample.

Initially, descriptive statistics were referenced. Then, an exploratory factor
analysis using principal axis analysis with varimax rotation was computed to determine

how many attitudinal dimensions existed. Five factors with eigenvalues greater than 1.0

76

Table 5: Research @estions, Study Variables and the Items on the Questionnaire

 

Variable Name Research Question Item(s) on
Survey
Attitude Q1: What attitudinal groups exist in the Q22a-t
(independent) watershed?
Q2: What is the composition of each group? Q22a—t: Q23-Q30
Q3: What are the fairness evaluations? Q22a-t
Personal Q4: What is the relationship b/w attitude groups Q22a-t; Q3a-d
Responsibility and Personal Responsibility?
Knowledge of Q5: What is the relationship b/w attitude groups Q6-Q10
Issues and Knowledge of Issues?
Knowledge of Q6: What is the relationship b/w attitude groups Q1 1-Q16
Action and Knowledge of Action Strategies?
Strategies
Locus of Q7: What is the relationship b/w attitude groups Q4a-f
Control and PR?
Intent to Act Q8: Do attitudinal groups differ in their Intent Q22a-t; Q20a-g
(dependent) to Act?
Q9: What is the relationship among all the All above items
variables?

Q10: Do the attitudinal groups differ in terms of All above items

variables that influence Intention to Act?

 

were produced. After dropping items that failed to load at greater than .40 on any of the
five factors along with those items not meeting face validity inspection, there remained
three factors. Confirmatory factor analysis was performed on the remaining three factors
to establish internal consistency and to confirm the results of the principal axis analysis.
Based on this analysis, the three factor solution was reduced to two. Confirmatory factor
analysis was again run on the final two factors. Next, a hierarchical cluster analysis was
performed on the resulting factors to identify groups of individuals who responded
similarly to the attitude dimensions. An agglomeration schedule explaining the greatest
difference between clusters was computed using a squared Euclidean distance measure of
Ward’s method. The resulting dendrogram indicated there were two clusters Of

respondents. K-means cluster analysis with two clusters (based on the hierarchical

77

results), was run to produce a new variable indicating respondent’s cluster membership.
Finally, because the investigator was interested in the relationship between the resulting
clusters of “types of attitudinal orientations” and socio-demographic variables the
crosstabs procedure was run.

The researcher was also interested in the relationships between the other study
variables and the attitudinal orientations together with their influence on intention to act.
After either a scale or index was created through summing individual responses to each
item in a variable (see previous section on each variable), individual scores were assigned
to the respondents and a new variable created. In order to better understand the
relationships between all the variables, a correlation matrix was produced. Next, t-tests
were computed to better understand the differences between clusters and each study
variable. Finally, multiple regression (OLS) techniques were employed to evaluate the
independent effects of specific variables on individual assessments of how to implement
best management practices in the Sycamore Creek watershed. The regression analysis

calculates results for the entire sample as well as for each cluster group.

78

CHAPTER IV:

DATA ANALYSIS AND INTERPRETATION

The purpose of this study was to assess the relationships between the antecedents
of knowledge, attitude, locus of control and sense of responsibility on the intention to
support the implementation of watershed best management practices. It was also designed
to examine the fairness judgments underlying individual attitudes. This chapter has been
divided into the following sections: (a) sample profile and biases, (b) constructing the
attitudinal clusters, (c) cluster demographics, (d) fairness evaluations, (e) relationship
between knowledge, personal responsibility, locus of control and the individualistic and
egalitarian clusters, (1) the individualist and egalitarian clusters and the intention to act,
(g) correlations between the study variables, (h) variables having the greatest influence

on the intention to act, (1) non response survey and (j) study limitations.

Sample Profile and Biases

The following demographic variables were included in the study: age, income,
education, number of people per household and length of residency in their current home
and in the area. The purpose of profiling respondents was to address the
representativeness of the sample to the Ingham County population and to see if
demographic variables influence the intention to act. Population statistics were obtained
from the 1990 and 2000 Census collected by the US. Census Bureau. Biases are

addressed in the narrative. The results are in Table 5.

79

Table 6: Demographic Characteristics of Respondents Compared to the Population

 

 

 

 

 

 

 

of Ingham County
Demographic Sample Population

Characteristics % %
Age

< 18 yrs old 26.41
(1) 18-24 years 1.0 8.4“
(2) 25-34 years 13.6 11.0.'
(3) 35-44 years 20.4 17.1‘
(4) 45-54 years 26.4 18.41
(5) 55-64 years 18.4 10.0‘
(6) 65+ years 20.2 8.71
Education
(1)<HSGrad 3.3 16.12
(2) HS or GED 17.2 23.82
(3) Some College 30.5 30.9”
(4) College Grad 30.7 16.32
(5) Post Grad 18.4 12.92
Income
(1) < $10,000 2.3 10.52
(2) $10,000 - $14,999 2.4 5.7”
(3) $15,000 - $19,999 3.4 7.02'
(4) $20,000 - $29,999 12.1 14.72‘
(5) $30,000 - $39,999 17.3 15.12'
(6) $40,000 - $49,999 16.4 14.82’
(7) $50,000 - $59,999 14.3 8.02
(8) $60,000 - $69,999 10.9 8.02'
(9) $70,000 + 20.9 16.11"
Avg. Size of Household 2.2 people 2.6I peOple
Yrs Lived in Area 35.3 yrs NA
Yrs in Current Home 17.5 yrs ~ 15.5 yrs”

 

NA = Not Available

0 Percentages may not add up to 100% due to rounding

0 (1) Source: US. Census Bureau, Census 2000 (US. Census 2001)

0 (2) Source: US. Census Bureau, Census 1990 (US. Census 1991)

* Due to different categories this is an estimate based upon combining
census categories.

80

Demographic Characteristics
Age

Comparison between the study sample and Ingham County age categories reveals
that the median category for both is 45 to 54 years of age. The sample underrepresented
the age group of 18 to 24 years of age. This is most likely due to having drawn the
sample from a list of property owners. While few young adults have the means to own
property, many Older adults have made this investment. Since the study was interested in
the opinions of property owners, this deviation from the population age distribution is
acceptable. The other age categories for the sample are generally consistent with the

population percentages.

Education

The levels of education in the sample were somewhat greater than found in
Ingham County. Specifically, the sample population over represented the percentage of
college graduates as are found in the general population (30.7% vs. 16.3% respectively).
Furthermore, this trend continued at the post graduate level with 18.4% of the
respondents possessing a post graduate degree while only 12.9 % of the pOpulation claim
to be post graduates. Conversely, only 3.3% of the sample had less than high school
equivalence while there was closer to 16.1 % of the peOple in Ingham County in this
category.

There may be two reasons for having a more educated sample population. First,
the Ingham County data was drawn from Census 1990 which has the most available data
on people’s education levels and the category percentages might have changed since that
time. Still, it is doubtful that the levels seen in the sample have been attained throughout

the general population. The second possible explanation is that of self-selection bias. It is

81

consistent with the literature that better educated people have more of an interest in
environmental issues (Arcury et al., 1986, 1987) and would therefore be more likely to
take the time to fill out and return the questionnaire. This may result in biases associated
with the study variables that are educational in nature such as the desire for voluntary
and/or information and education programs. Also, the relationships between knowledge
and the other study variables of attitude, locus of control and personal responsibility may

be inflated compared to the general population.

Income

The median household income for Ingham County in 1998 was $41,743 (MIC,
1999). The 90% confidence interval surrounding the median is $38,919 to $44,551.
Income for this year was not broken down by category and Census 1990 figures are
therefore reflected in Table 6. Based upon these two sources of data, the study sample
appears to under represent lower income levels while slightly over representing higher
income levels. If the sample is divided at the $40,000 level (roughly the median), then
37.5% of the sample fall below this vs. 53.0% in the general population. Comparatively,
62.5% of the study sample is above the median, while only 46.9 % of the general
population is above the median. The results of this bias may be that the study sample,
which has a higher disposable income than the rest of Ingham County, might be more
willing to undertake best management practices requiring higher levels of public funds
because they maybe more willing to incur the perceived higher taxes required to pay for
these programs. More resistance to these types of programs by the general public would
be consistent with placing pro-environmental behavior as a secondary issue, only to be

attended to when all other concerns have been dealt with (Eden, 1993; Maslow, 1959).

82

Average Size of Household

The average size of the sample household is 2.2 peOple. The 95% confidence
interval ranges from 2.1 to 2.3 people. The average household size of owner-occupied
units in Ingham County is 2.6 peOple, which falls outside the samples confidence interval.
The population generally has more people per household than reflected by the sample.
This finding is consistent with having a slightly Older study sample where it would be
more likely that fewer children still reside at home. These results may create a similar
bias as was thought to arise from having respondents with higher incomes than in the
greater population. That is, the general population may have more basic needs, such as
taking care of one’s children, that they consider first before they think about their pro-
environmental behavior. Conversely, households with children may be more inclined to

consider inter-generational equity issues and support pro-environmental programs.

Number of Years One has Lived in the Area and in Their Current Home

The average years that a respondent has lived in the Sycamore Creek area was
35.3 years (SD = 19.6). The 95% confidence interval ranges from 33.7 to 36.9 years. An
average respondent is likely to have resided in their current home for 17.5 years (SD =
15.3, CI: 16.3 — 18.8 years) compared to 15.5 years for homeowners in the greater
Ingham County area. The fact that respondents have lived and invested in the area for
significantly long periods of time is likely to influence the study results. For example, it
is possible that because respondents have lived there for so long, they have developed a
sense of place that compelled them to respond to the survey. Beyond simply being self-
selecting, respondents’ length of residence might also influence their level of knowledge

about the area otherwise not found in the general Ingham County population.

83

Constructing the Attitudinal Clusters

The twenty fairness heuristic statements were factor analyzed (Table 7) using
principle axis analysis with varimax rotation. The final solution produced five factors
with eigenvalues greater than 1.0 and explained 49.3% of the variance. Based on
convention, items with factor loadings greater than .40 were selected for each factor.
Also, any inconsistent items (i.e. lacking face validity) were not selected. Consequently,
the statement concerning “public involvement in the decision making process” was
dropped by convention. And the statements on “saving waterways for the future is more
important than making money now” and “long-term health of waterways should be
achieved even if it reduced short-term business profit” were dropped due to lack of face
validity.

The eigenvalues for the five components ranged from 1.03 to 4.19. The Cronbach
alphas were somewhat less impressive with only three factors nearing the conventional
minimum threshold of .60. Consequently, Factors 3 and 5 were immediately drOpped due
to poor internal consistency. These results were not unexpected as Syme and his
colleagues (Syme & Fenton, 1993; Syme & Nancarrow, 1996, p. 1849) have noted “. ..
general scales of beliefs about equity may be too simplistic to reflect justice beliefs.”
Therefore, because of inherent difficulty of scaling equity and justice issues the
researcher expected both low internal consistencies and less than eloquent factors to
emerge from the analysis.

The remaining factors were reanalyzed using confirmatory factor analysis (CFA).
The CFA revealed there was considerable correlation (.79) between Factors 1 and 2
(Table 8). Upon further inspection of the items contained in both factors, it became

apparent that they did appear to be one construct. Based upon the results of the CFA and

84

Table 7: Principal Component Factor Analysis Results of 20 Heuristic Statements
Fairness Heuristic Factor Factor Factor Factor Factor

 

Concept 1 2 3 4 5
Q22N Upstream is morally .638 .191 -5.2E-02 .286 -5.9E-02
responsible to downstream
Q22M With a fair process the .611 4.1E-02 -9.4E-02 .130 1.8E—02
public should accept the

decision
Q22J Value other than dollar value .591 .155 -6.8E-02 -.170 6.9E—02
Q22E Polluters pay greatest share .516 .105 -.173 -.252 .1 l7

 

 

 

 

Q22R Saving waterways now is .501 .440 -. 162 -.204 -5.9E-02
more important than making
money

Q22] Long-term health over short- .499 .480 -3. lE-02 -. 149 -. 161
term profit

Q22C Analyzing costs and benefits 4.3E-02 .668 9.3E-02 -3.3E-02 .172

Q22T Cannot wait for exact .163 .535 -.211 -8.1E-02 -5.5E-02
knowledge

Q22B Make personal sacrifices .428 .523 -2.9E-02 5.4E-02 -5.9E-02

Q22F Environment same right to .104 .513 -.451 -. 124 .220
water

Q22D Manage for overall public .416 .450 -4.1E-02 2.5E-02 .231
benefit

Q228 Not all parts of environment -l.lE—02 -. 128 .815 -2.30E- .110
are valuable 02

Q22Q Clean waterways for -.235 4.8E-03 .609 .120 3.0E-02
environment second to
peOple

Q22H Pollution programs .105 -l.4E-02 4.0E-02 .781 -5.2E-02
maximize local economy

Q22L Minimize conflict in .270 -.228 .341 .500 .239
community

Q22P Groundwater is landowner's -.202 -.147 -. 100 .449 .336
property

Q22K Government not involved -.295 -2.9E-02 .339 .448 -2.9E-02

Q220 People should be -.177 9.2E-02 .112 .115 .684
compensated

Q22A Community has a right to .263 .104 2.2E-02 -4.7E-02 .659
say

Q22G/RC Public involvement not part .292 -.378 -.366 -.351 .390
of decision-making

 

Eigenvalues 4.19 1.97 1.44 1.22 1.03
Cronbach’s alpha .595 .627 .490 .508 .313
Percentage of variance explained 20.9 9.9 7.2 6.1 5.1

Cumulative variance explained 20.9 30.8 38.0 44.1 49.3
Overall mean 4.07 3.80 2.45 2.79 3.75

 

85

face validity it was decided to combine Factors 1 and 2 into one factor. This left two
factors, one with nine items and the second consisting of four items.

Table 8: Correlation Matrix of the Three Factors

 

 

Factor Factor Factor
1 2 4
Fa?" 1.00 .79 -.22
Fa?" .79 1.00 -.38
Fa?" -.22 -.38 1.00

 

 

 

 

A second CFA was run using a two-factor solution in order to determine whether
there was an improvement in the Cronbach alphas as well as other tests for internal
consistency. The Cronbach’s alpha for Factor 1 was .73, which by convention is
considered acceptable (Boster, 1999). When Factor 1 was compared to Factor 4 for
parallelism it was found there was a departure from both flatness and gradient. Hence, the
two factors were not measuring the same construct.

Factor 4 had a Cronbach’s alpha of .50 indicating poor internal consistency.
Closer inspection through the CFA analysis indicated that the model did not have a flat
pattern (X = 26.22, p = .000) but there was a gradient (X = 5.08, p = .406). As mentioned
above, the parallelism tests also indicated they were not measuring the same construct as
Factor 1. Despite low internal consistency, Factor 4 was retained based on its
eigenvalues, presence of an internal gradient and face validity.

The combination of the initial Factors 1 and 2 converges on the rights of the
environment and the moral obligation to protect it. It grouped those items that focused on
how to manage the process and the criteria that should be used for decision-making. This

factor could be labeled MORAL IMPERATIVES. The second factor focused on many of

86

the notions associated with the libertarian values of minimal government intervention and

the inalienability of prOperty rights. This factor could therefore be called RESISTANCE

TO CHANGE. The correlation between the two dimensions was -.33. The items

contained in these scales are:

Table 9: Attitudinal Factor Items

 

Factor 1: MORAL IMPERATIVES

 

Q22h

Everyone may have to make some personal sacrifices if we are going to
have effective water resource programs.

 

Q220

You cannot really solve water pollution problems by analyzing the costs
and benefits in dollars.

 

Q22d

Everyone owns rivers and streams and therefore they should be managed
for the overall public benefit.

 

Q22e

Those who pollute the most should pay the greatest share of clean up and
protection costs.

 

szf

The environment has the same right to water as people have.

 

QZZJ

Water has value other than its dollar value.

 

Q22m

Those upstream have a moral responsibility to look after the interests of
those downstream.

 

Q22n

If the decision-making process is fair, people should accept its decisions
for addressing water pollution.

 

Q22t

There is not time to wait for exact environmental knowledge, we need to
act now.

 

Factor 2: RESISTANCE TO CHANGE

 

Q22h

Water pollution programs should be made to maximize the overall
economic income of the community.

 

Q22k

The government does not need to be involved in cleaning up and
protecting water resources.

 

Q22]

If we are to clean up and protect our rivers, it should be done so as to
minimize conflict in communities.

 

 

Q23)

 

Groundwater (water under land) is the property of the landowner above.

 

The two dimensions only accounted for a total of about 37.0% of the variance

explained. MORAL IMPERATIVES alone accounted for 30.8% of the variance. The

RESISTANCE TO CHANGE dimension explained 6.1 % of the variance.

Again, these low explanatory results were anticipated due to the interrelated nature of the

concepts underlying the fairness heuristics.

87

 

The overall means of the five dimensions ranged from 2.79 to 4.07 and indicated
that MORAL IMPERATIVES were somewhat important to respondents (mean is 4.07 +
3.80/2 = 3.93) while RESISTANCE TO CHANGE was less so (mean is 2.79). However,
as will be shown shortly, different clusters of people did place importance on .

RESISTANCE TO CHANGE.

Identification of the Clusters

Cluster analysis was applied to identify groups of respondents with similar
responses to the two fairness heuristic dimensions. Initially, a ward’s hierarchical
clustering method was used to determine the number of clusters. Examination of the
dendrogram and agglomeration coefficients suggested two clusters. This number of
clusters was used a priori in a follow up non-hierarchical (K-means) cluster analysis. The
results were one cluster of 320 respondents and another cluster of 279 respondents.

In order to further classify the results of the cluster analysis, t-tests and a scatter
plot were used. The t-tests indicated a statistically significant difference between the two
clusters for each of the two fairness heuristic dimensions (Table 10). More specifically,
cluster 2 was more likely to place slightly more importance on the MORAL
IMPERATIVES (mean = 4.02, SD = .453) of managing the watershed than was luster 1
(mean = 3.82, SD = .522). Conversely, cluster 1 was more likely to resist change (mean =
3.30, SD = .409) than was Cluster 2 (mean 2.25, SD = .414). Further examination also
shows there is not any overlap between the 95% confidence intervals (CI) for each factor.

Figure six depicts the relationship between the clusters in two-dimensional space.
The MORAL IMPERATIVE factor is found on the x-axis while the RESISTANCE TO

CHANGE factor is on the y-axis. Inspection of the cluster centers clearly shows there is a

88

Table 10: Means of Fairness Heuristic Factors by Clusters

 

 

 

Fairness Cluster 1 . Cluster 2 Si
Heuristic (112320) (112279) t-test Lesa
Factors Mean SD Mean SD
MORAL 3.82 .522 4.02 .453 -5.10 .000“
IMPERATIVES 95%CI = 3.76 — 3.87 95%CI = 3.97 —— 4.07
RESISTANCE 3.30 .409 2.25 .414 31.30 .000*
TO CHANGE 95%CI = 3.26 -— 3.35 95%CI = 2.20 -— 2.29

 

* Significant differences between means at the .05 level of significance

HDNVHC) OJ. EIDNVlSISElH

 

 

 

 

6
5 ~ I: a 1:1
:1 a
1:1 1:1 1::
4 . 1:1 :1 a 1:1 a: cum
:1 o m an: o
u: 1::an Cluster
. mm
3 Membership Center
In D D WV
WVV W
l
1. Wm... ° *
WV
v
v w v 2 +
l - V
V V
V
0
0 2 3 4 5 6
MORAL IMPERATIVE

Figure 6: Cluster Membership and Centers by Factors

89

small (but significant) difference with regard to the MORAL IMPERATIVE factor with
Cluster 2 believing that the environment has slightly more rights than do the people in
Cluster 1. Regarding the RESISTANCE TO CHANGE factor, Figure six illustrates there
is a significant difference between Clusters l and 2 with the former being more resistant
to change than the latter.

Based on the means each cluster had for the two factors, the clusters can
tentatively be named individualists (Cluster 1) and egalitarians (Cluster 2). These names
are drawn from Dake’s (1991,1992) worldview categories based on interpersonal
relationships. As the analysis progresses, we will look for further evidence in terms of the

grid/ group nomenclature that might support these labels.

Cluster Demographics

Table 11 presents the demographic profiles of the individualist and egalitarian
cluster groups. The demographic categories available for analysis were age, education,
income, household size, years lived in the area and years in their current home. The two
clusters were statistically different on each of the demographic variables.

The individualist cluster is significantly Older than the egalitarian cluster. The
mean for the individualists cluster was 45-55 years of age whereas for the egalitarian
cluster it was only 35—45 years. Forty-five point five percent of the individualist
respondents were older than 55 years of age while only 30.7% of the egalitarian cluster
fell into this category. Examinations of the 95 % confidence intervals shows there is no
overlap between the cluster’s mean ages and are therefore distinctly different. The t-test

results and significance level at the 95% level of confidence supports these observations.

90

Table 11: Demographic Profiles of Cluster Groups

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Characteristics Total Individualist Egalitarian
Respondents (n= 320) Q1=279)
% % % t-test Sig.
Level

Age Categories
Mean/Std. Dev. 4.08 SD = 1.35 4.33 SD = 1.33 3.80 SD = 1.32 4.847 .000*
(1)18-24 years 1.0 1.0% 1.1%
(2) 25-34 years 13.6 8.7 19.0 95 % CI
(3) 35-44 years 20.4 18.7 22.3 CLI 4.18 —— 4.48
(4) 45-54 years 26.4 26.1 26.7 CL2 3.64 — 3.96
(5) 55-64 years 18.4 18.7 17.9
(6) 65+ years 20.2 26.8 12.8
Education
Mean/Std. Dev. 3.44 SD=1.07 3.22 SD=1.07 3.68 SD=1.03 -5 .188 .000*
(1) < HS Grad. 3.3 4.5% 1.8%
(2) HS or GED 17.2 21.4 12.4 95% CI
(3) Some College 30.5 34.7 25.8 CLI 3.10 — 3.34
(4) College Grad 30.7 25.6 36.4 CL2 3.55 — 3.80
(5) Post Grad 18.4 13.6 23.6
Income
Mean/Std. Dev. 6.26 SD: 2.78 6.04 SD=2.16 6.49 SD=1.97 -2.494 .013*
(1) < $10,000 2.3% 3.3% 1.1%
(2) $10,000/$14,999 2.4 3.3 1.5 95% CI
(3) $15,000/$l9,999 3.4 3.3 3.4 CLI 5.78 —- 6.30
(4) $20,000/$29,999 12.1 13.8 10.3 CL2 6.25 — 6.73
(5) $30,000/$39,999 17.3 19.0 15.6
(6) $40,000/S 16.4 14.5 18.3
49,999
(7) $50,000/$59,999 14.3 12.6 16.0
(8) $60,000/$69,000 10.9 12.3 9.5
(9) $70,000 + 20.9 ' 17.8 24.0
Mean/Std. Dev. 2.07 SD: 1.15 2.31 SD =1.43
Household Size 2'18 SD‘ “'29 (:11 1.94 — 2.20 CI' 2.14 — 2.48 '2'21 I 027*
Mean/Std. Dev. 39.02 SD: 194530.90 SD: 18.82
Yrs Lived in Area 35°18 SD_ ”'56 CI'36.84 — 41.22 CI'28.65 — 33.14 5'09] 000*
Mean/Std. Dev. 20.91 SD: 162513.44 SD: 12.67
Yrs Current Home ”'35 SD‘ 15’“ CI‘19.08 — 22.76 Cl'l 1.94 -— 14.94 6'127 000*

 

* Significant differences between the cluster means at the .05 level of significance
(1) 95% confidence interval

The egalitarian cluster is more educated (mean = 3.68) than the individualist

cluster (mean = 3.22). Even though both of these means represent the "Some College"

91

category, it becomes apparent that the egalitarian cluster is significantly more educated
when the "College Graduate" and "Post Graduate" categories are summed for both
clusters. When this is done, it is found that a full 60% of the egalitarian cluster are at least
college graduates versus only 39.2% for the individualist cluster. Again, analysis of the
95% confidence intervals shows there is no overlap between the two clusters and they are
statistically different. The t-test result of—5. 188 with a significance level of .000 at the

95 % level of confidence support these findings.

The income difference between the clusters is less pronounced than the two
previous demographic categories. Specifically, the mean household income for both
groups was the $40,000 to $49,000 range, with the individualist cluster’s mean being in
the low $40,000 and the egalitarian cluster’s being in the mid $40,000. Inspection of the
confidence intervals reveals that indeed there is a small overlap (.05) at the 95% level of
confidence. The higher, but still significantly different significance level of the t-test
(.013) would also indicate that there is less variation between the cluster means.

The results on household size are similar to the findings on income. Here the
mean household size for the individualist respondents was 2.07 people and 2.31 for the
egalitarian respondents. The t-test results were —2.21 with a significance level of .027;
indicating there was a statistically significant difference between the two clusters. As was
expected from the t-test results, there is an overlap of .06 between the two 95%
confidence intervals.

There was a statistical difference between the clusters for both the mean number
of years lived in the area and mean number of years in their current home variables. The

mean number of years for living in the area for the individualists cluster was 39.02 years,

92

while the mean number of years for the egalitarian respondents were almost 10 years less
at 30.90 years. This pattern was mirrored in the number of years that respondents lived in
their current home. In this case, the individualist group resided in their current home for
an average of 20.91 years versus 13.44 for the egalitarian group. Inspection of the
confidence intervals and t-test indicates the two groups are distinct with regard to these
two variables that may or may not have any relevance since the residency time is
relatively long for both groups.

In summary, the individualist group is Older, less educated, has less household
income, fewer children and resided in both their home and the area longer. Conversely,
the egalitarian group is younger, better educated, has a higher household income, more
children and have not resided in their homes or the area for as long. The fact that the
individualists group was older and had longer residency in the area and in their home is
consistent with their being more likely to resist change. With regard to the egalitarian
cluster, the fact they are younger, better educated and had higher household income is
consistent with their feeling more strongly about the rights of the environment (Van Liere

& Dunlap, 1980; Christianson & Arcury, 1992).

Fairness Evaluations

The percentage frequency of responses to each of the twenty items on the
“agreement scale” is shown in Table 12. It is noticeable that the responses to most items
were skewed, indicating a high level of agreement among all respondents. Those items
that exhibited the most normal distribution were "you cannot solve pollution problems by

analyzing costs and benefits", "pollution programs should maximize the local economy",

93

Table 12: Percentage Distribution, Mean Score and Standard Deviation of
Responses to Fairness Heuristic Statements
Strongly Disagree Neutral Agree Strongly

 

 

 

 

Number Disagree Agree

Fairness Heuristic of (1) (2) (3) (4) (5) Std.
Statements Responses % % °/o % % Mean Dev.
Communilyhasa 597 0.2 2.3 9.9 65.00 22.6 4.08 .655
righttosay
Tomakcpcmnal 600 1.00 2.00 10.8 69.2 17.00 3.99 .670
sacrificesmi
Analylingcostsand 593 2.9 13.7 27.0 43.7 12.8 3.50 .976
benefitsmli
Manageforovcra" 598 1.0 5.4 12.2 59.2 22.2 3.96 .805

. mi
public benefit
Polluterspay 600 0.0 1.3 7.2 42.5 49.0 4.39 .680
_greatestsharem‘
Environmenthas 593 1.2 5.1 16.4 49.2 28.2 3.98 .869
. mi
right to water
Publicnotinvolved 596 2.0 4.7 13.8 53.9 25.7 3.96 .874
in decision
Pollutionpmgmms 586 5.1 22.2 39.1 29.4 4.3 3.05 .943
maximize local

11c

economy
Long-tennhcalthvs- 592 0.7 4.4 21.1 57.6 16.2 3.84 .766
short-term profits
Valueotherthan 596 1.3 0.8 3.7 49.7 44.5 4.35 .717
dollarvaluemi
Govemmentnot 597 1.8 5.0 12.6 46.9 33.7 4.06 .910
involved "c
Minimizeconflictin 595 1.3 11.3 29.2 49.7 8.4 3.53 .851
community I“
Upstreamismomlly 588 1.2 3.6 9.5 60.7 25.0 4.05 .771
responsible to
downstream
Fail’PFOCCSSthC 586 0.5 4.1 14.8 70.6 9.9 3.85 .661
public should accept
decision
PC01319811091dbe 585 2.7 15.6 45.1 32.6 3.9 3.19 .844
compensated . .
Groundwaterism 584 10.1 36.1 33.6 16.6 3.6 2.67 .985
landowners property
CleanwatefwaYSfOT 578 14.7 52.9 20.8 10.2 1.4 2.31 .892
environment second
topeople
SaVingwaterwaYS 590 2.2 2.5 14.6 57.3 23.4 3.97 .825
more important than
money
1801311133418“ 588 2.4 16.2 30.3 40.8 10.4 3.41 .957
envrronmentare
valuable
Cannotwaitfor 586 2.4 12.1 28.8 43.2 13.5 3.53 .952

exact knowledje mi

 

mi = MORAL IMPERATIVE, rtc = RESISTANCE TO CHANGE

94

"water management programs should minimize conflict in the community , pe0p1e
should be compensated for water quality programs that hurt their livelihood",
“groundwater is a landowner’s property", "not all parts of the environment are valuable"
and "we cannot wait for exact knowledge before we act".

Those items pertaining to "the community’s right to a say on how to manage the
watershed", "the polluter pay principle", "water’s value other than in economic terms",
and "those upstream having a moral responsibility to those down stream” tended to be
highly skewed. The skewed pattern of responses for the items support Syme and
Nancarrow’s (1996) assertion, at least prima facie, that these items would cause problems
in analysis based on the assumptions of univariate and multivariate normality.

Product moment correlations among the items not included in either of the two
attitudinal factors were generally small (mean = 0.146, variance = 0.011) and were not
significant at the .05 level. This is consistent with respondents viewing the concepts these
items conveyed as being highly discriminated and generally independent of other items.

Table 13 further analyzes the results of those items of the fairness heuristic in
which 80% of the respondents (in total and by cluster) either agreed or strongly agreed
with the fairness statement. These items were first ranked based upon the total sample
and were then again ranked based upon cluster membership. This allowed for the
similarity and differences between the cluster’s responses to the fairness heuristic to
become more apparent.

While these statements are only a portion of the of the fairness heuristic, they
reflect the general phiIOSOphies which emerged from the research. At this universal

fairness level five themes emerged: (l) the value of waterways, (2) procedural justice

95

_0>0_ 00000588 mo. 05 00 805000 0:0 805 8080.50 5008588 0.3 80820 ..

 

 

L _ 8 N. 20 a E
a 5 2 8 _ _ a:
..w 08 a S 2 m8
2 _m 0 cm a wow
1 8 _ _ 80 w 0.8
.0 8 8 N8 5 :8
2 0.8 a E: 0 4. a

5 c8 4 0.8 m 0.8
a. .8 m S 4 08
m 8 m 8 m 8
..m a N 8 N m. a
..m 03 _ 8 _ 8
08:8 3 0:85. .x. 8358 .x.
#5 83 88
N .8830 n .8820 0.95% .308

8508 803.58 8000008 8 .: 805 005300 00 2:08
883.0003 800. .80 580: 8808:0880 8.8800— 05.
0>0n

08000 8 80.03 3 Emt 088 05 mm: 80800880 05.
800882 80503 880008 80.5 80008 8308
803000 05 «0 can 00 50: 2008 80802038 0555
088:3 8083 9880500 80% 008600 8 80000
2008 0800a 85 a 80008 8308808000 05 .3
800888 80003 8500008 000 a: 88020

8 00202: 00 9 0000 50: 8000 80850.38 0:...

30: 30:08 mac—08

005 85.508 0.08 8 0885 05 80.0 803.5003 w8>0m
50000 0:58 5830 05 80.5 008808 00 2:08

>05 0850805 000 88008 08 80>: 830 080b0>m
80080300 805 .«0 808058 05

.050 x00— 3 mam—£80080“ .808 0 08: 800.8% 89:.
880808 00888 8083 0>500b0 038 00 w80w 08 03
.5 8005808 580808 088 0508 8 038 >08 000b0>m
80808808 80003 w8£0>8 808$ 00

m8 805 8 Ewt 0 038 8858800 05 m0 800808 =<
800 00500508 000 a: 000—0 m0 088

805008 05 38 2:08 808 05 08:00 0.3 805.
02? 8:00 8 :05 ..050 020> 8: 80503

80
880
080
88
680
80
080
880
£8
£3

680
.80

808085 03880: 808.8”—

 

8080§m 0.885: 8088,,— nomm 80m 008080.505 0880M 0:0 53 0008< 055 88000080“ m0 0380803” “2 050...

96

considerations, (3) distributional justice consideration, (4) environmental and human
rights and (5) governmental involvement.

The number one ranked item by all respondents was “water has value other than
its dollar value”. Other items that could be included in how respondents valued their
waterways are Q22e, Q22r and Q22i. These items were ranked second, seventh and
twelfth respectively. Taken as a group, these items focus on the trade off between the
environment and the economy and how programs should be funded. There was strong
support at the sample level for the polluter pays principle.

Procedural justice considerations appeared third in the ranking by the entire
sample. The statement “all members have a right to their say on issues involving water
management” was agreed with by 88% of all respondents. Other items relating to
procedural justice were Q22n and Q22g, ranked ninth and tenth respectively. These items
as a whole spoke to the right to have voice, public involvement and a fair process.

The third theme to emerge concerned itself with distributional justice
considerations. There were two items relating to this theme: (Q22b) “everyone may have
to make some personal sacrifices if we are going to have effective water resource
programs” and (Q22d) “everyone owns rivers and streams and therefore they should be
managed for the overall public benefit”. The former speaks to the distribution of costs,
while the latter speaks to the distribution of benefits.

Question 22m on the moral responsibility to look after your neighbors
downstream (Q22m) (ranked as the fifth item by the entire sample), and the “environment

having the same right to water as pe0p1e have” (Q220, constitutes a rights based theme.

97

Question 22m reflects human rights while Q22f concerned itself with environmental
rights.

The last theme to be considered important by all respondents was governmental
involvement (Q22k). It was ranked eighth by all respondents.

There were five of the twelve fairness heuristic items that appeared in the
Australian water allocation study "top eight" rankings (Syme et al., 1999). At the
universal fairness level, the Australian studies had ongoing support for the community’s
rights to have a voice in the allocation decisions, the rights of the environment and having
appropriate outcomes of procedural justice (Syme et al., 1999). Some of the similarities
include an emphasis on the environment’s right to water with two of the Australian
studies having about 78% - 81% of respondents supporting this notion, a similar result to
this study’s findings. Another similarity between the studies was the weight placed on
accepting the decisions made by a fair process. In the Australian studies support for these
concepts ranged from 64% - 83%, while this study found 80% - 81% support. In terms of
the differences between the two countries, those items contained in both studies that
specifically traded off economic prosperity with environmental health appeared in this

study’s ranking, but not in the Australian standings.

Comparison Between Individualist and Egalitarian Cluster’s Fairness Evaluations

There are statistically significant differences between how the two clusters of
respondents, individualists and egalitarian, ranked these fairness heuristic items. The
most glaring difference between the groups is their views on governmental involvement.
While the egalitarian cluster ranked this item (Q22K) number one, the individualist

cluster ranked it eleventh. This finding is consistent with the individualist view that

98

deregulation is a rational management strategy because they view nature as being benign.
Also, individualists have a low group orientation and believe value decisions stem from
personal judgments rather than collective control (Rayner, 19883) as might be
represented by governmental involvement. Conversely, egalitarians might see
governmental involvement as being able to help foster the equality of outcomes (Rayner,
1988b). It is also consistent with their high group philosophy (i.e. everyone should be
involved).

The differences between the cluster rankings on other fairness heuristic items
would seem to support these findings. For instance, there were statistically significant
differences between clusters on the items in the value of waterways theme. Most telling
were items Q22r and Q22i that directly pitted environmental protection of our waterways
against economic profit. In both these questions, the individualists were considerably less
willing to forgo economic gain for the sake of a cleaner environment. Specifically, on the
“saving waterways for the future being more important than making money” item only
73.2% were in agreement. And on the “long-term health of local waterways should be
achieved even if it reduces short-term profits” item only 65.1% were in agreement. The
egalitarians were 89% and 84% in agreement with these items respectively. The two
clusters also differed on their level of agreement about “water having a value other than
its dollar value” (Q22j) and the polluter pays principle (Q22e) but considerably less than
the items just mentioned. Despite these apparent differences, two-thirds to three-quarters
of the sample population are in agreement about these fairness heuristic concepts.

The two clusters also differ significantly (t = -4.638, p = .000) on whether public

involvement should be part of the decision making process. The individualist cluster does

99

not agree as strongly (73%) as the egalitarian group (87.3%) that the public should be
involved. This is particularly interesting since both clusters do not differ (statistically) on
“people’s right to have their say” (Q22a) and “accepting the decision of a fair process”
(Q22n).

The item on everyone having to make “personal sacrifices in order to have
effective water resource program” (Q22b) was less supported by the individualist cluster
(82%) than the egalitarian cluster (92%). This is consistent with individualists being less
willing to forgo income to protect their waterways as discussed earlier.

The last item on which the two clusters differ is on the “environments right to
water being the same as people’s rights”. Seventy-three percent of the individualists
agreed with this statement versus 87% by egalitarians.

In summary, the two clusters were in agreement about water having a value other
than its dollar value and the polluter pays principle. The items that they differed on
supported our initial labeling of the clusters as individualist and egalitarian. That is, the
individualist responses to the fairness heuristic items clearly demonstrated they favored
less government involvement, were less willing to forgo economic gains for a cleaner
environment, believed in the public's right to voice their Opinions but were less willing to
be involved in decision making and placed people’s rights over that of the environment
relative to their egalitarian counterparts. On the other hand, the egalitarian cluster was in
favor of government involvement, were more willing to forgo economic gain for a
cleaner environment, believed fair decision making processes involved the public and
their right to voice themselves, and awarded the environment a higher level of rights, all

relative to the individualist cluster. More importantly, despite the differences in the

100

degree of support indicated by the respective clusters, more than two-thirds to three-
quarters of the sample population was in agreement with the concepts presented by each
of the fairness heuristic items. These similarities can provide the basis for designing
fiJture deliberative planning processes.

Relationships Between Knowledge, Locus of Control and Personal Responsibility
and the Individualist and Egalitarian Clusters

If one thinks about the variables of knowledge, locus of control and personal
responsibility in the context of the individualist and egalitarian clusters, one might be
able to deduce what relationships might exist. For example, it would not be surprising if
the data revealed that the individualist cluster was higher on locus of control than their
egalitarian counterparts. This is because individualists value decisions stemming from
personal judgments (Rayner, 1988a) and defend their freedom to bid and bargain in self-
regulated networks (Thompson, 1992). In other words, they desire to be in control of
their surroundings.

Regarding personal responsibility, the egalitarian cluster is probably more likely
to assume more personal responsibility for the environment than the individualist cluster.
There are two reasons for this assertion: (1) egalitarians are thought to frame natural
resource issues in ethical terms because it allows them to focus on the social and political
dimensions and to criticize the institutions responsible for resource management (i.e. an
“if I don’t take responsibility who will” attitude) (Dake, 1992) and (2) because
individualists view nature as being benign, therefore not needing their attention.

In terms of knowledge, there is little in the literature to guide us as to how this
variable may relate to the two clusters. It could be argued that if egalitarians have more

personal responsibility, then they are more likely to have more knowledge because the

101

latter is thought to be a pre-requisite for the former (Hines et al., 1986). Also, the
previous demographic results indicating that the egalitarian cluster was more educated
than the individualist cluster suggest this might be the case. Table 14 reports the

differences between cluster means for the four psychological study variables.

Table 14: Comparison of Study Variables by Attitudinal Groups

 

Total In di vidualis ts Egalitarian t-test for Equality of
Responses Means
Std. Std. Std Sig.
Mean Dev. Mean Dev. Mean Dev. t—test df (2-tailed)

 

KNOWISS 2.56 1.45 2.50 1.46 2.66 1.44 ~1.399 597 .162

1 = no knowledge of issues
5 = knowledgeable of issues

KNOWACT 3.07 1.59 2.92 1.61 3.28 1.54 -2.806 582 .005*

CI' 2.33 -— 2.66CI‘ 2.50 — 2.83

1 = no knowledge of actions
5 = knowledgeable of actions

LOC 3.64 .593 3.57 0.64 3.73 0.51 -3.367 597 .001*

CI' 2.73—3.10CI‘ 3.10—3.47

1 = low perceived internal LOC
5 = high perceived internal Cl' 3.50 — 3.64 CI‘ 3.67 — 3.79
LOC

PERSONAL 3.75 .597 3.63 0.60 3.88 0.57 -5.257 597 .000*

1 = low personal responsibility CI‘
5 = high personal responsibility

* Clusters are statistically different from one another at the .05 significance level
1 95% confidence interval

3.56 — 3.69 CI' 3.8] — 3.95

Environmental Knowledge
The mean score for knowledge of the issue (KNOWISS) was 2.56 (SD = 1.45)

with a minimum of one and a maximum of five. The mean score for the individualist
cluster was 2.50 (SD = 1.46) and for the egalitarian cluster it was 2.66 (SD = 1.44). An

independent groups t-test revealed there was no difference between the two cluster

102

groups with regard to their knowledge of the issues. The overlap between the clusters’
95% confidence intervals confirms that there is little difference between the groups.

The mean score for knowledge of action strategies (KNOWACT) was 3.07 (SD =
1.59) also with a minimum of one and maximum of five. The individualist’s mean score
was 2.92 (SD = 1.61) while the egalitarian’s mean score was 3.28 (SD = 1.54). The
individual group t-test indicated there was a statistically significant difference between
the two clusters with the egalitarian group having more knowledge of action strategies.
Inspection of the 95% confidence intervals shows there is no overlap between the two

clusters and it can therefore be concluded the two groups do differ on this variable.

Locus of Control
The mean score on the locus of control (LOC) variable was 3.74 (SD = .89)

ranging from a low of one to a high of five. The individualist cluster’s mean was 3.60
(SD = 0.67) while the egalitarian cluster’s score was 3.83 (SD = 0.81). The t-test for
differences between groups indicated there was a difference and that the egalitarian
cluster had a higher sense of locus of control than did their individualist counter parts.
There was not a common range between the two confidence intervals. This is contrary to
the anticipated results discussed previously, but lends support to Eden’s (1993) view that
a person’s sense of environmental responsibility and efficacy are interdependent on how

they influence pro-environmental behavior.

Personal Responsibility
The mean personal responsibility (PERSONAL) score was 3.91 (SD = 1.02) with

a minimum of one and a maximum of five. Again, we see the egalitarian cluster as having

103

a larger mean score of 3.99 (SD = 0.78) than the individualist cluster’s mean of 3.75 (SD
= 0.96). The individual group t-test indicated these to be two distinct groups as was
shown by the separate confidence interval ranges. The result that the egalitarian cluster

was higher on this variable was as anticipated and helps explain the LOC results.

Summary

The analysis of the data indicated that the egalitarian cluster was more
knowledgeable about action strategies, believed it was in their power to make a
difference (higher internal LOC), and assumed more personal responsibility for the
management of local waterways relative to their individualist counterparts. Therefore,
egalitarians were higher on all the psychological variables. This was almost what had
been predicted, but originally it was thought that the individualist cluster would be higher
on LOC, than the egalitarian cluster. A possible explanation for this finding is Eden’s
(1993) observation that it seems meaningless to ascribe oneself responsibility for

something one has no control over.

The Individualist and Egalitarian Clusters and the Intention to Act
Analysis of the data so far supports there being two distinct clusters of individuals

in the study population with regard to their worldviews surrounding the management of
the Sycamore Creek watershed. One significant question that remains to be answered is if
these differing worldviews result in different intentions towards implementing best
management practices. Table 15 presents how the study sample population and each of

the clusters ranked the seven best management practices.

104

Table 15: Percentage Of Respondents Who Support and Ranked Preference for
Each Of the Best Management Practice Options
Total Sample Individualists Egalitarians

 

(N=608) (N=320) (N2279)
% Ranking % Ranking % Ranking
Q20d Fines for polluting 86.6 1 82.2 1 94.3 1*

Q20e Increased enforcement of
environmental regulations

Q20f Public information and
education programs

Q20g Voluntary programs to help

landowners adOpt *
environmentally friendly 80'] 3 '5 79-4 7- 83-8 5

practices

QZOO Stricter regulations on
activities that impact
waterways during
development

Q20a Zoning requirements for
some open space to be
preserved on undeveloped
land

Q20b Subsidies to landowners for
environmentally fiiendly 54.9 6 50.6 7 61.6 7*
practices

* Significant difference between means Of the cluster at the .01 level.

80.9 2 72.9 5 92.8 2*

80.1 3.5 76.9 3 86.4 4*

78.7 4 73.1 4 87.8 3*

71.0 5 64.3 6 81.0 6*

When an independent group t—test was run between the cluster groups on the
INTENT] variable (i.e. the sum Of all the BMP scores) there was a significant difference
(t = -6.820, p = .000). In order to further analyze the BMP options, the percentage support
for each option (would support and strongly support) was calculated and then each was
ranked against one another.

As a group, the respondents ranked fines for polluting as the number one BMP
Option, followed by increased enforcement Of regulations. There was a tie for the third
Option between public information and education and voluntary programs. Stricter

regulations were fourth, followed by the zoning of Open spaces and lastly subsidies. The

105

fact that fines for polluting was ranked first by both clusters again reflect the polluter
pays principle so prevalent in the results Of the analysis of the fairness heuristic.

The two clusters began to diverge after this initial agreement in terms Of the types
of programs they favored. The individualist cluster ranked both voluntary and public
information programs as their second and third options respectively. They also gave
lower rankings to increased enforcement (fifth) and stricter regulations (fourth). The
egalitarian cluster was consistent in their desire for fines, increased enforcement and
stricter regulation, ranking them first, second and third respectively.

Neither cluster indicated a preference for the zoning of Open space nor subsidies
and they were consistently ranked last. Subsidies were the most negatively received
Option with 8.7 % Of the respondents indicated they would not support this option.

The individualist cluster consistently had a lower percentage Of respondents who
would indicate their support on all the items. All the differences in the rankings by the
clusters were found to be statistically different by independent group t-tests. The “DO Not
Know” category never exceeded 10% Of the responses both for the entire sample
population and by cluster. Again despite the differences between the groups, five Of the
six BMP Options received between two-third to three-quarters of the sample population’s

support for implementation.

Correlations Between Study Variables

The seven study variables analyzed in this section are knowledge of the issues
(KNOWISS), knowledge Of action strategies (KNOWACT), the attitudes of moral
imperatives (MORAL), and resistance to change (RESISTI) that should be negatively

correlated with the other variables, locus Of control (LOC), personal responsibility

106

(PERSONAL) and intention to act (INTENTl). Pearson Correlation analysis was
conducted to explore the relationships among the variables (Table 16).

Table 16: Pearson Correlations (N = 608) Between Study Variables
(1) (2) (3) (4) (5) (6) (7)

 

(1) KNOWISS 1
(2) KNOWACT .394** 1

(3) MORAL .146 .132" 1

(4) RESISTl -.076 -.079 -.127" 1

(5) LOC .111" .058 .370** -.090* 1

(6) PERSONAL .158M .134** .383" -.189** .460" 1

(7) INTENT] .091* .066 .414M -.260** .449" .364" 1

 

** Correlation is significant at the 0.01 level (2-tailed)
* Correlation is significant at the 0.05 level (2-tailed)
The intention to support BMP implementation scores correlated significantly (p <
.01) with (in order Of importance) LOC, MORAL, PERSONAL and RESISTl. Intuitively
one would think that there would be a significant relationship between having knowledge
and the intention to act, yet surprisingly there was not at the .01 level. Several other
significant relationships were also found. KNOWISS was found to correlate with
KNOWACT, MORAL, LOC and PERSONAL. Relationships were found between
KNOWACT and the MORAL and PERSONAL variables. MORAL was found to
correlate negatively with RESISTl and positively with LOC and PERSONAL. The
RESIST] attitudinal measure exhibited a negative relationship with PERSONAL. And
lastly, LOC was positively correlated with PERSONAL. The magnitudes Of the
relationships are generally consistent with those studies reported in Table 4 at the end Of
Chapter II. Nevertheless when certain variables were controlled for, many Of these
relationships were no longer significant.
Table 17 presents the partial correlation coefficients, significance levels and the

controlled for variables. It can be seen amongst the antecedent variables to intention to

act, the strongest relationships were between KNOWISS and KNOWACT, MORAL and

107

 

Table 17: Partial Correlations Coefficients

 

 

 

 

Variables Coefficients Variables Controlled For
KNOWISS - KNOWACT 374 p Z 000 MORAL, RESIST], LOC,
' ' PERSONAL
KNOWISS - MORAL 034 Z 423 KNOWACT,RESIST1, LOC,
° 1’ ° PERSONAL
KNOWISS — LOC 037 p Z 379 KNOWACT, MORAL, RESISTl,
' ' PERSONAL
KNOWISS — PERSONAL 062 _ KNOWACT, MORAL, RESISTI,
. p -— .141 LOC
KNOWISS — RESIST] _ 017 Z 690 KNOWISS, MORAL, LOC,
' p ' PERSONAL
KNOWACT - MORAL 049 p Z 238 KNOWISS, RESISTI, LOC,
' ' PERSONAL
KNOWACT - RESIST] _ 054 Z 203 KNOWISS, MORAL, LOC,
' p ' PERSONAL
KNOWACT — LOC _ 026 p Z 540 KNOWISS, MORAL, RESIST],
' ' PERSONAL
KNOWACT — 067 p Z 112 KNOWISS, MORAL, RESIST],
PERSONAL ° ' LOC
MORAL - RESISTl _ 049 Z 240 KNOWISS, KNOWACT, LOC,
‘ P ° PERSONAL
MORAL — LOC 231 Z 000 KNOWISS, KNOWACT, RESISTl,
' p ' PERSONAL
MORAL - PERSONAL __ KNOWISS, KNOWACT, RESIST],
.233 p - .000 LOC
RESIST] — PERSONAL _ KNOWISS, KNOWACT, MORAL,
-.149 p — .000 LOC
RESIST] - LOC 088 Z 037 KNOWISS, KNOWACT, MORAL,
' p ' PERSONAL
LOC — PERSONAL 367 Z 000 KNOWISS, KNOWACT, MORAL,
. ' p ' RESIST]
KNOWISS — INTENTI 010 Z 818 KNOWACT, MORAL, RESISTI,
' p ' LOC, PERSONAL
KNOWACT — INTENTl _ 007 Z 867 KNOWISS, MORAL, RESISTI,
' p ' LOC, PERSONAL
MORAL - INTENTI 268 _ 000 KNOWISS, KNOWACT, RESIST],
' p "‘ LOC, PERSONAL
RESISTI - INTENT] _ 214 _ 000 KNOWISS, KNOWACT, MORAL,
‘ p " LOC, PERSONAL
LOC -— INTENTI 297 __ 000 KNOWISS, KNOWACT, MORAL,
' P "' RESISTl, PERSONAL
PERSONAL -— INTENT] 120 _ 004 KNOWISS, KNOWACT, MORAL,
° P ‘° RESIST], LOC
108

LOC, MORAL and PERSONAL, RESIST] and PERSONAL, RESIST] and LOC, and
between LOC and PERSONAL. Regarding the intention to act variable (INTENTl), it
can be seen that MORAL, RESIST], LOC and PERSONAL were all significant.

The difference between the partial correlation that controlled for all other study
variables and the Pearson Correlation on the relationship between KNOWISS and
KNOWACT was only .022. This insignificant difference between the results suggests
that these two measures of knowledge may not be as distinct as had originally been
intended. Therefore, these two measures will be summed together to create one variable
for use in the forth-coming regression analysis.

The attitudinal variables of MORAL and RESIST] along with the other
psychological variables Of LOC and PERSONAL interacted as was anticipated through a
review of the literature. Specifically, the following can be said about these relationships:
(1) the relationship between the two attitudinal variables of MORAL and RESIST] were
not significant in the partial correlation, (2) three Of the correlations, MORAL - LOC,
MORAL - PERSONAL and RESIST] - PERSONAL were all substantially lower in the
partial correlation analysis and (3) the LOC — PERSONAL relationship remained
relatively strong in the partial correlation analysis. These Observations would seem to
indicate a high level Of interrelatedness between the psychological variables used in the
study. The MORAL attitudinal factor was clearly viewed by respondents as being similar
to both the LOC and PERSONAL (responsibility) scales. The most robust relationship
was between locus of control and personal responsibility (.367, p = .000).

When each of the variables was correlated with the overall intention to support

best management practices while controlling for all the other variables, it was again

109

found that the relationships were weaker than in the Pearson Correlation. The variable
most resistant to change in the partial correlation was RESIST] while the most

significant variable to change was PERSONAL.

Variables Having the Greatest Influence on the Intention to Act

Multiple regression (OLS) analysis was used to assess the independent effects Of
specific variables on individual intention to support BMPS designed to improve water
quality. For the independent variable assessing cluster membership, a dummy variable
was constructed for use in the regression analysis.

Table 18 presents the mean and standard deviation Of each variable, their Beta
weights, t-tests and Significance levels. The F- test results indicate the model is
statistically significant. For the various demographic variables included in the model, it
was found that people with higher levels Of education were more likely to support the
implementation Of BMPS. The remaining variables Of age, income, length Of residency in
the area and in their current home were not found to be significant.

The knowledge of issues and action strategies variables were also not significant.
This is consistent with other studies that have found that knowledge has a weak
relationship with either attitudes or intentions (McFarlane & Boxall, 2000).

The next sets Of variables included in the model were the psychological measures.
This included the GROUP variable that encompassed the two attitudinal factors of moral
imperatives and resistance tO change and the other psychological variables Of locus Of
control and personal responsibility.

All three of the variables had the expected effect on the intention to support BMP

implementation and were statistically significant. Memberships in the egalitarian cluster

110

Table 18: Multiple Regression Estimates for the Intention tO Support BMP

KNOWALL

LOC

PERSONAL

GROUP

AREA

HOME

AGE

EDUC

INCOME

INTENT

Implementation

Knowledge Of Issues & Action
Strategies

1 - no knowledge

5 -— loiowledgeable

Locus of Control

1 - low internal control

5 —— high internal control
Personal Responsibility

1 - low sense of personal
responsibility

5 -— high sense Of personal
responsibility

Cluster Membership

0 - individualists

] —— egalitarians

Length Of Residency in Area
continuous

Length Of Residency in Current
House

continuous

Age Of Respondent

l -— 18 - 24

4 - 45 - 54

6 - 65 years Of age and up
Education Level

1 — less than high school

3 — some college

5 — post graduate degree
Household Income

1 — less than $10,000/year
6 - $40,000 -— $49,000/year
9 -— $70,000t/year

Intent to Support BMPS Options

1 — would not support
3 - would support
4 — would stroggly support

Mean

2.78

3.66

3.77

.502

33.89

16.05

4.02

3.48

6.24

3.23

Std.

Dev.

.954

.589

.608

.500

18.96

13.93

1.32

1.05

2.05

.538

Beta

-.009

.355

.179

.137

-.038

-.060

-.026

.094

.002

t-test

-.227

8.310

4.070

3 .426
-.686

1.111

-.492

2.090

.059

Sig.
Level

.821

.000*

.000*

.001 *

.493

.267

.623

.037*

.953

R = .557
R2 = .310
Adjusted R2 = .297

* Significant at the .05 level
F = 24.272
p = .000

 

would have a greater propensity towards supporting the implementation of BMPS than

would being in the individualist cluster. The locus Of control indicator was found to be

111

the strongest predictor Of whether an individual will support BMPS or not. For the final
indicator in this set of variables, personal responsibility, it was found that if a person felt
responsible for the health of local waterways then they were more likely to support the
implementation of BMP, although this was the weakest predictor of the significant
variables.

The model had an R of .557 and an adjusted R2 of .3 1. These figures, although
below the range that Sutton (1998) reported in his meta-analysis, still provide support for
the argument that environmental intention to act is primarily influenced by the
psychological underpinnings of worldviews. The psychological indicators were the most
important determinants of the intention to support implementing BMPS. Not only do all
the indicators have significant effects on the intention to support in the model, but the
largest standardized regression coefficients was locus of control (B = .355).

When the sample was subdivided by the clusters and the regression model
applied, there is only a small difference in the model’s predictability (Table 19).
Specifically, the model better predicts the individualist cluster (R = .563) than it does the
egalitarian cluster (R = .555) and with fewer variables. The models for both clusters were
significant with the F — test for individualists being 14.3444 (p = .000) and for
egalitarians it was 11.926 (p = .000).

In order Of importance, the variables Of LOC, MORAL and RESIST are the
significant predictors Of an individualist member's intention to support the
implementation Of BMPS. For egalitarians the significant predictor variables, in order Of
importance are LOC, MORAL, EDUC and PERSONAL. Note that there are differences

in the list Of variables that are effective in predicting the intention to act for each cluster.

112

Table 19: By Cluster Regression Estimates for the Intention to Support BMP
Implementation

Individualist Egalitarian
b Beta b Beta

KNOWALL Knowledge OfIssues&Act1on -.039 -.067 -.009 -.019

Strategies
1 - no knowledge
5 — knowledgeable

MORAL Moral Imperatives .352 .322a .259 .245m
0 — Low moral priority
5 — high moral priority

RESIST] Resistance to Change -.]72 .068c .020 -.017
O —— low resistance to change
5 -— high resistance to change

LOC Locus Of Control .290 .324“I .256 .272”|
1 — low internal control
5 — high internal control

PERSONAL Personal Responsibility .001 .018 .104 .122c
1 - low sense of personal responsibility
5 - high sense Of personal responsibility

AREA Length of Residency in Area -.000 —.026 .000 .000
continuous

HOME Length Of Residency in Current House .000 .026 .002 -.057
continuous .

AGE Age Of Respondent .000 -.127 -.025 -.067
l — 18 - 24
4 — 45 - 54
6 - 65 years of age and up

EDUC Education Level .030 .042 .009 .198"
1 - less than high school
3 - some college
5 - post graduate degree

INCOME Household Income .000 .003 .003 .014
1 — less than $10,000/year
6 - $40,000 — $49,000
9 - $70,000+/year

 

R = .563 R = .555
R2=.3l7 R2=.3O8

Adjusted R2 = .295 R2 = .282
F: 14.344 F: 11.926
p = .000 p = .000

 

a: Difference Between the Clusters is Significant at the .001 level
b: Difference Between the Clusters is Significant at the .0] level
0: Difference Between the Clusters is Significant at the .05 level
There are also similarities in that the variables Of LOC and MORAL are common to both

clusters. Both LOC and MORAL beta weights are stronger for individualists than for

113

egalitarians. These findings suggest that the clusters do indeed think differently about

their watershed and by extension, how it should be managed.

Non-Response Results

TO examine the somewhat low response rate, twenty phone contacts were made to
determine why the surveys were not completed and returned. The most common reason
for non-response was that the individual was “too busy.” Four respondents were no
longer at that mailing address. Three people felt they were too Old to participate. Two
respondents indicated they did not know enough to participate and the remainder Of the
pe0p1e did not respond for various other reasons. Not knowing enough about watershed
issues to adequately answer the survey was also one of the reasons given for refiasal when

trying tO recruit participants for pre-testing.

Study Limitations

Before advancing to the conclusions Of this study, there are several limitations
that should be acknowledged. It is important to address these limitations to fully
understand the conclusions and recommendations that have come about as a result of this
research.

First, the study chose to use the fairness heuristic in order to capture the
interaction between respondent attitudes and fairness judgments. Although use of this
instrument enriched the analysis in terms Of respondent worldviews, it weakened the
subsequent analysis between attitudes and the other psychological study variables. That

is, the attitudinal factors Obtained were less defined than one might like, although they

114

still met normal conventions. Also, there was some collinearity between the other study
variables but this is Often the case in psychological studies.

Second, personal responsibility was used as a proxy for the social norm variable
in the theory Of planned behavior. Although personal responsibility is a social norm
(Geller, 1995; Guagnano, 1995; Schwartz, 1970, 1977) it may be that it is two narrow a
construct to adequately capture the nuances that Ajzen and Fishbein (1977) envisioned
the construct to mean. Furthermore there was some collinearity with the other study
variables Of moral imperatives and LOC.

Another limitation Of the study was the ability to capture the situational factors
surrounding the implementation Of best management practices. The details associated
with BMPS such as size and location if it is a structural BMP or the type of payment
vehicle for both structural and non-structural BMPS was not reflected in the
questionnaire. This means that a respondent’s intention to support the implementation Of
a BMP may change once they are in possession of all the facts. Resource managers need
to be aware of the potential impact Of these details and not rely solely on the conclusions
drawn by this study.

Finally, the actual watershed used in the study, Sycamore Creek, has been the site
for state extension programs since the late 1980s. The previous efforts to inform and
educate the public may have influenced respondent knowledge, attitudes, locus Of control
and their sense Of personal responsibility that in turn influenced their intention to support
BMP implementation. Managers from other watersheds reading these conclusions and
recommendations need to recognize that each watershed is unique and not simply assume

it will be the same in their watershed.

115

CHAPTER V:

CONCLUSIONS, DISCUSSION AND RECOMMENDATIONS

Conclusions and Discussion

There has been a limited amount of research using psychological analysis to
predict future behavioral intentions towards watershed management programs. This study
sought to take one psychological research instrument (the fairness heuristic) and one
psychological theory (the theory Of planned behavior) and apply them in the context Of
watershed best management practices.

The study set out to determine groups of people who held similar worldviews on
how to manage their watershed. Analysis Of respondents discovered two attitudinal
factors, MORAL IMPERATIVES and RESISTANCE TO CHANGE. Using these two
factors, respondents were clustered into two groups, individualists and egalitarian,
borrowing from Dake’s (1991, 1992) worldview typology. The analysis further defined
the characteristics of these groups in terms of demographics, fairness evaluations and
psychological variables. When the groups were broken down by these variables and
interpreted through the lens of grid/ group nomenclature, it was concluded that the
worldviews represented by these names indeed fit the respondent groups. The last stage
of the analysis sought to determine if the two groups of respondents differed in their
support for implementing BMPS and what study variables best predicted their support. It
was concluded that they did differ on both the level of their support and the types Of
programs preferred. Also, the clusters used different study variables in making these

determinations.

116

The use Of the fairness heuristic and the theory Of planned behavior were used tO
examine the effects of psychological variables on the intention to support BMP
implementation. The use Of these different instruments helped to reinforce the difficulty
Of separating the procedural justice aspects Of decision-making from those associated
with attitudes and outcomes (Folger, 1996; Syme et al., 1999). Still, the combination Of
these approaches yielded many important findings. These discoveries are discussed
below.

Exploratory factor analysis Of the fairness heuristic yielded two attitudes held by
respondents in terms Of managing their watershed. These were labeled MORAL
IMPERATIVES and RESISTENACE TO CHANGE. The former factor seemed similar
to the findings that people believe in the “rights Of the environment” found in the
Australian water allocation studies (Syme et al., 1999). This bodes well for resource
managers in that they can focus on those similarities when working with the community
to formulate policy. Regarding the resistance to change factor, it seemed to indicate a
desire for minimal government and the inalienability of property rights both Of which are
tenants of libertarian doctrine (Wenz, 1988). Again, this might help resource managers
when they are working with the community. If the pOpulation is in agreement about
desired outcomes (i.e. clean water) then it is only the ways in which to achieve this that
are contentious. Knowing there is a large portion of the population who will be resistant
to certain types of change, the manager can then suggest Offering a suite Of
implementation Options aimed at cleaning up waterways. Concerning the other items in
the fairness heuristic, they were only modestly correlated which suggests that people

think about these items somewhat independently.

117

The cluster analysis based upon these two attitudes showed two clusters existed in
the study population. These were labeled individualist and egalitarian. The individualist
and egalitarian groups Of respondents were relatively close in terms of their MORAL
IMPERATIVES attitude. This is encouraging because it suggests that people generally
agree on a common set Of ethical principles upon which to decide how to manage their
watershed (Seligrnan et al., 1994). The two groups were not as close in their attitude Of
RESISTANCE TO CHANGE, with the individualist group indicating they were more
resistant than their egalitarian counter parts. If one assumes that part Of a person’s
resistance to change is somewhat related to “fear Of the unknown”, then the implications
Of this finding for the watershed manager is a need to anticipate the consequences Of
proposed programs and communicate these effectively to watershed residents.
Furthermore, it can be anticipated that a portion of the population will support programs
that have the least impact on themselves, regardless of program effectiveness.

In order to further understand and confirm the worldviews held by the two groups
Of respondents, the study variables were analyzed at both the sample population level and
at the cluster level. These yielded numerous interesting results beginning with the
demographic profiles. Beginning at the sample population level, it was shown that the
sample had obtained higher levels of education, earned slightly more money and had
fewer children than did the general pOpulation Of homeowners in Ingham County. Based
on previous research stating that young, educated liberals were more likely to possess
pro-environmental attitudes, (Arcury & Johnson, 1987; Arcury et al., 1986; Christianson

& Arcury, 1992) we might wish to conclude that selection bias is an issue for the study.

118

Fortunately, subsequent analysis Of the groups revealed substantial differences in
responses suggesting that this effect was minimal.

The individualist cluster was found to be Older, less educated, have a lower
household income, fewer children and resided in the area and their home longer. Relative
to these findings, it was shown that the egalitarian cluster was younger, more educated,
had a higher household income and resided in the area and their home for shorter periods
Of time. The multiple regression analysis indicated that of these variables, only the level
of education for egalitarians was a significant predictor of the intent to support BMP
implementation.

There are two conclusions to be drawn from these findings on demographic
variables: (1) the profiles Of each group may help a watershed manager identify which
group an individual or group of individuals is likely to be associated with, but that (2) the
results only partially supported previous research suggesting that younger, more
educated, urban dwelling liberals are more likely to engage in pro-environmental
behavior (Arcury & Johnson, 1987; Arcury et al., 1986; Christianson & Arcury, 1992).
Therefore, demographic variables may only be useful in helping to identify individuals
who possess similar worldviews and the type of interventions to use.

Those fairness heuristic items in which there appeared to be the greatest
variability in the population sample related to human use. There was a significant spread
in thoughts about water and water pollution in an economic context. In general, most
people were neutral on whether people should receive compensation for programs that
hurt an individual’s livelihood, but were more variable about using programs to

maximize the local economy. Furthermore, this variability carried over into using

119

cost/benefit analysis to solve water pollution problems and to trading Off different parts
Of the environment for human use. The variability Of these items suggests that
respondents were unsure about viewing water resources in economic terms and that there
may be other considerations that entered their reasoning.

The most highly skewed responses involved moral Obligations (to those
downstream), community voice, non-economic value and distributive justice. The strong
desire by respondents to adhere to these concepts when managing the watershed seems to
indicate they held a rights based viewpoint, as is associated with a universal value
orientation (Axelrod, 1994). This has immediate implications for water pollution trading
programs. If the public has the perception Of clean water as a right, then they may not be
receptive to economic programs, especially if the programs are not understood. At the
very least, it can be anticipated that there will be conflict between environmental rights
and other constitutional or legal rights (Montada & Kals, 2000).

From this point forward in the analysis, the grid/ group nomenclature associated
with Dake’s (1991,1992) worldviews was used as a lens for interpretation. Specifically,
the social relations surrounding individualists are hypothesized to hold the myth Of nature
as benign so that if pe0p1e are released from artificial constraints there will be few limits
to abundance for all with surplus tO provide compensation for any hazards created in the
process (Dake, 1992). Deregulation is the rational management strategy in this low-
grid/low-group culture because individualists value decisions stemming from personal
judgments rather than collective control (Rayner, 1988a). The term individual in this
context refers tO social beings generating and stabilizing a form of social relations and

institutions that defend their freedom to bid and bargain in self-regulated networks with

120

few prescriptions (Dake, 1992). On the other hand, egalitarian groups are those with
strong group boundaries (high group), but with prescriptions that do not vary by rank and
station (Dake, 1992). They believe the myth that nature is fragile and because they view
nature as being ephemeral it justifies their precautionary approach to management. The
egalitarian group prefers approaches to management that foster equity of outcomes
(Rayner, 1988b) and are hypothesized to frame natural resource issues in ethical terms.
When egalitarian social relations prevail, they are ofien critical Of the institutions
responsible for natural resource management and in the extreme form can be strict
preservationists (Dake, 1992). The data analysis findings did support that the two study
clusters fit these social patterns and the findings will be subsequently discussed. It should
be noted that measures of a worldview are related to personality traits and personal values
as well as to social attitudes and policy preferences and as such are not meant to be
mutually exclusive categories.

Ranking the responses to the fairness heuristic at both the sample population and
cluster levels was the next step in the analysis. Consistent with the highly skewed
responses, the number one ranked item by the sample population was that “water has a
value other than its dollar value.” The individualist cluster also ranked this item first
while the egalitarian cluster ranked it second. Question 22f directly asked respondents to
consider whether the environment has the same rights to water as people; it was ranked
ninth by the egalitarian cluster and tenth by the individualist cluster. Therefore, the same
conclusion that was already made about respondents awarding the environment at least

some rights can be drawn.

121

Either the second or third ranked item by all respondents was the desire for
polluter pay programs. This stated preference indicates the public Operates in a command
and control management paradigm and wish to internalize the extemalities of pollution.
The limitations Of command and control programs have been well documented (Montada
& Kals, 2000; Tenbrunsel et al., 1997), but the desire by the public for these types of
programs would seem to indicate a lack of comprehension about their limitations. Also,
consistent with water having a value other than its dollar value, the desire for command
and control programs can be interpreted as support for rights based viewpoints: people
who pollute and effect others in the community should have to make restitution (Montada
& Kals, 2000).

The finding that some respondents were more concerned about forgoing economic
gain for the sake of a cleaner environment than they were about process considerations is
contrary to the Australian water allocation studies (Syme et al., 1999). In the Australian
studies, it was found that fair decision-making processes were of paramount importance
and that economic arguments were Of a lesser importance to process considerations
(Syme et al., 1999). In this study, more than half the respondents (individualists, N =
320) thought economic arguments more important than process consideration.
Furthermore, their moderate support for the long-term health of the environment if it
reduces short-term profits (Q22i = 65.1%) indicated a desired to for polluter pay policies
that have minimal immediate economic impact. These frndings are consistent with the
individualist view that if artificial constraints (regulation and enforcement) are removed,
there will be few limits to abundance for all and that this will more than compensate for

any hazards created in the process (Rayner, 19883). Conversely, egalitarian support for

122

these items can be interpreted as being consistent with advocating prescriptions that do
not vary by rank and station (Dake, 199]). The implications of this finding for watershed
managers is that they need to be particularly aware Of program impacts and need to find
innovative ways Of firnding their initiatives.

The egalitarian and individualist clusters differ significantly on the procedural
justice consideration of whether the public should be involved in decision-making. The
individualist cluster does not agree as strongly (73%) as the egalitarian group (87 .3%)
that the public should be involved. This is particularly interesting since both clusters dO
not differ (statistically) on people’s right to have their say (Q22a) and accepting the
decision of a fair process (Q22n). This would seem to indicate that there is a difference in
what the two clusters consider a fair process. If individualists wish to have their say but
not be as involved in the decision making process, then their criteria for a fair process
probably includes less up front public involvement. Again, this is consistent with the
individualist worldview: they seek to generate and stabilize a form of social relations and
institutions that defend their freedom to bid and bargain in self-regulated networks with
few prescriptions (Thompson, 1992). This low group/low grid finding would suggest that
individualists would support the current American legal system where they can voice
their concerns after a problem is perceived.

The two groups also differed significantly on whether the government should be
involved in decision-making processes with the individualist group indicating a greater
desire for minimal governmental intervention. Therefore, watershed-planning initiatives
should strive to be seen as non-governmental agencies and autonomous Of political

affiliation. Furthermore, planning bodies need to recognize the necessity of soliciting

123

help from all facets Of the community, not just those willing to participate (i.e., most
likely egalitarian groups). A trusted spokesperson that can voice individualist concerns
and communicate the initiative’s intentions back to all factions of the community will go
a long way towards having programs gain public acceptance.

Having established the differences between the clusters based on the fairness
heuristic responses, it is necessary to point out that more than two-thirds to three-quarters
Of the sample population agreed with these statements. Watershed managers need to
recognize that there are more similarities than differences between these groups. In most
cases the differences that do exist are more a matter of degree and are concerned with
how to proceed with implementation rather than being irreconcilable differences. As will
be shown, this trend towards the presence of group differences within the context Of
overall general agreement is a common element to most Of the study variables.

Comparisons Of this study’s conclusions to those found by the Australian studies
indicate some differences. Although both agreed that a large portion Of people believe in
the “rights of the environment” and its preservation for a range of uses for future
generations (Syme et al., 1999), they differed on their value Of process. That is, the
Australian study put process above economic concerns while at least half of the
respondents in this study felt the Opposite. Furthermore, not all eight items common to
both studies and as ranked by the Australian studies appeared in this study’s rankings and
those that did were ranked differently. These differences probably indicate there are
significant cultural distinctions between the countries on how they view and manage the

environment and in particular, water resource management.

124

 

 

A country’s laws reflect its culture and a brief look at the differences between the
United States and Australian judicial systems may help explain the differences noted
above. In the United States there is an “absolutist approach” with a heavy emphasis on
the supremacy Of the law, particularly the Constitution, and an attempt to avoid
substantive issues (Bossehnann, 1997). On the other hand, Australia adOpts a “balance of
interests” approach that attempts to weigh all the various interests (Bosselrnann, 1997).
Specifically, Ministerial discretion is awarded to environmental agencies on when and
when not to institute certain procedures (i.e. environmental impact assessments) (Meyer,
1996). The public then has the right to appeal ministerial decisions. Consequently there is
a lack Of judicial review and an emphasis on collaborative solutions in Australia (Meyer,
1996). Conversely, in the United States the importance of the law creates an atmosphere
of scientific dominance where scientific results are used to support varying positions
surrounding an issue and leaving the courts to sort out solutions. This adversarial system
Of checks and balances pits individuals and institutions against one another and Offers a
possible explanation as to why this study found process consideration ranked lower than
economic considerations by half the respondents.

There is an implicit belief that information causes pro-environmental behavior
(Eden, 1993). This research supports previous studies (Beckwith & Ray], 2002; Hwang et
al., 2000; Steel et al., 1990; Tarrant et al., 1997) that have found the impact Of policy-
relevant knowledge on behavioral intent to have little bearing. Even though there was a
statistical difference between the knowledge Of action strategies between the two clusters,
it was not a determinant in predicting the intention tO support BMP implementation. The

partial correlations between attitude, locus Of control and personal responsibility with

125

intent to act changed little when knowledge was controlled for. This means that
knowledge is mediated, almost completely, by these variables. Interestingly, respondents
cited a lack of knowledge as one reason for not wishing to participate in the study. If
people are not making decisions based on what they know (or aware of), then other
variables must be influencing their attitudes. There two possible explanations for this
finding which more than likely are interdependent. First, the measures used tO indicate
knowledge in this study could be characterized as awareness about both the issues and
action strategies. Although awareness does constitute a form of knowledge, it is arguably
at a surface level and therefore may not invoke cognitive interaction with other study
variables by respondents. The second possible explanation for the weak relationship
between knowledge and the other variables might be due to Steel et al.’s (1990)
suggestion that Americans are highly influenced by ideological and environmental value
orientations (Steel et al., 1990). That is, the values the American people assign to
environmental issues and the management decisions they support may have more to do
with political ideology than how they value the environment.

The issue Of transmitting knowledge to the public about BMP implementation is
an important one. It was shown earlier that there is a real need to educate the public about
the limitations Of command and control regulations and the possibility of other types of
programs. Yet, there exist two distinct ideological groups within the watershed each with

their own worldview of the environment. This suggests that not only is the content Of the
message provided to the different constituencies complicated and problematic, but so is

the issue of how to get the groups to even examine the message in the first place.

126

The theory Of planned behavior was found tO provide quite an accurate prediction
Of the intention to implement BMPS. In accordance with the theory, attitudes towards
watershed management, subjective norms (personal responsibility) and perceptions Of
behavior control were significant determinants of the intent to support the
implementations Of BMPS. The successful application of the theory of planned behavior
to the intention to support the implementation of BMPS is consistent with other research
in which the theory effectively predicted intentions. The correlation for the sample
pOpulation was .56 and for the individualist and egalitarian groups .56 and .55
respectively. These results are on the low end Of previous studies reported in Sutton’s
(1998) meta-analysis. There are a few reasons why this might have occurred.

The first reason for the low (but still significant) correlations could be due to the
fact that attitudinal factors derived from the exploratory factor analysis were not as
tightly defined as if they had been determined a priori. Second, the variable Of personal
responsibility was used as a proxy for social norms. Although used in this way in the past
(Geller, 1995; Guagnano, 1995; Schwartz, 1970, 1977), personal responsibility is only
one aspect Of social norms and may be tOO closely associated with both locus Of control
and the attitudes used in this study. Further evidence of these possible limitations comes
from the reduced partial correlations and collinearity between the attitudinal factors
(MORAL IMPERATIV ES and RESISTANCE TO CHANGE) and the locus of control
(LOC) and personal responsibility variables. Having couched the results in this context,
the finding that egalitarians had a higher internal LOC than individualists and that there
was collinearity between the study variables is consistent with Eden’s (1993) supposition

that feeling responsible is often dependent upon a belief in efficacy. That is, individuals

127

feel they can have some impact through their pro-environmental behavior as well as
being able to choose what they undertake. According to Eden:

“Where efficacy is not perceived, responsibility is weakened because, without

impact, individual acts are futile. It becomes, for most, pointless to ascribe

responsibility to the self to undertake behavior which has no effect.”

(Eden, 1993, p. 1748)

This link with impact rather than moral Obligation has a utilitarian ethos to it
(Eden, 1993) since responsibility is not initiated because of morality, but because of
perceiving the ability to limit environmental damage. This is important to this study
because in the multiple regression analysis, personal responsibility was not a determining
variable in predicting the individualist cluster’s intent to implement BMPS. Eden Offers
this possible explanation:

the perceived impact is significant and this supports Schwartz’s (1970, 1977)

emphasis on the individual’s awareness of the consequences of his or her actions

as contributing to responsibility ascription. Contexuality is therefore influential in

that it affects how efficacy is perceived. Efficacy is reinforced for activists by

their group situation, underlying the necessity of setting individual pro-

environmental behaviors within social circumstances.”

(Eden, 1993, p. 1749)

Therefore, even though individualists saw it in their control to dO something about the
problem Of degraded water quality, they were unlikely to act upon it (relative to the
egalitarian cluster) because they did not feel as responsible. The relative lack of personal
responsibility felt by the individualist cluster was probably a result of their cultural
relations that reinforce their belief that the environment is benign and does not need their
attention. The implication Of this insight for the watershed manager is that they must

create and help maintain social organizations that are supported by all facets Of the

community so as to instill a sense Of personal responsibility across all groups.

128

The other difference between the variables that helped predict the intent to
implement BMPS was that education level was a determinant for egalitarians but not for
individualists. One possibility for this result mentioned previously is that advanced levels
of education are Ofien associated with more liberal worldviews (Arcury et al., 1986;
Christianson & Arcury, 1992), a variable nOt measured in this study. Still, there is
evidence that this could be true. Recall in the analysis of the clusters, the egalitarians that
are assumed tO exhibit more liberal tendencies were also more likely to support all the
BMP Options. These results lead to the conclusion that simply possessing knowledge
(awareness) about the issues and the implementation Of BMPS is not sufficient to get
people to support these types Of policy initiatives.

Again it was shown that there were group differences, but that generally the study
population was in agreement in terms of their psychological orientations towards the
BMP implementation. The correlations between the study variables Of attitudes, LOC and
personal responsibility were all positive and substantial and both the knowledge
categories were not significant. SO even though there were statistical differences between
the individualist and egalitarian clusters, their mean values for each Of the respective
scales indicated they both had a positive moral imperative attitudes, a negative resistance
to change attitudes, positive LOC and personal responsibility.

There were some significant differences between individualists and egalitarians in
terms of the types of BMP programs they preferred after their initial agreement on fines
for polluters. Specifically, individualists sought voluntary and public information and
education programs as their second and third choices and were less supportive Of all the

programs than the egalitarian cluster. This finding is consistent with their low grid

129

orientation, their general desire for fewer prescriptions and being more resistant to
change. The resistance to change attitude involved issues such as minimal governmental
intervention (i.e. more freedom) and the property rights. Therefore, it would seem
individualists exhibit integrative or “trade-Off" reasoning (Peterson, 1994) as is associated
with a value pluralism approach (Sniderman et al., 1991; Sniderman & Tetlock, 1986b;
Tetlock, 1986) to decision-making. Recall, this is where all their core values are highly
and equally prized in the decision-making frame of reference.

Egalitarians preferred increased enforcement and stricter regulations in terms Of
their policy preferences and generally indicated greater support for all the policy
categories. This conflicts with their low grid orientation but is consistent with their desire
to create an equitable playing field within the community. One might expect these policy
choices as they are seen as being generally valid and guarantee more equality for
members of the community (Montada & Kals, 2000). The egalitarian’s frame for
decision-making is more focused on the environment and therefore more monastic. As
was seen previously, they were also more Open to procedural justice initiatives. These
Observations suggest that an egalitarian’s frame of reference for decision-making
involves both self-ascription and having an implicit assessment Of efficacy (Eden, 1993)
compared to their individualist counterparts.

Neither cluster indicated a preference for the zoning Of Open space or subsidies
with these options consistently being ranked last. The strong economic values that came
forth in the analysis are probably responsible for the negative evaluation of subsidies,
while lack Of support for increased zoning restrictions is more than likely a result Of the

strong support for property rights. Once more we see that despite the group differences,

130

there was relatively strong support for five of the seven proposed BMPS by all the
respondents.

In general, invoking the concept Of framing refers tO the perceptual lenses,
worldviews or underlying assumptions that guide communal interpretation and definition
Of particular values. Use Of the concept of framing reflects the growing acknowledgement
that how societies view the environment is not simply given by nature, but reflects
collective moral choices about the legitimacy of the myriad of intersections between
natural and human systems (Bardwell, 199]; Miller, 2000). This research has
demonstrated that the worldviews of the two clusters of respondents, individualists and
egalitarians, influenced their assessments of environmental change including the terms Of
participation, the range Of policy Options considered and the nature Of the political
intervention. These Often divergent worldviews indicate there is an “interpretive
flexibility” (Collins & Pinch, 1982) surrounding watershed management. It has already
been shown that at the macro scale a country’s laws can serve to establish context and
that previous efforts within the planning of Sycamore Creek watershed might have helped
establish local context but neither of these by themselves or in cooperation can explain
the subtle nuances surrounding the differences between the two cluster’s worldviews.
One needs to better understand the social histories and dynamics of the area in order to
achieve a deeper understanding of these groups. Obviously if there is general agreement
about the desired outcomes, but not on how to achieve it, then a more in-depth look at
how the groups frame the issues is needed so planning initiatives can move forward. It
may be simply that individuals do draw on political ideology (Steel et al., 1990) to frame

watershed issues but this may also not be the case. The Obvious implication of this

131

conclusion is that deliberation must occur in order to create policies that reflect local

context and make them acceptable to the community.

Implications for the Management of Sycamore Creek Watershed

With the growing tendency for the public tO become more involved in decisions
regarding the management of natural resources, there are practical implications to a
greater understanding of the relationships among public values, attitudes and knowledge.
First, attitudinal information can help managers understand the diverse sides Of watershed
management issues. Increasingly, the management Of viable natural ecosystems can
represent a multiplicity of public values (Bengston, 1994). Given that they must manage
natural resources in the public interest, managers must recognize the extent tO which
these values and value orientations drive public attitudes towards specific issues. In this
study, it was shown there were two groups Of peOple with different value orientations or
worldviews within the watershed and that the public is not simply monolithic in their
views Of how to manage watershed even if there is mutually agreed upon goal.

Second, a significant amount Of research in social psychology and natural
resource management has supported the notion that attitudes predispose or predict
behavior. Such behavior can take an active form, as in participating in decision-making
processes, or a more passive form such as support for specific watershed management
practices. This is important because many decisions regarding resource management are
being brought forward tO the public through ballot initiatives. Regardless Of the outcome
of these initiatives, it is apparent that watershed managers must understand the nature Of
public attitudes and the resulting behavior. Using the identified cluster’s worldviews as a

lens to anticipate public reaction to prOposed policies is one way for watershed managers

132

to understand public attitudes and behavior. For example, any proposed policy calling for
immediate implementation that had an associated cost for the landowner attached tO it,
and administered by a government agency, would probably be Opposed by the
individualist cluster. Conversely, any proposed policy that did not have equality Of
outcomes for all individuals would probably meet Opposition from the egalitarian cluster.
Both of these reactions can be anticipated regardless Of whether the groups agreed upon
and desired the same end result.

If justice considerations moderate the relationship between attitude and behavior
as it is theorized, then managers must also pay close attention to the processes in which
policy is deliberated. Fairness judgments are an estimation of whether the right mix Of
“fairness ingredients” have been incorporated into the process (Syme et al., 1999). Key to
the success of a project is the effective communication that the process in which
decisions were reached was both Open and transparent.

Understanding the nature of attitudes is complicated by the lack Of a relationship
between knowledge and behavior. The fact that knowledge is mediated by other
psychological constructs complicates matters for two reasons. First, worldviews and
attitudes were shown not to be dependent upon awareness of the issues or Of the action
strategies, thus confounding the issue of what to include in a persuasive communication
to people in the watershed. Second, if the direct path between knowledge and attitudes is
muddy, then either contextual factors or other value domains (i.e. political ideology) or
both must be having an influence. It is the managers job to uncover the degree to which
each Of these might be involved in individual decision-making and make process

adjustments to account for them so implementation goals can be set.

133

Recommendations for Future Research

This study produced two important findings: (1) that attitudes appeared to be
moderated by fairness judgments and (2) that the theory Of planned behavior was an
effective tOOl for predicting behavioral intent. There were, however, some ways the
results could be improved upon. First, the attitudinal factors need to be better defined.
This researcher is confident the two factors that were derived indeed capture the essence
of unique attitudes, but acknowledges there was collinearity between the other
psychological variables that with some effort may be minimized. Focus groups and/or
interviews may help in this regard while at the same time helping to better understand the
local context.

Second, knowledge was measured through the indexing Of several yes or no
questions. Although this is an acceptable method Of eliciting information, it is possible
that a more complex instrument might yield different results. Furthermore, this study
tried to break down knowledge into two components: knowledge Of issues and
knowledge of action strategies. This conceptual distinction in theory only yielded a minor
difference in terms of responses. For these reasons, there needs to be further research into
the structure of knowledge and its relationship with other psychological variables in
environmental decision-making.

Third, past studies have indicated that the American public may formulate their
attitudes based upon political ideology (Steel et al., 1990; Steel, 1996). Given the
apparent lack Of a relationship between knowledge and behavior and the weak
relationship between knowledge and the other mediating variables Of attitudes, locus of
control, and responsibility, it seems prudent for future studies to incorporate political

items aimed at measuring respondent ideology.

134

And lastly, although considerable effort went into capturing the situational factors
that were known to influence decision-making, this study was still limited in the specifies
that were introduced. Variables such as location, accessibility and payment vehicles are
all important factors not considered in this study. Future studies might wish to
incorporate these variables.

In all, the results Of this study serve as a starting point for better understanding Of
how people think about managing their watershed. The study has been especially useful

in isolating several issues needing attention when managing a watershed.

135

APPENDICES

136

Appendix A:

Survey And Response Rate Data

137

 

OFFICE OF
RESEARCH
AND
GRADUATE
STUDIES

    

246 unhistatim Building
Ea Lansing Michigan
48824-1046

517/355-21w
FAX: 51 71353-2976

MICHIGAN STATE
0 N l v E R s 1 T Y

 

June 26, 2001

TO: Scott G. WITTER
308 Natural Resources Building
MSU

RE: IRB if 00-567 CATEGORY: 2-G EXPEDITED

TITLE: PREDICTING SUPPORT FOR BEST MANAGEMENT PRACTICES IN THE RED CEDAR
WATERSHED BASED UPON INDIVIDUAL VALUES

ANNUAL APPROVAL DATE: OCtOPOI' 10. 2000
REVISION REQUESTED: June 8. 2001

REVISION APPROVAL DATE: June 25. 2001

The University Committee on Research Involving Human Subjects' (UCRIHS) review of this project is
complete and I am pleased to advise that the rights and welfare of the human subjects appear to be
adequately protected and methods to Obtain informed consent are appropriate. Therefore. the
UCRIHS APPROVED THIS PROJECT‘S REVISION.

This letter approves the addition of the mail survey.

RENEWALS: UCRIHS approval is valid for one calendar year, beginning with the approval date
shown above. Projects continuing beyond one year must be renewed with the green renewal form.
A maximum of four such expedited renewal are possible. Investigators wishing to continue a project

beyond that time need to submit it again for a complete review.

REVISIONS: UCRIHS must review any changes in procedures involving human subjects, prior to
initiation of the change. If this is done at the time Of renewal. please use the green renewal form. .
To revise an approved protocol at any other time during the year, send your written request to the
UCRIHS Chair, requesting revised approval and referencing the project's IRB# and title. Include in
your request a description of the change and any revised instruments, consent forms or
advertisements that are applicable.

PROBLEMS/CHANGES: Should either of the following arise during the course of the work, notify
UCRIHS promptly: 1) problems (unexpected side effects. complaints, etc.) involving human
subjects or 2) changes in the research environment or new information indicating greater risk to the
human subjects than existed when the protocol was previously reviewed and approved.

If we can be of further assistance, please contact us at 517 355-2180 or via email:

Web. numeral/1mm *

E—Mail: ucrihsOmsuedu

mama-am
0&me
Minuet

“Slim
WW

UCRIHS@pilot.msu.edu.
Sincerely, M
47/ M -
UC I
Ashir Kumar M. D. T3:§yé§:“%¥§" F93
Interim Chair UCRIHS ~21: w r. «IRES.
01.11" 1 0 200]

. SUleraErtEWALAPPu ATI
AK. bd . ONE oump pm 10300 0N
cc: Stephen Pennington ABOVE DATE TO CONTINUE

1900 Pepper Tree Lane

Lansing, MI 48912 a & a , ,,

(beg Lisa

138

STATE

MICHIGAN
u N l R s I T v

VE

July 7?, 2001
Dear Mers Stakeholder:

You are invited to participate in a study of the Red Cedar Watershed. The Red Cedar
Watershed Coordinating Committee is a partnership of State, county and local governmental
agencies together with the Department Of Resource Development at Michigan State
University. Our mission is to work together to better manage the biophysical and socio-
economic resources of the Red Cedar Watershed in a sustainable. fair, and cost-effective
manner. The primary goal of the prOposed project is the creation of a watershed
management plan that will be adopted by all community stakeholders. townships and county
governments within the watershed. In order to foster widespread adoptiOn and maintenance
of the best management practices we ask for your participation by completing and returning
the enclosed survey in the return addressed envelope provided.

Any information that is obtained in connection with this study will remain confidential. Only
the investigators will have access to the raw data and any reports will convey information in
the aggregate. Your privacy will be protected to the maximum extent allowable by the law.

Your decision whether or not to participate in this research component will not prejudice your
future relations with the Red Cedar Watershed Coordination Committee or Michigan State

University.

 

   
      

 

 

I O O O O O O O
: ‘\‘ If you have any questions now or during the process. please feel free to contact us at one of
: ‘ the numbers listed below. If you have any questions regarding your rights as a research
, " participant, please contact the University Committee on Research involving Human Subjects
0 fl (UCRIHS) chair David E. Wright at 355-2180. ‘
0
Your signature indicates that you have read the information provided above and are willing to
coulseEOF participate.
AGRICULTURE
AND NATURAL Thank you very much for your time and consideration]
RESOURCES
Mira
Develop-rod Participant
new Star um -
”ninja”... Dr. Scott G. wttter Dr. Mike Kaplowitz Stephen Pennington
East wig. man 319 Natural Resources Bldg 3113 Natural Resources Bldg 302 Natural Resources Bldg
mum Michigan State University Michigan State University Michigan State University
517ml Phone (517) 355—3421 Phone (517) 355-0101 Phone (517) 355-3415
FAX: 51713530994 Fax (517) 353-8994 Fax (517) 353-8994 Fax (517) 353-8994
www.mmodu witter@msu.edu kalowitz@msu.edu pennin32@msu.edu
UCRIHS APPROVAL FOR
THIS prolect EXPIRES:
OCT 1 0 2001
Wis-1W
SUBMIT RENEWAL
cadmium 139 ONEMClrl' Rafa???)

ABOVE DATE TO counuue

MICHIGAN STATE

UNIVERSITY

 

The Red Cedar River and its Streams

 

i
- N
I POW
3"" 1 Shran‘ r .,
C I _ _x Q7 _ _ _I’ """"""""" r """""""""
------------ lilk'l'" I (‘n L X” (
a,“ Haslett i
East Lansing '
Lanslng E . I
1 1,1..1In : ‘- ‘E I
M] "
her-:2“... 4’ Okcmos i g g '1
""""‘"’ ‘11,. Williamslon 2 i 1
r ”u. 5 4’
J, 1, "It
\ 1 °
C,» ‘c We- . : Illa
.fi ‘ i Fowlervllle
9
Red Cedar Rim ”'3'“ 3 g, ‘
_ 1
F 2‘
Finn: Lonqllt 40 miles ' 3 {E “In: How
‘72 to mlu "~ 3 I; "
’3” I! Tr,
. Ii
1
a the Grand River in l
' 0
1
Stockbridge I t and air
1
I

 

 

 

 

This is an Opportunity for you to provide information for local water

resource planning. Your input will help planners make informed water
quality decisions.

Thank you for your participation. This booklet contains several sections Of
brief questions that should take about 20 minutes to complete.

Please return your completed questionnaire in the enclosed envelope to:
Red Cedar Project - K, Department of Resource Development, 323

Natural Resources Building, Michigan State University, East Lansing,
MI 48824-1222

140

Section 1: Uses And Thoughts About The Red Cedar River
1) How often do you do the following activities in the Red Cedar Area? (Mark E

one response for each item)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

' 2 ti 3 More
Once a . o than 4
Never times a
year times a
year
, , . . Z -.. _1 year
a GO fishing [:1 Cl E [I
b Use river water for lawns or gardens [:1 Cl E] [:1
c GO swimming Cl C] Cl C]
d Irrigate crops [:1 E] C1 [I
6 Drain excess water into the river Cl E] El [:1
f Use river as dnnkrng supply for [:1 Cl E] E
animals/pets
Z Use area for nature walks or wrldl1fe E] E] El E1
vrewrng
h :3; )recreattonal boating (canoe, kayak, E El [1 El
i Use well water for household use E E Cl C]
Use area for hunting [:1 El 1] E]
k Other ( ) Cl [:1 El E1

 

 

 

2) In your opinion, how would you characterize the water quality of the Red Cedar

River and it streams? (Please mark E one)

[:1 Poor

CI Fair

B Good

Cl Excellent
[I Don’t know

3) Please read each statement below and indicate your level of agreement with each

statement. (Please mark E one for each statement)

 

 

 

 

 

 

 

activities that violate our water protection laws.

 

 

 

 

e 22 :1 $.13
8’ a 8' E i: 0:: 8
‘a E P = i a? a
8 .< 8 53- '4
It is my personal responsibility to protect our rivers
a and streams for other peOple even if they seem E] El Cl E] El
unconcerned.
b It is not my responsibility to ensure the well being of El
other species on earth.
C I am partly responsible for the degraded state Of our
local rivers and streams.
d It is my responsibility to inform authorities about

 

 

141

 

streams support.... (Mark E one response for each activity)

4) In your opinion, how important is it to you that the Red Cedar River and its

 

Not Somewhat
Important Important

Extremely
Important

Know

Don’t

 

Fishing [:1

E

1:]

Cl

 

Watering lawn/garden

 

Swimming

 

A drain for excess water

 

Water supply for livestock/pets

 

Nature appreciation

 

Recreational Boating

 

Hunting

 

 

DEEEDEEE

Other ( )

 

 

 

 

EEEDEEEE

 

DEEEEEEE

 

EEEEEEEE

 

 

5) Complete the following sentence with each of the statements below and indicate

your level of agreement with each. (Mark E one response for each complete

 

 

 

 

 

 

 

 

 

 

 

 

 

statement)
114)» individual actions would improve water g '3’ E 1 g > > g.’
qualityin local rivers and streams if I were to n: S of. E. "3 "3 s.
g ".9. g 3 8 8 ".9.
'< ""' '<
attend a community meeting that involves El [:1 E! El El
concern over our local streams and rivers.
buy resource conservation devices, such as low- [:1 CI E E El
flow faucet for my sinks and shower heads.
report someone who violates a law or laws that
protect our rivers and streams (e. g. illegal fishing, [3 [:1 Cl C] E
polluting) to the prOper authorities.
convince someone to sign a petition regarding E] El El E] El
an issue surrounding our rivers and streams.
...convince someone to buy household cleaning
and/or laundry products that don't harm the El E E CI [:1
environment.
convince someone to conserve water by not
running the water while brushing their teeth or El El E E E
shaving and/or installing a water saving devices.

 

 

142

 

Section 2: Water Quality Concerns

Experts have identified four concerns for the Red Cedar River and its
streams. Pollution control prograrrrs are addressing some of these concerns.
Other concerns require additional management practices.

 

2.1 Human and Animal Waste: Testing has found high levels of human

and animal waste matter in the Red Cedar River and its streams. These levels make it
unsafe to swim in the river 40% of the time. The problem of human and animal waste
pollution is being addressed by 1) stormwater plans that separate sewer pipes from
stormwater pipes, and 2) farm Operations adopting generally accepted agricultural

management practices.

6) Have you heard about your community’s combined stormwater overflow
plan?
NO [I
Yes [:1

7) Have you heard about farmers in your community adopting generally
accepted management practices?
No D
Yes 1:]

 

 

 

 

2.2 N on-Point Source Pollution (N PS): Non-Point Source pollution)

the result of such things as oil, gas, salts, fertilizers, pesticides, and other materials from
homeowners, industry and agriculture being washed into surface waterbodies by rain or
snowmelt. These pollutants are spread over wide areas and cannot be traced to a single
source. While not visible, NPS pollution degrades water quality, impairs fish habitat

and raises health concerns.

8) Have you heard about non-point source pollutants in the Red Cedar?
NO [:1
\ Yes [:1 J

143

 

 

Water Quality Concerns (continued)

 

(Blncreased Flow and Flooding: There has been increasing urbaD

rural growth around the Red Cedar River and its streams. The additional hard surfaces
(e. g. pavement) associated with urban and rural growth increases the amount and speedf
Of water entering directly into the river and streams. Water moving directly into

waterways destroys plant and animal habitat and causes flooding.

9) Are you aware of flooding problems in the Red Cedar or its streams?
NO [:1
Yes D

 

 

\ J

arosion and Sedimentation: Erosion is the process of water washing

soil into waterways. Sedimentation is the settling of soil particles on the bottom Of a

 

water body (e.g. rivers, lakes). The removal of vegetative cover (e.g. forest, grass,
brush, etc.) increases soil erosion and sedimentation. The murky water seen during and
afier storm events is a result of erosion. Too much sedimentation can result in the loss
of plant and animal habitat. One cause of erosion and sedimentation is construction in

developing areas.

10) Are you aware of erosion prone areas of the Red Cedar River or its streams?

No D
Yes 1:]

 

 

144

 

Section 3: Best Management Practices.

Experts recommend combinations of practices to improve water quality in the Red Cedar
River system. The best management practices (BMPS) below control non-point source
pollution, flooding and erosion. Other programs address animal and human waste
concerns. All BMP projects include education programs for landowners, builders and
Others in the community. Some BMPS use upland areas; others use lowland areas near

the river and streams; while still others use land along stream banks.

UPLAND BMPS

 

Dry Basins: Dry basins are designed to
hold runoff water. A dry basin allows water to
seep into the ground and slows runoff into
drainage systems. They empty after a storm

event and are dry most of the time. Dry basins

 

 

 

 

have a limited ability to remove soil particles

and pollutants.

Plans with MANY dry basins would have dry basins throughout available areas.
Plans with SOME dry basins would have dry basins in only the most critical areas.

11) Have you seen a Dry Basin in your community? E No El Yes El Don’t know

 

 

 

Wet Ponds: Wet ponds have a permanent
pool of water and keep most stormwater runoff
on site until it evaporates or seeps into the
surrounding soils. Soil particles and some

pollutants may settle in wet ponds

 

 

 

 

Plans using MANY wet ponds would have wet ponds throughout available areas.

 

Plans using SOME wet ponds would have wet ponds in only the most critical areas.
12) Have you seen a Wet Pond in your community? CI No I] Yes [:1 Don’t know

J

 

145

LOWLAND BMPS
r 1

Wetlands: Wetlands refer to areas with

 

 

wet soil and the plants and animals that live
there. Examples Of wetlands include wooded

wetlands and marshes. Natural and man-made

 

wetlands may filter pollutants, slow water

 

 

 

flow and reduce flooding. Wetlands may be

found near to rivers and streams.

Plans with MANY wetlands would have wetlands in almost all available areas.

Plans with SOME wetlands would have wetlands in only the most critical areas.

 

 

13) Have you seen a Wetland in your community? E] No El Yes Cl Don’t know
L J

 

 

Filter Strips: Filter strips are areas of
land 5’ to 30’ wide near rivers and streams.
Grasses, shrubs or trees can be used for filter

strips. They allow stormwater to seep into the

 

soil, reduce erosion and slow water entering

 

 

 

rivers. Filter strips trap some pollutants.

Plans with MANY filter ships would have strips in almost all available areas.
Plans with SOME filter strips would have strips in only the most critical areas.
14) Have you seen a Filter Strip in your community? [:1 No [:1 Yes CI Don’t know

 

 

L J

146

STREAM BANK BMPS
r 1
Rip Rap: Rip Rap are large stones placed

 

 

along stream banks and stormwater inlets to
protect them from flowing water. Rip Rap

slows down water flow and reduces stream

 

bank erosion and sedimentation. Rip Rap does

 

 

 

not usually remove pollutants other than erosion.

Plans with MANY rip raps would have rip rap along almost all available stream banks.
Plans with SOME rip raps would have rip rap along only the most critical

stream banks.

 

 

CS) Have you seen Rip Rap in your community? El No El Yes [:1 Don’t knovD

 

 

 

W

Streambank Naturalization:
Streambank natural-ization uses native
grasses, plants, trees, rocks, and tree stumps to

rebuild the banks Of streams or rivers.

 

 

 

Streambank natural-ization Slows runoff, traps

 

pollutants and sediments and allows water to seep into the soil.

Plans with MANY naturalizations would have it along almost all available stream
banks.

Plans with SOME naturalizations would have it along only the most critical stream
banks.

16) Have you seen Streambank Naturalization in your community?

El No D Yes [I Don’t Know

 

 

147

Section 4: BMP Combinations
Additional BMPS Needed

 

fl

adequately protect water quality. Additional BMPS such as dry basins, wet

(The current management practices in the Red Cedar River area do no

ponds, filter strips, wetlands, rip rap and stream bank naturalization will
improve water quality. There are areas available for additional BMPS in the
Red Cedar River area.

 

 

 

 

in J
Combining BMPS
r n

Individual BMPS are combined together to improve water quality. Water
resource experts recommend different combinations of BMPS. For the Red
Cedar River and its streams, several different combinations can improve the

Red Cedar’s water quality so that it will become swirnable and fishable.
L J

 

 

Achieve Same Water Quality

f R
BMP combinations are known as plans. Each plan suggested by a panel of

 

experts will help achieve the same water quality for the Red Cedar.
However, each Of the plans differ in the particular BMPS they use. That is,

each plan uses different levels of the six possible BMPS discussed above in

 

 

éwtron 3. J

Same Cost to Community
r 1

 

Each of the proposed plans is estimated to cost the same amount to your
community. That is, each plan will result in the same water quality

improvements, cost the same to your community and use different BMPs.
L J

 

 

148

Section 5: Implementation Information

The use Of BMPS are Often part of larger programs to improve water quality and protect
natural resources. We would like your input on some complimentary elements and
approaches that may be used for improving water quality in the Red Cedar River and its
streams.

20) In your opinion, how supportive would you be of the following?
(Mark E one response for each item)

 

Would Might Would Strongly Do
Not Support Support Support Not
Support g, . gKgnow

 

Zoning requirements for some
a Open space to be preserved on Cl C] E C] E
undeveloped land.
Subsidies to landowners for
b environmentally friendly [:1 E El E El
practices.

Stricter regulations on activities
c that impact waterways during [:1 Cl C] E] El
develOpment.

d Fines for polluting. Cl C] E] Cl C]

 

 

 

 

Increased enforcement of E] El 1:] (:1 El
envrronmental regulations.

 

Public information and education [:1 [:1 El E] El
programs.

Voluntary programs to help
landowners adopt

environmentally fiiendly D D D D [:1
practices.

 

 

 

 

 

 

 

 

 

 

21) Mark your three preferred information sources about water quality issues in
the Red Cedar River and its streams?

Michigan Department of Agriculture

Natural Resource Conservation Service
Michigan State University (other than Extension)
Michigan State University Extension

Michigan Department Of Environmental Quality
Soil Conservation District

Drain Commissioners

Farm Bureau

County Health Department

Michigan Environmental Council

Local Newspapers

Broadcast Media TV/Radio

El Other ( )

CID

EEEEEEEEDE

149

Section 6: Attitudes and Beliefs

Designing and implementing water quality plans can be improved when managers
understand the Opinions of the people living in the area. To help give policy makers and
planners for the Red Cedar River a better understanding of how you feel, please answer

the following questions.

22) For the statements below, please share with us your opinion on how
much you agree with each. (Mark E one response for each item)

 

 

 

 

 

 

 

 

 

 

 

a a, 9. 2 a
at s s a is it: a
a E a = s s a
8 3< . 8 9;. 5<

a All members of the communrty have a rrght to then say E E E C] E
on 1ssues involvrng water management.
Everyone may have to make some personal sacrifices if

b we are going to have effective water resource El El [:1 Cl Cl
programs.
You cannot really solve water pollution problems by

c analyzing the costs and benefits in dollars. E] D D D D
EveryOne owns rivers and streams and therefore they

d should be managed for the overall public benefit. D D D D D

3 Those who pollute the most should pay the greatest El E] Cl C] I]
share of clean up and protection costs.

f {he envrronment has the same rrght to water as people 1:] E] D E] El

ave.

g Public involvement should. not be part Of the decrsron- Cl Cl Cl [:1 El
making process for managing water resources.

h Water pollution programs should be made tomaxrmrze Cl C] E E El
the overall economrc income of the communrty.
The long-term environmental health Of local waterways

i should be achieved even if it reduces short-term E] E 1:] El E1
business profits.

j Water has value other than its dollar value. [:1 Cl E E E

 

 

 

 

 

 

 

 

 

150

 

aerfiesrq
Alfiuorrg
aorfiesrq
113-linen
oarfiv
as: v

Alfiuorrs

 

The government does not need to be involved in
cleaning up and protecting water resources.

E
Cl
C]
E
E

 

If we are to clean up and protect our rivers, it should be
done so as to minimize conflict in communities.

[:1
El
[:1
E
El

 

 

 

 

 

Those upstream have a moral responsibility to look
m after the interests of those downstream. '3 D D D D
u If the decrsron—makrng process is farr, people should E] Cl C] E] El
accept 1ts decrsrons for addressrng water pollution.
If new water pollution programs hurt some peOple's
o livelihoods, they should receive compensation. D D D D D
p Groundwater (water under land) is the property of the [:1 Cl Cl Cl C]
landowner above.
q When rt comes to clean waterways, the envrronment E [:1 E Cl C]

should be a secondary consideration to peOple.

 

Saving waterways for the firture is more important than [:1 Cl [:1 E] El
making money now.

 

While some parts Of the natural environment are
s valuable and should be preserved, some are not so [:1 E] Cl [:1 El
valuable and should not be preserved.

 

t There rs not time to wart for exact envrronmental E] CI [:1 Cl C]
knowledge, we need to act now.

 

 

 

 

 

 

 

 

 

Section 7: Demographics

23) How long have you lived in the Ingham/Livingston County area?

years

24) How long have you lived in your current home?

years

151

25) Do you currently own or rent your home?

[:1 Own
[:1 Rent

26) Which category includes your age?

18 to 24 years
25 to 34 years
35 to 44 years
45 to 54 years
55 to 64 years
65 +

DECIDED

27) Please indicate your highest level of education.

CI Less than High School Graduate
El High School or GED

[:1 Some College

El College Graduate

El Post Graduate Degree

28) Which category includes your household’s gross income?

Less than $10,000
$10,000 - $14,999
$15,000 — $19,999
$20,000 — $29,999
$30,000 — $39,999
$40,000 - $49,999
$50,000 — $59,999
$60,000 — $69,999
$70,000 +

EEEEEEEE

Cl

29) How many adults (18 years of age and over) currently live in your
household?

30) How many children under 18 currently live in your household?

152

F eel free to use the back cover if you have any further comments.

Thank you for your time!

Please place the survey in the envelope provided and return it to:

Red Cedar Project, Department of Resource Development, 323 Natural

Resources Building, Michigan State University, East Lansing, MI

48824-1222

Comments:

 

 

Original number of surveys 1650
wave 1 wave 2 wave 3 (Choice Exp. Only)

undeliverable 21 1 5 10
Return to Sender (RTS) 15 8 10
Blanks 9 10 10
returns 423 185 Total 608
response rate (rr) 26.0% 11.3%
wave rr 26% 16%

response rate 0.3767

rr if RTS = undeliverable 0.3943

 

153

Appendix B:

Response Frequencies For Survey Questions

154

Standard Deviation and Variance for all

Valid N ' ' Mean Std. Deviation
1.5 1.6
1.3 1.57
1. 1.64
1.4 1.7
1.5 1.831
1.4 1.
2. 1.5
l. 1.65
1.5 1.74
1.4 1.69

3.
4.
4.41
2.
4.1
2.
2.5
2.5
3
3.
3.
2.5
2.6

3.3
3.6
4.
3.5
3.
3.7

-.51
.6
-.l
3.5
2.3
1.4
3.
1.
3.1
2.

—l

 

y—ag—sp—er—du—s—sg—s

O

155

Mean Std. Deviation
2.]
2.201
1.
l.
1.57
1.25
1.43
1.4
1.45

4.1 . .85
4. . .7
3.6 1.65
4. . 1.04
4 4 .8 .73
. 5 1.34
4. .11 1.2
3.2 . 2.09
3.9 . 1.25
4. . .
4.1 . 1.2
3. . 1.33
4.21 . 1.35
4. . 1.
3.41 . 1.91
2. . 2.45
2. . 2.861
4.1 . 1.3
3.5 . 1.88
3.7 . 1.91
37.995 539.352]
20.829 22.0846] 487.7300

 

156

Valid N ' ' Mean Std. Deviation Variance
l. 1. 2.13
4. 1.58 2.51
3.61 1.43 2

7.7 4.561 20.
2. 1.58 2.5
.7 1.83 3.38

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequencies
Qla Flow often do you go fishin r in the Red Cedar River?
Frguency Percent Valid Percent Cumulative %
Never 513 84.4 84.4 84.4
Once a Year 26 4.3 4.3 88.7
2 to 3 times a yr 29 4.8 4.8 93.4
> 4 times a year 17 2.8 2.8 96.2
no response 23 3.8 3.8 100.0
Total 608 100.0 100.0
lb THow often do you use water from the Red Cedar River for your lawn or garden?
Frequency Percent Valid Percent Cumulative %
Never 566 93.1 93.1 93.]
Once a Year 1 .2 .2 93.3
2 to 3 times a yr 9 1.5 1.5 94.7
> 4 times a year 9 1.5 1.5 96.2
no response 23 3.8 3.8 100.0
Total 608 100.0 100.0
1c THow Often do you go swimming in Red Cedar River?
Frequency Percent Valid Percent Cumulative %
Never 547 90.0 90.0 90.0
Once a Year 14 2.3 2.3 92.3
2tO3timesayr 11 1.8 1.8 94.1
> 4 times a year 11 1.8 1.8 95.9
no response 25 4.1 4.1 100.0
Total 608 100.0 100.0
1d [How often do you irrigate your crops with water from the Red Cedar River?
Frequency Percent Valid Percent Cumulative %
Never 573 94.2 94.2 94.2
2 to 3 times a yr 1 .2 .2 94.4
> 4 times a year 4 .7 .7 95.1
no response 30 4.9 4.9 100.0
Total 608 100.0 100.0

 

 

 

 

 

 

157

 

 

 

Qle Jl-Iow Often do you drain excess water into the Red Cedar River?

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
Never 559 91.9 91.9 91.9
2 to 3 times a yr 3 .5 .5 92.4
> 4 times a year 14 2.3 2.3 94.7
no response 32 5.3 5.3 100.0
Total 608 100.0 100.0

 

 

Qlf [How Often do you use the Red Cedar River as drinking supply for animals/pets?

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
Never 560 92.1 92.1 92.]
Once a Yr 2 .3 .3 92.4
2 t0 3 “ms 3 7 1.2 1.2 93.6
year
> 4 times a year 8 1.3 1.3 94.9
no response 31 5.] 5.1 100.0
Total 608 100.0 100.0

 

 

1g Il-low Often do use the Red Cedar River area for nature walks or wildlife viewing?

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
Never 165 27.1 27.1 27.]
Once a Year 90 14.8 14.8 41.9
2 to 3 times a yr 139 22.9 22.9 64.8
> 4 times a year 197 32.4 32.4 97.2
no response 17 2.8 2.8 100.0
Total 608 100.0 100.0

 

 

lh [How Often do you go recreational boating in the Red Cedar River?

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
Never 442 72.7 72.7 72.7
Once a Year 74 12.2 12.2 84.9
2tO3timesayr 43 7.1 7.1 91.9
> 4 times a year 26 4.3 4.3 96.2
no response 23 3.8 3.8 100.0
Total 608 100.0 100.0

 

 

Qli jI-Iow often do you use well water for household use?

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
Never 539 88.7 88.7 88.7
2 to 3 times a yr 2 .3 .3 89.0
>4 timesayear 4] 6.7 6.7 95.7
no response 26 4.3 4.3 100.0
Total 608 100.0 100.0

 

158

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q] j [How Often do you use the Red Cedar River area for hunting?
Frequency Percent Valid Percent Cumulative %
Never 540 88.8 88.8 88.8
Once a Year 10 1.6 1.6 90.5
2 tO 3 times a yr 15 2.5 2.5 92.9
> 4 times a year 17 2.8 2.8 95.7
no response 26 4.3 4.3 100.0
Total 608 100.0 100.0
Qlk [How often do you the Red Cedar River or surrounding area for other activities?
Frequency Percent Valid Cumulative %
Percent

Blank 585 96.2 96.2 96.2
Bicycling 1 .2 .2 96.4
'Biking Qx) l .2 .2 96.5
Bird Watching (4x) 1 .2 .2 96.7
Bird, wildlife watching 1 .2 .2 96.9
Cross-country skiing (4x) 1 .2 .2 97.0
Dogs go swimming (>4x) ] .2 .2 97.2
Drive over (4x) 1 .2 .2 97.4

xercise (4x) 1 .2 .2 97.5
Flood relief (4x) 1 .2 .2 97.7
Golfing (4x) 1 .2 .2 97.9
'Mountain biking (4x) 1 .2 .2 98.0
Other (3x) 1 .2 .2 98.2
Other (4x) 1 .2 .2 98.4
Photo opts (4x) 1 .2 .2 98.5
River trail for biking (4x) 1 .2 .2 98.7
Running, walking (4x) 1 .2 .2 98.8
Scenic/Bird watching (4x) 1 .2 .2 99.0
Sight seeing, hiking (1x) 1 .2 .2 99.2
Visit city overlooks (3x) 1 .2 .2 99.3
Wading (4x) 1 .2 .2 99.5
Walk doggsit by MSU (4x) 1 .2 .2 99.7
Walks at MSU (3x) 1 .2 .2 99.8
XX skiing, snowshoeing (4x) 1 .2 .2 100.0

Total 608 100.0 100.0

 

 

159

 

 

eams

Q2 [How would you characterize the water quality of the Red Cedar River and its
st

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
Poor 145 23.8 23.8 23.8
Fair 198 32.6 32.6 56.4
Good 90 14.8 14.8 71.2
Excellent 5 .8 .8 72.0
Don't Know 150 24.7 24.7 96.7
NO Response 20 3.3 3.3 100.0
Total 608 100.0 100.0
Q33 It is my personal responsibility to protect our rivers and streams for other pe0p1e
even if they seem unconcerned.
Frequency Percent Valid Percent Cumulative %t
Sirong‘y 8 1.3 1.3 1.3
Drsagree
Disagree 26 4.3 4.3 5.6
Neutral 104 17.1 17.1 22.7
Agree 316 52.0 52.0 74.7
Strongly Agree 136 22.4 22.4 97.0
NO Response 18 3.0 3.0 100.0
Total 608 100.0 100.0

 

 

 

Q3b- It is not my responsibility to ensure the well being of other species on earth.

 

 

 

 

 

 

 

 

 

 

 

 

recoded
Freguency Percent Valid Percent Cumulative %
Stimg'y 4 .7 .7 .7
Drsagree
Disagree 29 4.8 4.8 5.4
Neutral 52 8.6 8.6 14.0
Agree 242 39.8 39.8 53.8
Strongly Agree 259 42.6 42.6 96.4
No Response 22 3.6 3.6 100.0
Total 608 100.0 100.0

 

 

Q3c I I am partly responsible for the degraded state of our local rivers and stream.

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
Simg'y 97 16.0 16.0 16.0
Drsagree
Disagree 144 23.7 23.7 39.6
Neutral 169 27.8 27.8 67.4
Agree 156 25.7 25.7 93.1
Strongly Agree 19 3.1 3.1 96.2
NO Response 23 3.8 3.8 100.0
Total 608 100.0 100.0

 

I60

 

 

 

 

 

It is my responsibility to inform authorities about activities that violate our water

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q3d protection laws.
Frequency Percent Valid Percent Cumulative %
Simngly 9 1.5 1.5 1.5
Drsagree
Disagree 16 2.6 2.6 4.]
Neutral 67 11.0 11.0 15.1
Agree 333 54.8 54.8 69.9
Strongly Agree 168 27.6 27 .6 97.5
No Response 15 2.5 2.5 100.0
Total 608 100.0 100.0
How important to you is it that the Red Cedar River and its stream support
Q43 fi h' 2
s mg.
F reguency Percent Valid Percent Cumulative %
Not lrnportant 72 11.8 11.8 11.8
Somewhat Important 287 47.2 47 .2 59.0
Extremely Important 207 34.0 34.0 93.]
Don't Know 27 4.4 4.4 97.5
No Response 15 2.5 2.5 100.0
Total 608 100.0 100.0
Q 4b How important to you is it that the Red Cedar River and its stream support
wateringlawns and gardens?
Frequency Percent Valid Percent Cumulative %
Not Important 252 41.4 41.4 41.4
Somewhat
lmpo rtan t 222 36.5 36.5 78.0
Extremely
Important 49 8.1 8.] . 86.0
Don't Know 67 11.0 11.0 97.0
No Response 18 3.0 3.0 100.0
Total 608 100.0 100.0
Q40 How important to you is it that the Red Cedar River and its stream support
swimming?
Frequency Percent Valid Percent Cumulative %
Not Important 182 29.9 29.9 29.9
Somewhat Important 252 41.4 41.4 71.4
Extremely Important 1 13 18.6 18.6 90.0
Don't Know 44 7.2 7.2 97.2
No Response 17 2.8 2.8 100.0
Total 608 100.0 100.0

 

 

 

161

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q4 d How important to you is it that the Red Cedar River and its stream support
draining excess water?
Frequency Percent Valid Percent Cumulative %
Not Important 88 14.5 14.5 14.5
Somewhat
”@0113!“ 208 34.2 34.2 48.7
Extremely
Important 208 34.2 34.2 82.9
Don't Know 88 14.5 14.5 97.4
NO Response 16 2.6 2.6 100.0
Total 608 100.0 100.0 -
How important to you is it that the Red Cedar River and its stream be a water
Q4e .
supply for livestock or pets?
Frequency Percent Valid Percent Cumulative %
Not Important 124 20.4 20.4 20.4
Somewhat Important 269 44.2 44.2 64.6
Extremely Important 109 1 7.9 17.9 82.6
Don't Know 90 14.8 14.8 97.4
NO Response 16 2.6 2.6 100.0
Total 608 100.0 100.0
Q4 f How important to you is it that the Red Cedar River and its stream support
nature appreciation?
Frequency Percent Valid Percent Cumulative %
Not Important 1] 1.8 1.8 1.8
Somewhat Important 138 22.7 22.7 24.5
Extremely Important 429 70.6 70.6 95. 1
Don't Know 21 3.5 3.5 98.5
NO Response 9 1.5 1.5 100.0
Total 608 100.0 100.0
How important to you is it that the Red Cedar River and its stream support
Q4g recreational boating?
Frequency Percent Valid Percent Cumulative %
Not Important 88 14.5 14.5 14.5
Somewhat Important 294 48.4 48.4 62.8
Extremely Important 184 30.3 30.3 93. 1
Don't Know 29 4.8 4.8 97.9
NO Response 13 2.] 2.1 100.0
Total 608 100.0 100.0

 

 

 

 

 

 

162

 

 

 

 

How important to you is it that the Red Cedar River and its stream support

 

 

 

 

 

 

 

 

 

 

 

 

Q4h hunting?
Frequency Percent Valid Percent Cumulative %

Not Important 219 36.0 36.0 36.0
Somewhat Important 225 37.0 37.0 73.0
Extremely Important 79 13 .0 13.0 86.0
Don't Know 69 11.3 11.3 97.4

NO Response 16 2.6 2.6 100.0

Total 608 100.0 100.0

 

 

 

How important to you is it that the Red Cedar River and its stream support other

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q‘“ activities?
Frequency Percent Valid Percent Cumulative %
583 95.9 95.9 95.9
Clean environment for
Wildlife (3) 1 .2 .2 96.]
Disease control (3) l .2 .2 96.2
[Habitat for wildlife (2) 1 .2 .2 98.0
[Hiking L3) 1 .2 .2 98.2
Illiking, rec. use (2) I .2 .2 98.4
“nake attractive ptS of l .2 .2 98.5
ountain biking (3) l .2 .2 98.7
ature Trial (3) 1 .2 .2 98.8
Support wildlife (3) 1 .2 .2 99.7
Sustainable env. (3) 1 .2 .2 99.8
The view (3) 1 .2 .2 100.0
Total 608 100.0 100.0
2 = Somewhat Important, 3 = Extremely Important

 

 

 

 

My individual actions would improve water quality in local rivers and streams if

 

 

 

 

 

 

 

 

Q 5 a I were to attend a community meeting that involves concern over our local
streams and rivers.
Frequency Percent Valid Percent Cumulative %
Slmngly 23 3.8 3.8 3.8
Drsagree
Disagree 90 14.8 14.8 18.6
Neutral 238 39.1 39.1 57.7
Agree 223 36.7 36.7 94.4
Strongly Agree 22 3.6 3.6 98.0
NO Response 12 2.0 2.0 100.0
Total 608 100.0 100.0

 

 

 

 

 

 

163

 

 

 

 

My individual actions would improve water quality in local rivers and streams if

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

QSb I were to buy resource conservation devices such as.
Frequency Percent Valid Percent Cumulative %
Sirong'y 19 3.1 3.1 3.1
Drsagree
Disagree 74 12.2 12.2 15.3
Neutral 145 23.8 23.8 39.]
Agree 297 48.8 48.8 88.0
Strongly Agree 60 9.9 9.9 97 .9
No Response 13 2.1 2.1 100.0
Total 608 100.0 100.0
My individual actions would improve water quality in local rivers and streams if
Q5c I were report someone who violates a law or laws that protect our rivers and
streams to the prOper authorities.
Frequency Percent Valid Percent Cumulative %
Sirongly 5 .8 .8 .8
Drsagree
Disagree 8 1.3 1.3 2.]
Neutral 53 8.7 8.7 10.9
Agree 342 56.3 56.3 67.]
Strongly Agree 190 31.3 31.3 98.4
NO Response 10 1.6 1.6 100.0
Total 608 100.0 100.0
My individual actions would improve water quality in local rivers and streams if
Q5d l were convince someone to Sign a petition regarding an issue surrounding our
rivers and streams.
Frequency Percent Valid Percent Cumulative %
31mg” 14 2.3 2.3 2.3
Drsagree
Disagree 48 7 .9 7.9 10.2
Neutral 266 43.8 43.8 53.9
Agree 225 37.0 37.0 91.0
Strongly Agree 43 7.1 7.1 98.0
NO Response 12 2.0 2.0 100.0
Total 608 100.0 100.0

 

 

164

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

My individual actions would improve water quality in local rivers and streams if
Q5e I were convince someone tO buy household cleaning and/or laundry products that
don’t harm the environment.
Frequency Percent Valid Percent Cumulative %
80°“le 6 1.0 1.0 1.0
Drsagree
Disagree 27 4.4 4.4 5.4
Neutral 116 19.] 19.] 24.5
Agree 352 57.9 57.9 82.4
Strongly Agree 96 15.8 15.8 98.2
NO Response 11 1.8 1.8 100.0
Total 608 100.0 100.0
My individual actions would improve water quality in local rivers and streams if
Q5f I were convince someone to conserve water by not running the water while
brushing their teeth or shaving and/or installing a water saving device.
Frequency Percent Valid Percent Cllt)mulatlve
ercent
Sirong‘y 17 2.8 2.8 2.8
Drsagree
Disagree 47 7.7 7.7 10.5
Neutral 135 22.2 22.2 32.7
Agree 306 50.3 50.3 83.1
Strongly Agree 92 15.1 15.] 98.2
NO Response 1] 1.8 1.8 100.0
Total 608 100.0 100.0
Q6 1 Have ou heard about your community’s combined stormwater overflow plan?
Frequency Percent Valid Percent Cumulative %
no 226 37.2 37.2 37.2
yes 374 61.5 61.5 98.7
NO Response 8 1.3 1.3 100.0
Total 608 100.0 100.0
Q7 Have you heard about farmers in your community adopting generally accepted
management practices?
Frequency Percent Valid Percent Cumulative %
no 516 84.9 84.9 84.9
yes 78 12.8 12.8 97.7
NO Response 14 2.3 2.3 100.0
Total 608 100.0 100.0

 

 

 

 

 

 

 

165

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q8 IHave you heard about non- Oint sourceiollutants in the Red Cedar?
Frequency Percent Valid Percent Cumulative %
no 350 57.6 57.6 57.6
yes 248 40.8 40.8 98.4
NO Response 10 1.6 1.6 100.0
Total 608 1 00.0 100.0
Q9 [Are you aware of flooding oroblems in the Red Cedar or its streams?
Frequency Percent Valid Percent Cumulative %
no 14] 23.2 23.2 23.2
yes 461 75.8 75.8 99.0
NO Response 6 1.0 1.0 100.0
Total 608 100.0 100.0

 

 

 

 

 

 

Q10 I Are you aware of erosion prone areas of the Red Cedar River or its streams?

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
no 379 62.3 62.3 62.3
yes 221 36.3 36.3 98.7
NO Response 8 1.3 1.3 100.0
Total 608 100.0 100.0

 

 

 

 

 

 

Q11 I Have you seen a DIyBasin in your community?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
no 169 27.8 27.8 27.8
yes 175 28.8 28.8 56.6
Don't Know 257 42.3 42.3 98.8
NO Response 7 1.2 1.2 100.0
Total 608 100.0 100.0
Q12 I Have you seen a Wet Pond in your communi ?
Frequency Percent Valid Percent Cumulative 5
no 121 19.9 19.9 19.9
yes 338 55.6 55.6 75.5
Don't Know 142 23.4 23.4 98.8
No Response 7 1.2 1.2 100.0
Total 608 100.0 100.0

 

 

 

 

 

 

Q13 I Have you seen a Wetland in your community?

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
no 58 9.5 9.5 9.5
yes 495 81.4 81.4 91.0
Don't Know 51 8.4 8.4 99.3
NO Response 4 .7 .7 100.0
Total 608 100.0 100.0

 

 

 

 

 

166

 

 

 

Q14 IHave you seen a Filter Strip in your community?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
no 161 26.5 26.5 26.5
yes 193 31.7 31.7 58.2
Don't Know 25] 41.3 41.3 99.5
NO Response 3 .5 .5 100.0
Total 608 100.0 100.0
Q15 I Have you seen Rip Rap in your community?
Frequency Percent Valid Percent Cumulative %
no 207 34.0 34.0 34.0
yes 257 42.3 42.3 76.3
Don't Know 14] 23.2 23.2 99.5
NO Response 3 .5 .5 100.0
Total 608 100.0 100.0

 

 

 

 

 

 

Q16 I Have you seen Streambank Naturalization in our community?

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
no 125 20.6 20.6 20.6
gyes 262 43.1 43.] 63.7
Don't Know 217 35.7 35.7 99.3
NO Response 4 .7 .7 100.0
Total 608 100.0 100.0

 

 

 

 

 

 

C17 TWhich of these two plans would you prefer in your community?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
Plan A 279 45.9 45 .9 45.9
Plan B 282 46.4 46.4 92.3
NO Response 47 7 .7 7.7 100.0
Total 608 100.0 100.0
C18 I Which of these two plans woulclyou prefer in your community?
Frequency Percent Valid Percent Cumulative %
Plan A 231 38.0 38.0 38.0
Plan B 321 52.8 52.8 90.8
NO Response 56 9.2 9.2 100.0
Total 608 100.0 100.0

 

 

 

 

 

 

C19 I Which of these two plans would you prefer in your community?

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
Plan A 268 44.1 44.1 44.]
Plan B 285 46.9 46.9 91.0
No Response 55 9.0 9.0 100.0
Total 608 100.0 100.0

 

 

 

 

 

167

 

 

 

 

 

 

How supportive would you be of zoning requirements for some Open space to be

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q2021 preserved on undeveloped land?
Frequency Percent Valid Percent Cumulative %
WW“ N“ 21 3.5 3.5 3.5
Support
Might Support 98 16.1 16.1 19.6
Would Supmrt 193 31.7 31.7 51.3
Strongly Support 239 39.3 39.3 90.6
Do Not Know 36 5.9 5.9 96.5
NO Response 21 3.5 3.5 100.0
Total 608 100.0 100.0
How supportive would you be of subsidies to landowners for environmentally
Q20b fii .
endly practlces?
Frequency Percent Valid Percent Cumulative %
WW“ N‘” 53 8.7 8.7 8.7
Support
Might Simport 163 26.8 26.8 35.5
Would Support 216 35.5 35.5 71.]
Strongly Support 1 18 19.4 19.4 90.5
Do Not Know 39 6.4 6.4 96.9
NO Response 19 3.] 3.1 100.0
Total 608 100.0 100.0
How supportive would you be Of stricter regulations on activities that impact
QZOc .
waterways durrng development?
Frequency Percent Valid Percent Cumulative %
WW“ N“ 16 2.6 2.6 2.6
Support
Might Support 65 10.7 10.7 13.3
Would Support 238 39.] 39.1 52.5
Strongly Support 24] 39.6 39.6 92.]
DO Not Know 30 4.9 4.9 97.0
NO Response 18 3.0 3.0 100.0
Total 608 100.0 100.0

 

168

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q20g How sup aortive would you be of fines for polluting?
Frequency Percent Valid Percent Cumulative %
W°“'d N“ 10 1.6 1.6 1.6
Support
Might Support 42 6.9 6.9 8.6
Would Support 151 24.8 24.8 33.4
Strongly Support 376 61.8 61.8 95.2
DO Not Know 16 2.6 2.6 97.9
NO Response 13 2.1 2.] 100.0
Total 608 100.0 100.0
Q20e How supportive would you be of increased enforcement of environmental
regulations?
Frequency Percent Valid Percent Cumulative %
would N“ 21 3.5 3.5 3.5
Support
Might Support 60 9.9 9.9 13.3
Would Support 206 33.9 33.9 47.2
Strongly Support 286 47.0 47.0 94.2
DO Not Know 19 3.1 3.1 97.4
NO Response 16 2.6 2.6 100.0
Total 608 100.0 100.0
Q20f [How sup aortive would you be public information and education programs?
Frequency Percent Valid Percent Cumulative %
WW“ N‘” 12 2.0 2.0 2.0
Support
Might Support 75 12.3 12.3 14.3
Would Support 260 42.8 42.8 57 .1
Strongly Support 227 37.3 37.3 94.4
DO Not Know 17 2.8 2.8 97.2
NO Response 17 2.8 2.8 100.0
Total 608 100.0 100.0
Q20g How supportive would you be voluntary programs tO help landowners adopt
environmentally fiiendly practices?
Frequency Percent Valid Percent Cumulative %
WW“ 1"“ 9 1.5 1.5 1.5
Support
Might Support 72 11.8 11.8 13.3
Would Support 278 45.7 45.7 59.0
Strongly Support 210 34.5 34.5 93.6
DO Not Know 23 3.8 3.8 97.4
NO Response 16 2.6 2.6 100.0
Total 608 100.0 100.0

 

 

 

 

 

169

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q21 I Three preferred information sources:
Organization Frequency Ranking
MDA 168 4
NRCS 164 6
MSU 119 7
MSUE 166 5
MDEQ 206 3
Soil CD 75 10
Drain Commission 86 9
Farm Bureau 27 12
CO. Health Dept. 114 8
MEC 58 11
Newspapers 226 2
TV/Radio 232 1

 

 

 

Others, MUCC, Ml Pork Producers, Sierra Club,
EPA, PIRGIM, Individual Mailings

 

 

Questions 22 a-t; Please see summary Table 10 or Appendix E.

Question 23 & 24 see mean, standard deviation and variances above.

 

Q25 IDO you currently own or rent your home?

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
Own 587 96.5 96.5 96.5
NO Response 21 3.5 3.5 100.0
Total 608 100.0 100.0

 

 

Q26 I Which category includes your age?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %

18-24 yrs 6 1.0 1.0 1.0
25 - 34 yrs 79 13.0 13.0 14.0
35 - 44 yrs 119 19.6 19.6 33.6
45 - 54 yrs 155 25.5 25.5 59.0
55 - 64 yrs 107 17.6 17.6 76.6
65 + yrs 123 20.2 20.2 96.9
NO Regponse 19 3.1 3.1 100.0

Total 608 100.0 100.0

 

170

 

 

 

Q27 I Please indicate your highest level of education.

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
< High Scool
Graduate 20 3.3 3.3 3.3
Some School or

GED 10] 16.6 16.6 19.9

Some College 179 29.4 29.4 49.3
College Graduate 181 29.8 29.8 79.]
P0“ graduate 108 17.8 17.8 96.9
Degree

NO Response 19 3.1 3.] 100.0

Total 608 100.0 100.0

 

 

 

 

 

Q28 I Which catego includes your household income?

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
<$10,000 /yr 13 2.] 2.1 2.1
10,000 - 14,999 / yr 13 2.1 2.] 4.3
15,000 - 19,999 /yr 18 3.0 3.0 7.2
20,000 - 29,999 /yr 64 10.5 10.5 17.8
30,000 - 39,999 /yr 92 15.1 15.1 32.9
40,000 - 49,999 /yr 88 14.5 14.5 47.4
50,000 - 59,999 /yr 77 12.7 12.7 60.0
60,000 - 69,999 /yr 59 9.7 9.7 69.7
70,000 + /yr 112 18.4 18.4 88.2
NO Response 72 11.8 1 1.8 100.0
Total 608 100.0 100.0

 

 

 

 

 

 

Q29 I How many adults (1 8 years of age and over) currently live in your household?

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %

0 27 4.4 4.4 4.4

l 172 28.3 28.3 32.7
2 339 55.8 55.8 88.5
3 36 5.9 5.9 94.4
4 7 1.2 1.2 95.6
5 2 .3 .3 95.9
6 2 .3 .3 96.2

NO Regionse 23 3.8 3.8 100.0
Total 608 100.0 100.0

 

 

 

 

 

I71

 

 

 

 

Q30 I How many children under 18 currently live in your household?

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative %
0 431 70.9 70.9 70.9
1 72 11.8 11.8 82.7
2 58 9.5 9.5 92.3
3 20 3.3 3.3 95.6
4 3 .5 .5 96.1
5 1 .2 .2 96.2
No Response 23 3.8 3.8 100.0
Total 608 100.0 100.0

 

 

 

 

 

172

 

Appendix C:

Exploratory Factor Analysis Output

173

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

174

Extraction Method: Principal Component Analysis with Varimax Rotation

Descriptive Statistics Communalities
Mean Std. Analysis N Initial Extraction
Deviation Q22A l .000 .5 ] 7
Q22A 4.07 .655 538 Q22B 1.000 .465
QZZB 4.00 .666 538 Q22C 1.000 .488
Q22C 3.51 .965 538 Q22D 1.000 .431
Q22D 3.97 .812 538 Q22E 1.000 .385
Q22E 4.39 .675 538 Q22F 1.000 .54]
Q22F 3.99 .859 538 Q22GJ RC 1.000 .637
Q22G_RC 3.98 .863 538 Q27-H 1-000 -625
Q22H 3.05 .948 538 Q22I 1.000 .528
Q22] 3.83 .773 538 Q22J 1.000 .412
Q22] 4.35 .728 538 Q22K 1.000 .404
Q22K 1.93 .890 538 Q22L 1.000 .548
Q22L 3.51 .853 538 Q22M 1.000 .401
@M 4.04 .777 538 Q22N 1.000 .531
Q22N 3.85 .666 538 Q220 1.000 .534
Q220 3.19 .838 538 (222P 1-000 -387
Q22P 2.66 .992 5 38 Q22Q l .000 .442
Q22Q 2.31 .906 538 Q22K 1.000 .516
Q22R 3.99 .793 5 38 Q228 1.000 .693
Q22S 2.59 .945 5 38 Q22T 1.000 .367
Q22T 3.52 .952 538

 

Total Variance Explained

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. . Extraction Rotation Sums
Inltlal Sums of
Bi enval s uar d °f 3‘19”“
g ues q .e Loadlngs
Loadlngs
%Of Cumulative %Of Cumulative %Of Cumulative
Component Total Variance % Total Variance % Total Variance %
1 4.190 20.952 20.952 4.190 20.952 20.952 2.745 13.727 13.727
2 1.975 9.874 30.825 1.975 9.874 30.825 2.222 11.112 24.839
3 1.440 7.198 38.023 1.440 7.198 38.023 1.756 8.78] 33.621
4 1.217 6.083 44.106 1.217 6.083 44.106 1.697 8.483 42.104
5 1.030 5.150 49.256 1.030 5.150 49.256 1.430 7.152 49.256
6 .957 4.786 54.042
7 .907 4.535 58.577
8 .875 4.375 62.952
9 .863 4.317 67.268
10 .800 4.000 71.268
11 .718 3.592 74.860
12 .698 3.491 78.352
13 .659 3.296 81.647
14 .630 3.150 84.797
15 .580 2.899 87.696
16 .567 2.833 . 90.530
17 .536 2.678 93.207
18 .495 2.477 95.684
19 .479 2.394 98.078
20 .384 l .922 100.000

 

Extraction Method: Principal Component Analysis

175

 

Components

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Components
1 2 3 4 5
Q22N .638 .191 -5.152E-02 .286 -5.921E-02
Q22M .61] 4.078E-02 -9.367E-02 .130 1.779E-02
Q221 .591 .155 -6.776E-02 -.170 6.879E-02
Q22E .516 .105 -.173 -.252 .117
Q22R .501 .440 -.162 -.204 -5.982E-02
Q221 .499 .480 -3.l33E-02 -.149 -.161
Q22C 4.296E-02 .668 9.287E-02 -3.346E-02 .172
Q22r .163 .535 -.211 -8.148E-02 -5.460E-02
Q2213 .428 .523 -2.892E-02 5.419E-02 -5.974E-02
Q22F .104 .513 -.451 -.124 .220
Q22D .416 .450 -4.083E-02 2.538E-02 .231
szs -1.] l6E-02 —.128 .815 -2.260E—02 .110
Q22Q -.235 4.855E—03 .609 .120 30141302
Q22H .105 ‘1°%;5E' 4.034E-02 .78] -5.241E-02
Q22L .270 -.228 .341 .500 .239
Q22P -.202 -.147 -.100 .449 .336
Q22K -.295 '2'%6225' .339 .448 -2.983E-02
Q220 -.177 9.259E-02 .112 .115 .684
Q22A .263 .104 2.184E-02 -4.672E-02 .659
Q22G_RC .292 -.378 -.366 -.351 .390

 

Extraction Method: Principal Component Analysis
Rotation Method: Varimax with Kaiser Normalization
A Rotation converged in 9 iterations.

Normal analysis requested
Items have been assigned to factors as follows:

501 =4 3

21
502:6958
2]

503:101

31]

Average correlation within clusters:
.2683333 .252 .205

Standard score coefficient alphas:
.595 .627 .508

Three Factor Analyses

176

 

Item and factor correlation matrix

GJU'IWOI—‘NL’JD

\I

10
12
13
11
501
502
503

4
22
34
24
17
13
17
33
11
34
16
12
~6

~13
47
43

3 2 1
34 24 17
32 25 26
25 30 35
26 35 24
10 16 12
19 14 18
25 27 28
15 27 28
20 24 23

O ~4 -6

7 0 ~6
~6 ~12 ~15
~9 ~17 ~26
56 55 49
35 43 43
~4 ~18 ~29

6
13
10
16
12
22
19
31
20
26
-2

3

-12
-2
25
47
-7

9
17
19
14
18
l9
19
27
27
18
~4

~14
~11
~22
33
44
~28

5
33
25
27
28
31
27
34
26
28

5
~2

~16
~15
54
58
~15

8
ll
15
27
28
20
27
26
27
30

~12
~17
~9
~22
39
52
~33

******* Sampling error analysis *******
Sample size = 608

Analysis for cluster number 1

Internal consistency analysis
Within cluster correlation matrix for this cluster

I-‘Nleb

501

4
22
34
24
17

47

3
34
32
25
26

56

2
24
25
30
35

55

1
17
26
35
24

49

Significance test for deviation from a flat

within cluster correlation matrix. That is, a test

7
34
20
24
23
26
18
28
30
26
~l

3
~9

~17
49
51
~13

for the compound hypothesis (a) the items are

unidimensional (all measure the same construct) and

(b) the items are uniform in quality.

The value of chisquare is 25.677

The degrees of freedom are 5

The tail probability is .000

10
16
O
~4
~6
~2
~4
5
~12
—l
33
34
15
23
3
~6
58

Significance test for deviation from a unidimensional

within cluster correlation matrix that allows for variation

12
12
7

O
~6
3
~14
~2
~17
3
34
27
14
20
6
~11
52

in quality. That is, a test of the hypothesis the items all
measure the same construct which allows for a gradient
in item quality.

177

13
~6
~6
~12
~15
~12
~11
~16
~9
~9
15
14
9
17
~19
~23
30

11 501 502
~13 47 43
~9 56 35
~17 55 43
~26 49 43
~2 25 47
~22 33 44
~15 54 58
~22 39 52
~17 49 51
23 3 ~6
20 6 ~11
17 ~19 ~23
18 ~31 ~31
~31 100 79
~31 79 100
43 ~22 ~38

503

~4
~18
~29
~7
~28
~15
~33
~13
58
52
3O
43
~22
~38
100

The value of chisquare is 22.822
The degrees of freedom are 5
The tail probability is .000

Analysis of parallelism

Item by factor correlation matrix for this cluster

4 3 2 1
501 ***
502 43 35 43 43
503 5 ~4 ~18 ~29

Qualities ~-
501 47 56 55 49

Tests for parallelism

Test for flat tests for parallel AND uniform quality
Test with gradient allows for variation in quality

Test for flat Test with g
Cluster Chisq DF p Chisq
501 ***
502 2.317 3 0.509 7.390
503 28.235 3 0.000 26.938
TOTAL 37.164 6 0.000 41.758

Analysis for cluster number 2
lntemal consistency analysis

Within cluster correlation matrix for this cluster

6 9 5 8 7

6 22 19 31 20 26
9 19 19 27 27 18
5 31 27 34 26 28
8 20 27 26 27 30
7 26 18 28 30 26
502 47 44 58 52 51

Significance test for deviation fiom a flat
within cluster correlation matrix. That is, a test
for the compound hypothesis (a) the items are
unidimensional (all measure the same construct) and
(b) the items are uniform in quality.

178

radient

DF p
3 0.060
3 0.000
6 0.000

The value of chisquare is 20.536

The degrees of freedom are 9
The tail probability is .015

Significance test for deviation from a unidimensional
within cluster correlation matrix that allows for variation
in quality. That is, a test of the hypothesis the items all
measure the same construct which allows for a gradient
in item quality.

The value of chisquare is 12.122
The degrees of freedom are 9
The tail probability is .207

Analysis of parallelism
Item by factor correlation matrix for this cluster

6 9 5 8 7
501 25 33 54 39 49
502 ***
503 —7 ~28 ~15 ~33 ~13

Qualities ~—
502 47 44 58 52 51

Tests for parallelism
Test for flat tests for parallel AND uniform quality
Test with gradient allows for variation in quality

Test for flat Test with gradient
Cluster Chisq DF p Chisq DF p
501 27.784 4 0.000 15.093 4 0.005
502 ***
503 18.951 4 0.000 20.467 4 0.000
TOTAL 52.496 8 0.000 39.944 8 0.000

Analysis for cluster number 3

Internal consistency analysis
Within cluster correlation matrix for this cluster

10 12 13 11
10 33 34 15 23
12 34 27 14 20
13 15 14 9 17
11 23 20 17 18

503 58 52 30 43

179

Significance test for deviation from a flat
within cluster correlation matrix. That is, a test
for the compound hypothesis (a) the items are
unidimensional (all measure the same construct) and
(b) the items are uniform in quality.

The value of chisquare is 26.224
The degrees of freedom are 5
The tail probability is .000

Significance test for deviation from a unidimensional
within cluster correlation matrix that allows for variation
in quality. That is, a test of the hypothesis the items all
measure the same construct which allows for a gradient
in item quality.

The value of chisquare is 5.084
The degrees of freedom are 5
The tail probability is .406

Analysis of parallelism
Item by factor correlation matrix for this cluster

10 12 13 11
501 3 6 ~19 ~31
502 ~6 ~11 ~23 ~31
503 ***

Qualities ~-
503 58 52 30 43

Tests for parallelism

Test for flat tests for parallel AND uniform quality
Test with gradient allows for variation in quality

Test for flat Test with gradient
Cluster Chisq DF p Chisq DF p
501 43.745 3 0.000 56.166 3 0.000
502 19.076 3 0.000 35.238 3 0.000
503 ***
TOTAL 32.338 6 0.000 47.051 6 0.000

180

Two factor Analysis
Normal analysis requested
Items have been assigned to factors as follows:

501:432169587

502:10121311

Average correlation within clusters:
.2297222 .205

Standard score coefficient alphas:

.729 .508

Item and factor correlation matrix

oomxooxr—‘ch.

\J

10
12
13
11
501
502

4
23
34
24
17
13
17
33
11
34
16
12
~6

~13
48
5

3 2 1
34 24 17
20 25 26
25 25 35
26 35 24
10 16 12
19 14 18
25 27 28
15 27 28
20 24 23

0 ~4 ~6

7 0 ~6
~6 ~12 ~15
~9 ~17 ~26
45 50 49
~4 ~18 ~29

6
13
10
16
12
14
19
31
20
26
~2

3

~12
~2
37
~7

9
17
19
14
18
19
17
27
27
18
~4
~14
~11
~22

41
~28

5
33
25
27
28
31
27
37
26
28
5
~2
~16
~15

61
~15

8
11
15
27
28
20
27
26
23
30

~12
~17
~9
-22 -
48
-33 _

******* Sampling error analysis *******
Sample size = 608

Analysis for cluster number 1

Internal consistency analysis
Within cluster correlation matrix for this cluster

\ImU'IWOTP-‘NUJb

501

4
23
34
24
17
13
17
33
11
34

48

3
34
20
25
26
10
19
25
15
20

45

2
24
25
25
35
16
14
27
27
24

50

1
17
26
35
24
12
18
28
28
23

49

6
13
10
16
12
14
19
31
20
26

37

9
17
19
14
18
19
17
27
27
18

41

5
33
25
27
28
31
27
37
26
28

61

8
11
15
27
28
20
27
26
23
30

48

181

7
34
20
24
23
26
18
28
30
29
~1

3
~9
17
54
13

7
34
20
24
23
26
18
28
30
29

54

10
16

~4
~6
~2
~4

~12
~l
33
34
15
23
~2
58

12
12

~6

~14
~2
~17

34
27
14
20
~3
52

13
~6
~6
~12
~15
~12
~11
~16
~9
~9
15
14

17
~22
30

11
~13
~9
~17
~26
~2
~22
~15
~22
~17
23
20
17
18
~33
43

501
48
45
50
49
37
41
61
48
54
~2
~3

~22

~33

100

-33

502

~4
~18
~29
~7
~28
~15
~33
~13
58
52
30
43
-33
100

Significance test for deviation from a flat
within cluster correlation matrix. That is, a test
for the compound hypothesis (a) the items are
unidimensional (all measure the same construct) and
(b) the items are uniform in quality.

The value of chisquare is 171.075
The degrees of freedom are 35
The tail probability is .000

Significance test for deviation from a unidimensional
within cluster correlation matrix that allows for variation
in quality. That is, a test of the hypothesis the items all
measure the same construct which allows for a gradient
in item quality.

The value of chisquare is 113.144
The degrees of freedom are 35
The tail probability is .000

Analysis of parallelism

Item by factor correlation matrix for this cluster

4 3 2 1 6 9 5 8 7
501 ***
502 5 ~4 ~18 ~29 ~7 ~28 ~15 ~33 ~13

Qualities ~—
501 48 45 50 49 37 41 61 48 54

Tests for parallelism
Test for flat tests for parallel AND uniform quality
Test with gradient allows for variation in quality

Test for flat Test with gradient
Cluster Chisq DF p Chisq DF p
501 ***
502 50.407 8 0.000 50.359 8 0.000
TOTAL 50.407 8 0.000 50.359 8 0.000

Analysis for cluster number 2
lntemal consistency analysis

182

Within cluster correlation matrix for this cluster

10 12 13 11
10 33 34 15 23
12 34 27 14 20
13 15 14 9 17
11 23 20 17 18

502 58 52 30 43
Significance test for deviation from a flat
within cluster correlation matrix. That is, a test
for the compound hypothesis (a) the items are
unidimensional (all measure the same construct) and
(b) the items are uniform in quality.

The value of chisquare is 26.224
The degrees of freedom are 5
The tail probability is .000

Significance test for deviation from a unidimensional
within cluster correlation matrix that allows for variation
in quality. That is, a test of the hypothesis the items all
measure the same construct which allows for a gradient
in item quality.

The value of chisquare is 5.084
The degrees of freedom are 5
The tail probability is .406

Analysis of parallelism
Item by factor correlation matrix for this cluster

10 12 13 11
501 ~2 ~3 ~22 ~33
502 ***

Qualities -~
502 58 52 30 43

Tests for parallelism

Test for flat tests for parallel AND uniform quality
Test with gradient allows for variation in quality

Test for flat Test with gradient
Cluster Chisq DF p Chisq DF p
501 38.462 3 0.000 58.311 3 0.000
502 ***
TOTAL 38.462 3 0.000 58.311 3 0.000

183

Appendix D:

Cluster Analysis Output

184

Initial Cluster Centers

Cluster
1 2
[FACTOR] 2.78 4.00
FACTOR3 5.00 .33

 

 

 

 

 

 

 

 

 

Iteration History
Change in Cluster Centers
Iteration 1 2

1 1.972 1.934

2 2.067E-02 2.437E-02

3 3.495E-03 4.106E-03

4 .000 .000
a Convergence achieved due to no or small distance change.
The maximum distance by which any center has changed is .000. The current iteration is
4. The minimum distance between initial centers is 4.824.

 

 

 

 

 

 

 

 

 

 

 

Final Cluster Centers

Cluster
1 2
[FACTOR] 3.82 4.02
[FACTOR3 3.30 2.25

 

 

 

 

 

 

 

 

 

Distances between Final Cluster Centers

 

 

 

 

 

 

 

 

 

 

 

 

Cluster 1 2

l 1 .075

2 1 .075

ANOVA
Cluster Error F Sig.
Mean df Mean df
Square Square

FACTORI 6.290 1 .241 597 26.060 .000
FACTOR3 165.871 1 .169 597 979.479 .000

 

 

 

 

 

 

 

 

 

The F tests should be used only for descriptive purposes because the clusters have been
chosen to maximize the differences among cases in different clusters. The observed
significance levels are not corrected for this and thus cannot be interpreted as tests of the
hypothesis that the cluster means are equal.

185

Number of Cases in each Cluster

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Cluster 1 320.000
2 279.000
Valid 599.000
Missing 9.000
Group Statistics
Cluster N Mean Std. Std. Error
Members Deviation Mean
hip
FACTOR] 1 320 3.8153 .5221] .02919
2 279 4.0207 .45336 .02714
FACTOR3 1 320 3.3010 .40889 .02286
2 279 2.2461 .4145] .02482
Independent Samples Test
Levene's
Test for
Equality t-test for
of Equality of
Variances Means
F Sig. t df Sig. Mean Std. Error 95%
(2- iDifl‘erencelDifl‘erence Confidence
tailed) Interval of the
Difference
Lower Upper
IFACTORllEqual 8.859.003 -5.105 597 .000 -.2054 .04024 -.2845 -.12640
variancesl
lassumed
ual -5.154596.991.000 -.2054 .03986 -.2837-.12715
ariancesw
0t
sumed
1FACTOR3 qua] 2.067.]5131297 597 .000 1.0549 .0337] .9887 1.12112
Eariancesr
sumed
ual 31267583678 .000 1.0549 .03374.988661.12119
ariancesr
0t
sumed

 

 

 

 

 

 

 

 

 

 

 

 

186

Appendix E:

Fairness Heuristic Analysis Output

187

Statistics

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

N Mean Std. Deviation
Valid Missing

Q22A 597 l 1 4.08 .655
Q22E 600 8 3.99 .670
Q22C 593 15 3.50 .976
Q22D 598 10 3 .96 .805
Q22E 600 8 4.39 .680
Q22F 593 15 3.98 .869
Q226_RC 596 12 3.96 .874
Q22H 586 22 3.05 .943
Q221 592 16 3.84 .766
Q22] 596 12 4.35 .717
Q22K 597 1] 1.94 .910
Q22L 595 13 3.53 .851
Q22M 588 20 4.05 .77]
Q22N 586 22 3.85 .661
Q220 585 23 3.19 .844
Q22P 584 24 2.67 .985
Q22Q 578 30 2.3] .892
Q22R 590 18 3.97 .825
Q2ZS 588 20 2.59 .957
Q22T 586 22 3.53 .952

 

188

 

Frequencies

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Strongly Strongly Total Mean Std.
Disagree Disagree Neutral Agree Agree Dev.
(1) (2) (3) (4) (5)
Q22a Frequency 1 14 59 388 135 597 4 .08 0.655
Valid Percent 0.2 2.3 9.9 65 22.6 100
Q22h Frequency 6 12 65 4] 5 102 600 3 .99 0.67
Valid Percent l 2 10.8 69.2 17 100
Q22e Frequency 1 7 81 I60 259 76 593 3.5 0.976
Valid Percent 2.9 13.7 27 43.7 12.8 100
Q22d Frequency 6 32 73 354 1 33 598 3.96 0.805
Valid Percent 1 5.4 12.2 59.2 22.2 100
Q22e Frequency 0 8 43 255 294 600 4.39 0.68
Valid Percent 0 1.3 7.2 42.5 49 100
Q22f Frequency 7 30 97 292 167 593 3 .98 0.869
Valid Percent 1.2 5.1 16.4 49.2 28.2 100
QZZg—RC Frequency 12 28 82 32] 153 596 3.96 0.874
Valid Percent 2 4.7 13.8 53.9 25.7 100
Q22h Frequency 30 130 229 l 72 25 586 3.05 0.943
Valid Percent 5.1 22.2 39.1 29.4 4.3 100
Q22i Frequency 4 26 l 25 341 96 592 3 .84 0.766
Valid Percent 0.7 4.4 21.1 57.6 16.2 100
Q22j Frequeng 8 5 22 296 265 596 4.35 0.717
Valid Percent 1.3 0.8 3.7 49.7 44.5 100
922k Frequency 201 280 75 30 l 1 597 1.94 0.91
Valid Percent 33.7 46.9 12.6 5 1.8 100
Q22I Frequency 8 67 174 296 50 595 3.53 0.85]
Valid Percent 1.3 11.3 29.2 49.7 8.4 100
QZZm Frequency 7 2] 56 357 147 588 4.05 0.77]
Valid Percent 1.2 3.6 9.5 60.7 25 100
Q22n Frequency 3 24 87 414 58 586 3.85 0.66]
Valid Percent 0.5 4.1 14.8 70.6 9.9 100
Q220 Frequency 16 91 264 191 23 585 3.19 0.844
Valid Percent 2.7 15.6 45.1 32.6 3.9 100
Q22p Frequency 59 21 1 196 97 21 584 2 .67 0.985
Valid Percent 10.] 36.1 33.6 16.6 3.6 100
Q221 Frequency 85 306 120 59 8 578 2.3] 0.893
Valid Percent 14.7 52.9 20.8 10.2 1.4 100
Q22r Frequency 13 15 86 338 138 590 3.97 0.825
Valid Percent 2.2 2.5 14.6 57.3 23.4 100
0223 Frequency 61 240 178 95 I4 588 2 .59 0.957
Valid Percent 10.4 40.8 30.3 16.2 2.4 100
Q22t Frequency 14 7] I69 253 79 586 3.53 0.952
Valid Percent 2.4 12.1 28.8 43.2 13.5 100

 

189

 

Agree and Strongly Agree Combined Responses for Fairness Heuristic

 

Levene's Test for

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Equality of t-test for Equality of Means
Variances

. Si .

F Sig. t (11' (2-taigl ed)
Q22AREDU Equal variances assumed 13.394 .000 -1.828 594 .068
Equal variances not assumed -l.860 587.395 .063
Q22BREDU Equal variances assumed 44.159 .000 -3.372 596 .00]
Equal variances not assumed -3.450 575.430 .00]
Q22CREDU Equal variances assumed .061 .805 -1.318 591 .188
Equal variances not assumed -I .318 579.206 .188
Q22DREDU Equal variances assumed 10.858 .001 -1.813 595 .070
Equal variances not assumed -1.829 594.931 .068
Q22EREDU Equal variances assumed 25.398 .000 -2.535 597 .01]
Equal variances not assumed -2.587 581.198 .010
Q22PREDU Equal variances assumed 72.61 I .000 -4.751 591 .000
Equal variances not assumed -4.847 576.795 .000
QZZGREDU Equal variances assumed 83.254 .000 -4.638 594 .000
Egal variances not assumed -4.774 558.056 .000
Q22HREDU Equal variances assumed 2.290 .131 21.486 584 .000
Era] variances not assumed 21.360 551.218 .000
Q221REDU Equal variances assumed 55.509 .000 -4.560 590 .000
Equal variances not assumed -4.628 588.347 .000
QZZJREDU Equal variances assumed .247 .619 -.303 593 .762
Egral variances not assumed -.301 564.008 .764
Q22KREDU Equal variances assumed 700.414 .000 10.519 595 .000
Equal variances not assumed I 1.252 344.719 .000
Q22LREDU Equal variances assumed 68.581 .000 15.813 593 .000
Eflal variances not assumed 15.207 418.055 .000
Q22MREDU Equal variances assumed 2.317 .128 -.903 583 .367
Equal variances not assumed -.904 576.633 .366
QQZNREDU Equal variances assumed 3.057 .081 -.793 581 .428
Equal variances not assumed -.799 580.997 .424
Q220REDU Equal variances assumed 16.214 .000 2.369 580 .018
Equal variances not assumed 2.384 578.922 .017
Q22PREDU Equal variances assumed .365 .546 3.887 579 .000
Equal variances not assumed 3.904 577.871 .000
QZQREDU Equal variances assumed 55.698 .000 5.241 573 .000
Equal variances not assumed 5.316 565.610 .000
Q22RREDU Equal variances assumed 63.481 .000 -4.300 583 .000
Equal variances not assumed -4.374 566.935 .000
Q22SREDU Equal variances assumed 7.370 .007 3.954 583 .000
Equal variances not assumed 3.984 582.940 .000
Q22TREDU Equal variances assumed 15.049 .000 -3.604 581 .000
Equal variances not assumed -3.636 580.465 .000

 

 

190

— ~.o H QO§E> .WV— .9 H HOHONM huemo S #0: nab: MO :50:
WOZRQAU OH. m—UZ/th—mmd H mac: BN8?“
mgukémms: 302 H mp5: “gov—om
_ sown—r :VOM. 2b.- Inna—r 86.- cows. snmafi enema». 9.9—N.- .I6NV~. toga. wmQr m8: nevhN. onhhq. .551 saamfi oaNhN. «N8. «N60
— 21.2.. usmmm. one. uno:. owner #8... :mmm. :OON. .uon:.r comic mmo. «soc—r suhwmr a-Nzr «cam: 68.-. v8.» hmo. mNNo
— .va—r now—_r mvor Ivnwn. DISN. :or IQNONI 30m... rochv. _wor nomO. 0J~ON. OOVNW. IQMOm. :No—. .TJOQM. nos. uNNO
— sacs. one. N59. cor—NF 6:2. :OVN. .13—r :ooN; :mfl. son—N; :3”- rim—... ooh—u: 03.- .652: 35.. UNNO
« .... ..8..__..§..u a... £2. ......Ea , . as, was ...:q-.. ..VE.........s.s. . as .. .31.... was a... , as.
_ mm... .. 86.- :5. MB. .87 .28. :o. «8. 8a.- 3a.. .62. 68.- :62. 38
— :mvm. :VN — . :cN _ r :ONM. a. v2. n8.. lama: . Inv:. :- wvm. :wN— . noonm. cnnz. ENNO
n nha. omgr Ion—u. onwV—. . QIOWN. CIOON. 008. comma. nomuu. INNO
, . . _ . , as. ......33; ,_ . 8. ..aa . . .....s as. _ as ..
. . .61.- . . a,

  

    
  

—. ..63; 3.3..» y
_ .32. ..E. [.83. .62. 38
o—or GOOBN. Gabon. §._m_. _NNO
.....E... p.33 .... «.35.........aq. ...H 3.8 ._
. _ 8:. . «as. . :32. 91.38
_ .33. .62. :8

coth. lummfi DNNO
«icwm. GIFVN. :8
— semen. «m8. 0N8

— o I — o— . has
38

 

_
.8 38 38 £8 .38 6N8 .38 330 R8 38 :8 a8 .38 8.1338 :8 £8 £8 38 .38 38
9:3— 0.323: 32:25 .5952. 25322.50

191

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(122A Frequency Percent Valid Percent Cumulative %
Valid Sirong‘y l .2 .2 .2
Drsagree
Disagree 14 2.3 2.3 2.5
Neutral 59 9.7 9.9 12.4
Agree 388 63.8 65.0 77.4
Strongly Agree 135 22.2 22.6 100.0
Total 597 98.2 100.0
Missing System 1 l 1.8
Total 608 100.0
QZZB Frequency Percent Valid Percent Cumulative %
Valid 330”” 6 1.0 1.0 1.0
Drsagree
Disagree 12 2.0 2.0 3.0
Neutral 65 10.7 10.8 13.8
Agree 415 68.3 69.2 83.0
Strongly Agree 102 16.8 17.0 100.0
Total 600 98.7 100.0
Missing System 8 1.3
Total 608 100.0
22C Frequency Percent Valid Percent Cumulative %
Valid 330”” 17 2.8 2.9 2.9
Drsagree
Disagree 8] 13.3 13.7 16.5
Neutral 160 26.3 27.0 43.5
Agree 259 42.6 43.7 87.2
SW‘gly 76 12.5 12.8 100.0
Agree
Total 593 97.5 100.0
Missing System 15 2.5
Total 608 100.0
22D Frequency Percent Valid Percent Cumulative %
Valid Sirong'y 6 1.0 1.0 1.0
Drsagree
Disagree 32 5.3 5.4 6.4
Neutral 73 12.0 12.2 18.6
Agree 354 58.2 59.2 77.8
Strongly Agree 133 21.9 22.2 100.0
Total 598 98.4 100.0
Missing System 10 1.6
Total 608 100.0

 

 

 

 

 

 

 

 

192

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

IQ22E Frequeng Percent Valid Percent Cumulative %
Valid Disagree 8 1 .3 1.3 1 .3
Neutral 43 7.] 7.2 8.5
Agree 255 41.9 42.5 51.0
Strongly Agree 294 48.4 49.0 100.0
Total 600 98.7 100.0
Missing System 8 l .3
Total 608 100.0
Q22F Frequency Percent Valid Percent Cumulative %
Valid Simng‘y 7 1.2 1.2 1.2
Drsagree
Disagree 30 4.9 5.1 6.2
Neutral 97 16.0 16.4 22.6
Agree 292 48.0 49.2 71.8
Strongly Agree 167 27.5 28.2 100.0
Total 593 97.5 100.0
Missing System 15 2.5
Total 608 100.0
Q226_RC Frequency Percent Valid Percent ICumulative %
Valid Sirong'y 12 2.0 2.0 2.0
Drsagree
Disagree 28 4.6 4.7 6.7
Neutral 82 13.5 13.8 20.5
Agree 32] 52.8 53.9 74.3
Strongly Agree 153 25.2 25.7 100.0
Total 596 98.0 100.0
Missing System 12 2.0
Total 608 100.0
22H Frequency Percent Valid Percent Cumulative %
Valid Simngly 30 4.9 5.1 5.1
Drsagree
Disagree 130 21.4 22.2 27.3
Neutral 229 37.7 39.] 66.4
Agree 172 28.3 29.4 95.7
Strongly Agree 25 4.1 4.3 100.0
Total 586 96.4 100.0
Missing System 22 3.6
Total 608 100.0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q22] Frequency Percent Valid Percent Cumulative %
Valid Sirong'y 4 .7 .7 .7
Drsagree
Disagree 26 4. 3 4.4 5.]
Neutral 125 20.6 21.] 26.2
Agree 34] 56.1 57.6 83.8
Strongly Agree 96 15.8 16.2 100.0
Total 592 97.4 100.0
Missing System 16 2.6
Total 608 100.0
Q22J Frequency Percent Valid Percent Cumulative %
Valid Sirong‘y 8 1.3 1.3 1.3
Drsagree
Disagree 5 .8 .8 2.2
Neutral 22 3.6 3.7 5.9
Agree 296 48.7 49.7 55.5
Strongly Agree 265 43.6 44.5 100.0
Total 596 98.0 100.0
Missing System 12 2.0
Total 608 100.0
22K Frequency Percent Valid Percent Cumulative %
Valid Si’°“g'y 201 33.1 33.7 33.7
Drsagree
Disagree 280 46.1 46.9 80.6
Neutral 75 12.3 12.6 93.]
Agree 30 4.9 5.0 98.2
Strongly Agree 11 1.8 1.8 100.0
Total 597 98.2 100.0
Missing System 1 1 1.8
Total 608 100.0
[622L FrecRency Percent Valid Percent Cumulative %
Valid girong‘y 8 1.3 1.3 1.3
lsagree
Disagree 67 11.0 11.3 12.6
Neutral 174 28.6 29.2 41.8
Agree 296 48.7 49.7 91.6
Strongly Agree 50 8.2 8.4 100.0
Total 595 97.9 100.0
Missing System 13 2.1
Total 608 100.0

 

 

 

 

 

 

 

 

194

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q22M Frequency Percent Valid Percent Cumulative %
Valid Sirong'y 7 1.2 1.2 1.2
Drsagree
Disagree 2] 3.5 3.6 4.8
Neutral 56 9.2 9.5 14.3
Agree 357 58.7 60.7 75.0
Strongly Agree 147 24.2 25.0 100.0
Total 588 96.7 100.0
Missing System 20 3.3
Total 608 100.0
Q22N Frequency Percent Valid Percent Cumulative %
Valid 330“!” 3 .5 .5 .5
Drsagee
Diszquee 24 3.9 4.] 4.6
Neutral 87 14.3 14.8 19.5
Agree 414 68.1 70.6 90.]
Strongly Agee 58 9.5 9.9 100.0
Total 586 96.4 100.0
Missing System 22 3.6
Total 608 100.0
[Q220 Frequency Percent Valid Percent Cumulative %
Valid 53°“g'y 16 2.6 2.7 2.7
Drsagree
Disagree 9] 15.0 15.6 18.3
Neutral 264 43.4 45.1 63.4
Agree 19] 31.4 32.6 96.1
Strongly Agree 23 3.8 3.9 100.0
Total 585 96.2 100.0
Missing System 23 3.8
Total 608 100.0
R222P Frequency Percent Valid Percent Cumulative %
Valid 580%” 59 9.7 10.1 10.1
Drsagree
Disagree 2] 1 34.7 36.1 46.2
Neutral 196 32.2 33.6 79.8
Agree 97 16.0 16.6 96.4
Strongly Agree 21 3.5 3.6 100.0
Total 584 96.1 100.0
Missing System 24 3.9
Total 608 100.0

 

 

 

195

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

|Q22Q Frequency Percent Valid Percent lCumulative %
Valid Sirongly 85 14.0 14.7 14.7
Drsagree
Disagree 306 50.3 52.9 67.6
Neutral 120 19.7 20.8 88.4
Agree 59 9.7 10.2 98.6
Strongly Agree 8 1.3 1.4 100.0
Total 578 95.1 100.0
Missing System 30 4.9
Total 608 100.0
[Q22R Frequency Percent Valid Percent Cumulative %
Valid 5‘30“” 13 2.1 2.2 2.2
Drsagree
Disagree 15 2.5 2.5 4.7
Neutral 86 14.] 14.6 19.3
Agree 338 55.6 57.3 76.6
Strongly Agree 138 22.7 23.4 100.0
Total 590 97.0 100.0
Missing System 18 3 .0
Total 608 100.0
|Q228 Frequency Percent Valid Percent Cumulative %
Valid 35mg" 61 10.0 10.4 10.4
Drsagree
Disagree 240 39.5 40.8 5 1.2
Neutral 178 29.3 30.3 81.5
Agree 95 15.6 16.2 97.6
Strongly Agree 14 2.3 2.4 100.0
Total 588 96.7 100.0
Missigg System 20 3.3
Total 608 100.0
Q22T Frequency Percent Valid Percent Cumulative %
Valid Simng‘y 14 2.3 2.4 2.4
Drsagree
Disagree 7] 11.7 12.] 14.5
Neutral 169 27.8 28.8 43.3
Agree 253 41.6 43.2 86.5
Strongly Agree 79 13.0 13.5 100.0
Total 586 96.4 100.0
Missing System 22 3.6
Total 608 100.0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q22AREDU Total
2.00 3.00 4.00
cm“ . 1 Count 11 35 273 319
embershrp
% Within
Cluster 3.4% 1 1.0% 85.6% 100.0%
'Membership
2 Count 4 24 249 277
% Within
Cluster 1.4% 8.7% 89.9% 100.0%
LMemberslmr
Total Count 15 59 y 522 596
% Within
Cluster 2.5% 9.9% 87.6% 100.0%
[Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided)
Pearson Chi-Sgare 3.479 2 .176
Likelihood Ratio 3.602 2 .165
L‘“ear'b¥'l:‘“ea' 3.329 1 .068
Assocratron
N of Valid Cases 596
0 cells (.0%) have expected count less than 5. The minimum expected count is 6.97.
Q22BREDU Total
2.00 3.00 4.00
1114311133111 1 Count 12 47 261 320
P
% Within
Cluster 3.8% 14.7% 81.6% 100.0%
Membership
2 Count 5 18 255 278
% Within
Cluster 1.8% 6.5% 91.7% 100.0%
Membershi
Total Count 1 7 65 5 16 598
% Within
Cluster 2.8% 10.9% 86.3% 100.0%
Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided
Pearson Chi-Square 13.005 2 .001
Likelihood Ratio 13.493 2 .001
L‘“ea"b.y“:‘“ea’ 1 1.179 1 .001
Assocratlon
N of Valid Cases 598

 

 

 

 

 

 

0 cells (.0%) have expected count less than 5. The minimum expected count is 7.90.

197

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q22CREDU Total
2.00 3.00 4.00
cm” . 1 Count 54 95 169 318
embershlp
% Within
Cluster 17.0% 29.9% 53.1% 100.0%
[Membership
2 Count 44 65 166 275
% Within
Cluster 16.0% 23.6% 60.4% 100.0%
embership
Total Count 98 160 335 593
% Within
Cluster 16.5% 27.0% 56.5% 100.0%
[Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided)1
Pearson Chi-Square 3.573 2 .168
Likelihood Ratio 3.587 2 .166
Ll’far'bY'Ii‘nea’ 1.735 1 .188
ssoclatlon
N of Valid Cases 593

 

 

 

 

 

 

0 cells (.0%) have expected count less than 5. The minimum expected count is 45.45.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q22DREDU Total
2.00 3.00 4.00
Meflgigip 1 Count 23 46 250 319
% Within
Cluster 7.2% 14.4% 78.4% 100.0%
LMembership
2 Count 15 27 236 278
% Within
Cluster 5.4% 9.7% 84.9% 100.0%
Membership
Total Count 38 73 486 597
% Within
Cluster 6.4% 12.2% 81.4% 100.0%
[Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 4.237 2 .120
Likelihood Ratio 4.285 2 .117
Li’far'bY'Irinea’ 3.276 1 .070
ssocratlon
N of Valid Cases 597

 

 

 

 

 

 

 

0 cells (.0%) have expected count less than 5. The minimum expected count is 17.70.

198

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q22EREDU Total
2.00 3.00 4.00
M62321“) 1 Count 5 32 283 320
% Within
Cluster 1.6% 10.0% 88.4% 100.0%
'Membership
2 Count 3 1 1 265 279
% Within
Cluster 1.1% 3.9% 95.0% 100.0%
'Membership
Total Count 8 43 548 599
% Within
Cluster 1.3% 7.2% 91.5% 100.0%
[Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided)
Pearson Chi-Sgare 8.581 2 .014
Likelihood Ratio 8.996 2 .01]
Linear'bY'Itinea’ 6.370 1 .012
Assocratlon
N of Valid Cases 599
0 cells (.0%) have expected count less than 5. The minimum expected count is 3.73.
QZZFREDU Total
2.00 3.00 4.00
mum . 1 Count 26 72 219 317
embershlp
% Within
Cluster 8.2% 22.7% 69.1% 100.0%
'Membership
2 Count 11 25 240 276
% Within
Cluster 4.0% 9.1% 87.0% 100.0%
'Membership
Total Count 37 97 459 593
% Within
Cluster 6.2% 16.4% 77.4% 100.0%
[Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 27.110 2 .000
Likelihood Ratio 28.144 2 .000
Law'bY'lfma’ 21.776 1 .000
ssocratlon
N of Valid Cases 593

 

 

 

 

 

 

0 cells (.0%) have expected count less than 5. The minimum expected count is 17.22.

199

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

QZZGREDU Total
2.00 3.00 4.00
Meggjghip 1 Count 32 55 233 320
% Within
Cluster 10.0% 17.2% 72.8% 100.0%
[Membership
2 Count 27 241 276
% Within
Cluster 2.9% 9.8% 87.3% 100.0%
Membership
Total Count 40 82 474 596
% Within
Cluster 6.7% 13.8% 79.5% 100.0%
[Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 20.962 2 .000
Likelihood Ratio 22.059 2 .000
me'bY'Ir‘m’ 20.795 1 .000
Assoclatlon
N of Valid Cases 596
0 cells (.0%) have expected count less than 5. The minimum expected count is 18.52.
Q22HREDU Total
2.00 3.00 4.00
11463232815 1 Count 1 1 126 182 319
% Within
Cluster 3.4% 39.5% 57.1% 100.0%
'Membership
2 Count 149 103 15 267
% Within
Cluster 55.8% 38.6% 5.6% 100.0%
'Membership
Total Count 160 229 197 586
% Within
Cluster 27.3% 39.1% 33.6% 100.0%
[Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided
Pearson Chi-Square 260.339 2 .000
Likelihood Ratio 306.392 2 .000
L‘Ikear'b¥"7‘“ea' 258.270 1 .000
ssoclatlon
N of Valid Cases 586
0 cells (.0%) have expected count less than 5. The minimum expected count is 72.90.

200

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q221REDU Total
2.00 3.00 4.00
Chm . 1 Count 19 92 207 318
embershrp
% Within
Cluster 6.0% 28.9% 65.1% 100.0%
LMembership
2 Count 1 1 33 230 274
% Within
Cluster 4.0% 12.0% 83.9% 100.0%
[Membership
Total Count 30 125 437 592
% Within
Cluster 5.1% 21.1% 73.8% 100.0%
[Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided)
Pearson Chi-Scplare 28.077 2 .000
Likelihood Ratio 29.084 2 .000
“‘far'bY'ITinear 20.1 19 1 .000
ssocratron
N of Valid Cases 592

 

 

 

 

 

0 cells (.0%) have expected count less than 5. The minimum expected count is 13.89.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q221REDU Total
2.00 3.00 4.00
Clum . 1 Count 5 17 297 319
embershlp
% Within
Cluster 1.6% 5.3% 93.1% 100.0%
embership
2 Count 8 5 263 276
% Within
Cluster 2.9% 1.8% 95.3% 100.0%
'Membership
Total Count 1 3 22 560 595
% Within
Cluster 2.2% 3.7% 94.1% 100.0%
[Membership
Chi—Square Tests Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 6.227 2 .044
Likelihood Ratio 6.570 2 .037
Llaear'b¥"7mea’ .092 1 .762
ssocratron
N of Valid Cases 595

 

 

 

0 cells (.0%) have expected count less than 5. The minimum expected count is 6.03.

201

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q22KREDU Total
2.00 3.00 4.00
L4 01‘8“” . 1 Count 209 70 41 320
embership
% Within
Cluster 65.3% 21.9% 12.8% 100.0%
Membership
2 Count 272 5 277
% Within
Cluster 98.2% 1.8% 100.0%
Membership
Total Count 48 l 75 4 l 597
% Within
Cluster 80.6% 12.6% 6.9% 100.0%
[Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided)1
Pearson Chi-Square 103.022 2 .000
Likelihood Ratio 129.246 2 .000
“‘fm'bflrma’ 93.456 1 .000
ssocratron
N of Valid Cases 597
0 cells (.0%) have expected count less than 5. The minimum expected count is 19.02.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

202

Q22LREDU Total
2.00 3.00 4.00
cm“ . 1 Count 4 52 264 320
embershlp
% Within
Cluster 1.3% 16.3% 82.5% 100.0%
'Membership
2 Count 71 122 82 275
% Within
Cluster 25.8% 44.4% 29.8% 100.0%
embership
Total Count 75 174 346 595
% Within
Cluster 12.6% 29.2% 58.2% 100.0%
[Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided)l
Pearson Chi-Square 181.382 2 .000
Likelihood Ratio 199.033 2 .000
”‘far'bl/‘liinea' 176.182 1 .000
ssocratron
N of Valid Cases 595
0 cells (.0%) have expected count less than 5. The minimum expected count is 34.66.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q22MREDU Total
2.00 3.00 4.00
mum" . 1 Count 14 37 260 31 1
embershrp
% Within
Cluster 4.5% 11.9% 83.6% 100.0%
embership
2 Count 14 19 241 274
% Within
Cluster 5.1% 6.9% 88.0% 100.0%
'Membership
Total Count 28 56 50] 585
% Within
Cluster 4.8% 9.6% 85.6% 100.0%
[Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 4.183 2 .124
Likelihood Ratio 4.269 2 .118
Lfiw‘bl'lrmea’ .815 1 .367
ssocratron
N of Valid Cases 585
0 cells (.0%) have expected count less than 5. The minimum expected count is 13.11.
Q22NREDU Total
2.00 3.00 4.00
[M Clus'f‘ . 1 Count 18 44 248 310
embershrp
% Within
Cluster 5.8% 14.2% 80.0% 100.0%
embership
2 Count 9 43 221 27 3
% Within
Cluster 3.3% 15.8% 81.0% 100.0%
lMembership
Total Count 27 87 469 583
% Within
Cluster 4.6% 14.9% 80.4% 100.0%
[Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided
Pearson Chi-Square 2.227 2 .328
Likelihood Ratio 2.275 2 .32]
“’far'W'Ir‘ma’ .629 1 .428
ssocratron
N of Valid Cases 583

 

 

 

 

 

 

0 cells (.0%) have expected count less than 5. The minimum expected count is 12.64.

203

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q220REDU Total
2.00 3.00 4.00
mum . 1 Count 57 120 134 311
embershrp
% Within
Cluster 18.3% 38.6% 43.1% 100.0%
Membership
2 Count 49 144 78 27]
% Within
Cluster 18.1% 53.1% 28.8% 100.0%
Membership
Total Count 106 264 2 1 2 5 82
% Within
Cluster 18.2% 45.4% 36.4% 100.0%
[Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 14.899 2 .001
Likelihood Ratio 15.007 2 .001
”riar‘bflfinear 5.567 1 .018
ssocratron
N of Valid Cases 582

 

 

 

 

 

0 cells (.0%) have expected count less than 5. The minimum expected count is 49.36.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q22PREDU Total
2.00 3.00 4.00
mum . 1 Count 122 110 76 308
embershrp
% Within
Cluster 39.6% 35.7% 24.7% 100.0%
Membership
2 Count 148 85 40 273
% Within
Cluster 54.2% 31.1% 14.7% 100.0%
Membership
Total Count 270 195 1 16 581
% Within
Cluster 46.5% 33.6% 20.0% 100.0%
[Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided
Pearson Chi-Square 14.827 2 .001
Likelihood Ratio 14.97] 2 .001
Lifar‘bY'lfinea’ 14.746 1 .000
ssocratron
N of Valid Cases 581

 

 

 

 

 

 

0 cells (.0%) have expected count less than 5. The minimum expected count is 54.51.

204

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q22QREDU Total
2.00 3.00 4.00
cm” . 1 Count 176 82 47 305
embershrp
% Within
Cluster 57.7% 26.9% 15.4% 100.0%
Membership
2 Count 213 38 19 270
% Within
Cluster 78.9% 14.1% 7.0% 100.0%
[Membership
Total Count 389 120 66 575
% Within
Cluster 67.7% 20.9% 11.5% 100.0%
[Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 29.510 2 .000
Likelihood Ratio 30.172 2 .000
“near'bY'If‘near 26.253 1 .000
Assocratlon
N of Valid Cases 575
0 cells (.0%) have expected count less than 5. The minimum expected count is 30.99.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q22RREDU Total
2.00 3.00 4.00
Clusw’ . 1 Count 18 65 227 310
embershrp
% Within
Cluster 5.8% 21.0% 73.2% 100.0%
embership
2 Count 9 21 245 275
% Within
Cluster 3.3% 7.6% 89.1% 100.0%
embership
Total Count 27 86 472 585
% Within
Cluster 4.6% 14.7% 80.7% 100.0%
[Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided
Pearson Chi-Square 24.191 2 .000
Likelihood Ratio 25.263 2 .000
”’far'bfgnea’ 17.950 1 .000
ssocratron
N of Valid Cases 585
0 cells (.0%) have expected count less than 5. The minimum expected count is 12.69.

205

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q22SREDU Total
2.00 3.00 4.00
Mgrllbsetresrhip 1 Count 139 97 74 310
% Within
Cluster 44.8% 31.3% 23.9% 100.0%
embership
2 Count 160 81 34 275
% Within
Cluster 58.2% 29.5% 12.4% 100.0%
Membership
Total Count 299 l 78 108 5 85
% Within
Cluster 51.1% 30.4% 18.5% 100.0%
[Membershi
Chi-Square Tests Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 15.690 2 .000
Likelihood Ratio 15.995 2 .000
Linear‘bY’Itinea’ 15.249 1 .000
Assocratlon
N of Valid Cases 585
0 cells (.0%) have expected count less than 5. The minimum expected count is 50.77.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

206

Q22TREDU Total
7 2.00 3.00 4.00
mum . 1 Count 58 94 157 309
embershrp
% Within
Cluster 18.8% 30.4% 50.8% 100.0%
Membership
2 Count 26 75 17 3 274
% Within
Cluster 9.5% 27.4% 63.1% 100.0%
Membership
Total Count 84 169 330 5 83
% Within
Cluster 14.4% 29.0% 56.6% 100.0%
[Membership
Chi-Square Tests Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 13.048 2 .001
Likelihood Ratio 13.318 2 .001
Li‘far'b¥'l:i“ea’ 12.725 1 .000
ssocratron
N of Valid Cases 583
0 cells (.0%) have expected count less than 5. The minimum expected count is 39.48.

Appendix F:

BMP Analysis

207

 

BMP Results and Preferences

 

Would

Do

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

PP° P90 PP0 “PP Strongly Know P
0208 Frequency 21 98 193 239 36 21 608
Valid Percent 3.5 16.1 31 .7 39.3 71 5.9 3.5 100
Q20b Frequency 53 163 216 1 18 39 19 608
Valid Percent 8.7 26.8 35.5 19.4 54.9 6.4 3.1 100
Q20e Frequency 1 6 65 238 24 l 30 l 8 608
Valid Percent 2.6 10.7 39.1 39.6 78.7 4.9 3 100
Q20d Frequency 10 42 151 376 l6 13 608
Valid Percent 1.6 6.9 24.8 61.8 86.6 2.6 2.1 100
Q20e Frequency 2 l 60 206 286 19 l 6 608
Valid Percent 3.5 9.9 33.9 47 80.9 3.1 2.6 100
Q20f Frequency 12 75 260 227 l 7 1 7 608
Valid Percent 2 12.3 42.8 37.3 80.1 2.8 2.8 100
Q20 Frequency 9 72 278 210 23 16 608
Valid Percent 1.5 l 1.8 45.7 34.5 80.2 3.8 2.6 100
Would Do
Individualist Cluster “é?“ 1:” SMigh; SWoulgt Ssh-0mg and Not R No Total
ppo uppo uppo uppo Strong] Know espouse
Q20a Frequency 16 59 l 15 91 31 8 320
Valid Percent 5 18.4 35.9 28.4 64.3 9.7 2.5 100
Q20b Frequency 3 l 93 l 20 42 26 8 320
Valid Percent 9.7 29.1 37.5 13.1 50.6 8.1 2.5 100
Q20c Frequency 13 48 135 99 20 5 320
Valid Percent 4.1 15 42.2 30.9 73.1 6.3 1.6 100
Q20d Frequency 8 33 90 l 73 12 4 320
Valid Percent 2.5 10.3 28.1 54.1 82.2 3.8 1.3 100
Q20e Frequency 19 48 l 17 l 16 16 4 320
Valid Percent 5.9 15 36.6 36.3 72.9 5 1.3 100
Q20f Frequency 8 48 142 104 l 3 5 320
Valid Percent 2.5 15 44.4 32.5 76.9 4.1 1.6 100
Q20; Frequency 5 41 l 61 93 l 5 5 320
Valid Percent 1.6 12.8 50.3 29.1 79.4 4.7 1.6 100

 

 

 

 

 

 

 

 

 

 

208

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Would . Would Do
Egalitarian Cluster Not SMlgh; SWoultll1 :tronglly and Not R No Total
Slpmn up po uppo "FPO Stron Iy Know esponse
Q2 08 Frequency 5 38 78 148 4 6 279
Valid Percent 1.8 13.6 28 53 81 1.4 2.2 100
Q20b Frequency 22 69 96 76 l 3 3 279
Valid Percent 7.9 24.7 34.4 27.2 61.6 4.7 1.1 100
Q20e Frequency 2 l 7 l 03 l 42 9 6 279
Valid Percent 0.7 6.1 36.9 50.9 87.8 3.2 2.2 100
Q20d Frequency 2 8 61 202 4 2 2 79
Valid Percent 0.7 2.9 21.9 72.4 94.3 1.4 0.7 100
Q20e Frequency 1 l 2 89 l 70 3 4 279
Valid Percent 0.4 4.3 31.9 60.9 92.8 1.1 1.4 100
Q20f Frequency 4 26 l 18 123 4 4 279
Valid Percent 1.4 9.3 42.3 44.1 86.4 1.4 1.4 100
Q20g Frequency 4 30 1 l7 1 l7 7 4 279
Valid Percent 1.4 10.8 41.9 41.9 83.8 2.5 1.4 100
Independent Samples Test of BMP at .01 Level of Significance
lLevene's Test
for Equality o t-test for Equality of Means
Variances
0
F Sig dr Sig' (2' Me“ S‘d‘ Em” 9:11:37???
° tailed) IfferenceDifference Difference
Lower Upper
[Q20Al Equal
variance 1.051 .306 -5.214 548 .000 -.3717 .07130 -.55603 -.1875
assume;
Equal variances not assumed -5.225 546.476 .000 -.3717 .071 15 -.55565 -.I 878
IQ20B] Equal
variancej .242 .623 -3.312 547 .001 -.2544 .0768] -.45296 -.0559
assume
Equal variances not assumed -3.302 533.990 .001 -.2544 .07704 -.45358 -.0553
IQ20C1 Equafl
variance .721 .396 ~5.926 557 .000 -.3736 .06304 -.53653 -.2106
assume
Equal variances not assumed -6.000 549.643 .000 -.3736 .06226 -.53452 -.2126
02001 Equais
variance 1.958 .000 -5.012 575 .000 -.2881 .05747 -.43661 -.1395
assume
Equal variances not assumed -5.101 547.945 .000 -.2881 .05648 -.43406 -. l 421
IQ20E] Equafi
variance 19.459 .000 -7.394 570 .000 -.4735 .06404 -.63905 -.3080
assume
Egal variances not assumed -7.533 526.218 .000 -.4735 .06286 -.63604 -.31 10
Q20F1 Equza
variance .113 .737 -3.156 571 .002 -.1960 .06210 -.35645 -.0355
assume
Equal variances not assumed -3.170 570.630 .002 -.l960 .06182 -.35573 -.0362
0206] Equal
variancej 5.410 .020 -2.582 566 .010 -.1548 .05994 -.30969 .0001
assume
Equal variances not assumed -2.578 555.067 .010 -.1548 .06003 -.30994 .0004

 

 

 

 

 

 

 

 

 

 

209

 

Appendix G:

Regression Analysis Output

210

Sample Population Results

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Descriptive Statistics
Mean Std. Deviation N
INTENT] 3.2345 .53775 496
KNOWALL 2.7883 .95428 496
GROUP .5020 .50050 496
LOC 3.6649 .58922 496
PERSONAL 3.7730 .60779 496
Q23 33.8931 18.96306 496
Q24 16.0464 13.93092 496
Q26 4.02 1 .324 496
Q27 3.48 1.054 496
Q28 6.24 2.049 496
Model Summary

 

Model R R Square Adjusted R Std. Error of the
Square Estlmate
l .557 .310 .297 _45077

Predictors: (Constant), Q28, LOC, Q26, KNOWALL, GROUP, Q27, PERSONAL, Q24,
Q23
Dependent Variable: INTENT]

 

 

 

 

 

 

 

 

ANOVA

M ode] Sum of d f Mean

Squares Square

1 Regression 44.388 9 4.932 24.272 .000

Residual 98.753 486 .203
Total 143.14] 495

Predictors: (Constant), Q28, LOC, Q26, KNOWALL, GROUP, Q27, PERSONAL, Q24,

Q23
Dependent Variable: INTENT]

 

 

 

 

 

 

 

 

 

 

 

 

 

211

Coefficients

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Unstandardized Standardized t Sig 95% Confidence
Coefficients Coefficients ' Interval for B

Std. Lower Upper

Model 1 B Error Beta Bound Bound
(Constant) 1.336 .188 7.116 .000 .967 1.705
KNOWALL -5.030E—03 .022 -.009 -.227 .821 -.049 .039
GROUP .148 .043 .137 3.426 .001 .063 .232
LOC .324 .039 .355 8.310 .000 .247 .400
PERSONAL .158 .039 .179 4.070 .000 .082 .234
Q23 -1.071E-03 .002 -.038 -.686 .493 -.004 .002
Q24 -2.322E—03 .002 -.060 -1.1 l l .267 -.006 .002
Q26 -1.065E-02 .022 -.026 -.492 .623 -.053 .032
Q27 4.816E-02 .023 .094 2.090 .037 .003 .093
Q28 6.289E-04 .01 1 .002 .059 .953 -.020 .022

Dependent Variable: INTENT]
Residual Statistics (dependent variable: INTENTI)
Unstandardized tandardized t Sig 95% Confidence
Coefficients Coefficients ' Interval for B

Std. Lower Upper

Model 1 B Error Beta Bound Bound
(Constant) 1.336 .188 7.116 .000 .967 1.705
KNOWALL -5.030E-03 .022 -.009 -.227 .821 -.049 .039
GROUP .148 .043 .137 3.426 .001 .063 .232
LOC .324 .039 .355 8.310 .000 .247 .400
PERSONAL .158 .039 .179 4.070 .000 .082 .234
Q23 -1.071E-03 .002 -.038 -.686 .493 -.004 .002
Q24 -2.322E-O3 .002 -.060 -1 .1 l 1 .267 -.006 .002
Q26 -1 .065E-02 .022 -.026 -.492 .623 -.053 .032
Q27 4.816E—02 .023 .094 2.090 .037 .003 .093
Q28 6.289E-04 .01 l .002 .059 .953 -.020 .022

 

 

212

 

Cluster 1: Individualists

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Descriptive Statistics
Mean Std. Deviation N
INTENT] 3.0596 .5712] 320
KNOWALL 2.6406 .98208 320
MORAL] 3.8153 .5221] 320
RESIST] 3.3010 .40889 320
LOC 3.5689 .63913 320
PERSONAL 3.6275 .59496 320
GROUP .0000 .00000 320
Q23 39.0280 18.95577 320
Q24 20.9189 15.79435 320
Q26 4.33 1.312 320
Q27 3.22 1.053 320
Q28 6.04 1.977 320
Model Summary
R Adjusted Std. Error Change
R Square R Of the Statistics
Square Estimate
R .
F Slg. F
Model (83:21:: Change dfl df2 Change
1 .563 .317 .295 .47964 .317 14.344 10 309 000

 

 

 

 

 

 

 

 

 

 

 

LOC, Q23, Q26
Dependent Variable: INTENT]

 

 

 

 

 

 

 

 

 

 

 

 

 

Predictors: (Constant), Q28, Q24, PERSONAL, RESIST], KNOWALL, Q27, MORAL],

ANOVA

Model Sum 0f df Mean 1= Sig.

Squares Square
1 Regression 32.999 10 3.300 14.344 .000
Residual 71.086 309 .230
Total 104.085 319

Predictors: (Constant), Q28, Q24, PERSONAL, RESIST], KNOWALL, Q27, MORAL] ,
LOC, Q23, Q26

Dependent Variable: INTENT]

213

 

Coefficients

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

nstandardized Standardized t Sig
Coefficients Coefficients '
Model B Std. Error Beta
1 (Constant) 1 .463 .338 4.325 .000
KNOWALL -3.91 1E-02 .029 -.067 -1.351 .178
MORAL] .352 .060 .322 5.828 .000
RESIST] -.172 .068 -.123 -2.523 .012
LOC .290 .052 .324 5.580 .000
PERSONAL 1.714E-02 .057 .018 .301 .764
Q23 -7.751E-04 .002 -.026 -.375 .708
Q24 9.441E-04 .003 .026 .363 .717
Q26 -5.537E-02 .030 -.127 -1.842 .066
Q27 2.297E-02 .030 .042 .761 .447
Q28 7.620E-04 .015 .003 .051 .959
Dependent Variable: INTENT]
Residual Statistics
Minimum Maximum Mean Std. Deviation N
Predicted Value 2.0310 3.8620 3.0596 .32163 320
Std. Predicted Value -3.198 2.495 .000 1.000 320
Standard Error of
Predicted Value .04432 .16521 .08639 .0211] 320
33:65“! Pred‘c‘ed 1.9828 3.8664 3.0586 .32267 320
Residual -1.6707 1.3591 .0000 .47206 320
Std. Residual -3.483 2.834 .000 .984 320
Stud. Residual -3.527 2.882 .001 1.003 320
Deleted Residual -1.7125 1.4058 .0010 .48994 320
it“? Deleted -3594 2.917 .000 1.007 320
esrdual
Mahal. Distance 1.727 36.850 9.969 5.623 320
Cook's Distance .000 .056 .003 .006 320
86mm“ ”average .005 .116 .031 .018 320
alue

 

 

 

 

 

 

 

Dependent Variable: INTENT]

214

 

 

Cluster 2: Egalitarians

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Descriptive Statistics
Mean Std. Deviation N

INTENT] 3.3622 .47909 279

KNOWALL 2.8566 .9450] 279

MORAL] 4.0207 .45336 279

RESIST] 2.2461 .4145 1 279

LOC 3.7307 .50986 279

PERSONAL 3.8806 .56444 279

Q23 30.8974 18.61404 279

Q24 13.4420 12.59847 279

Q26 3.80 1.306 279

Q27 3.68 1.018 279

Q28 6.49 1.911 279

Model Summary

. Std.
Ad usted
R R JR Error Of Change Statistics
Square Square the
Estlmate
R .
Model Square Ch; 6 dfl dt2 Sign:
Change g g
l .555 .308 .282 .4059] .308 11.926 10 268 .000

 

 

 

 

 

 

 

 

 

 

 

Predictors: (Constant), Q28, LOC, Q26, RESIST], KNOWALL, MORAL1,Q27,
PERSONAL, Q24, Q23

Dependent Variable: INTENT]

 

 

 

 

ANOVA
Model sum 0f df Mean F Sig.
Squares Square
1 Regression 19.650 10 1.965 1 1.926 .000
Residual 44. 15 7 268 . 165
Total 63.808 278

 

 

 

 

 

 

 

 

 

Predictors: (Constant), Q28, LOC, Q26, RESIST], KNOWALL, MORAL1,Q27,
PERSONAL, Q24, Q23

Dependent Variable: INTENT]

215

 

Coefficients

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Unstandardized Standardized t Sig
Coefficients Coefficients '
Model B Std. Error Beta
1 (Constant) .794 .340 2.339 .020
KNOWALL -9.398E-03 .027 -.019 -.351 .726
MORAL] .259 .057 .245 4.571 .000
RESIST] -2.022E-02 .060 -.017 -.339 .735
LOC .256 .052 .272 4.883 .000
PERSONAL .104 .048 .122 2.139 .033
Q23 -3. 143E-06 .002 .000 -.002 .999
Q24 -2.174E-03 .003 -.057 -.825 .410
Q26 -2.471E-02 .026 -.067 -.943 .347
Q27 9.313E-02 .029 .198 3.262 .001
Q28 3.500E-03 .014 .014 .257 .797
Dependent Variable: INTENT]
Residuals Statistics
. . . Std.
iMlnrmum Maxrrnum Mean Deviation N
Predicted Value 2.1044 4.0493 3.3622 .265 86 279
Std. Predicted Value -4.731 2.585 .000 1.000 279
Standard Error of
Predicted Value .03719 .22182 .0779] .02069 279
Adjuswd hemmed 1.7231 4.0861 3.3608 .27335 279
Value
Residual -1.7356 1.1130 .0000 .39855 279
Std. Residual -4.276 2.742 .000 .982 279
Stud. Residual -4.402 2.806 .002 1.007 279
Deleted Residual -1.8392 1.2769 .0014 .41972 279
St“? Deleted -4.561 2.843 .001 1.013 279
Resrdual
Mahal. Distance 1.337 82.025 9.964 7.163 279
Cook's Distance .000 .269 .005 .019 279
66”“ mange .005 .295 .036 .026 279
alue
Dependent Variable: INTENT]
216

 

 

BIBLIOGRAPHY

217

BIBLIOGRAPHY

Ajzen, I. (2002). Theory of Planned Behavior Diagram. Available: Http://www-
unix.oit.umass.edu/~ajzen/tpb.diag.html [2002, April 27]

Ajzen, l. (1991). The Theory of Planned Behavior. Organizational Behavior and Human
Decision Processes, 50, 179-211.

Ajzen, I. (1985). From Intentions to Actions: A Theory of Planned Behavior. in J.
Kuhland & J. Beckman (Eds), Action-Control: From Cognitions to Behavior (pp.
11-39). Heidelberg: Springer.

Ajzen, I. & M. Fishbein (1980). Understanding Attitudes and Predicting Social Behavior.
Englewood Cliffs: Prentice Hall.

Ajzen, I. & M. Fishbein (1977). Attitude - Behavior Relations: A Theoretical Analysis
and Review of Empirical Research. Psychological Bulletin, 84(5), 888-918.

Allen, J. B. & J .L. Ferrand (1999). Environmental Locus of Control, Sympathy, and
Proenvironmental Behavior. Environment and Behavior, 31(3), 338-353.

Arcury, T. A. & T.P. Johnson (1987). Public Environmental Knowledge: A Statewide
Survey. Journal of Environmental Education, 18, 31-37.

Arcury, T. A., Johnson, T. P. & SJ. Scollay (1986). Ecological Worldview and
Environmental Knowledge: An Examination of The New Environmental
Paradigm. Journal of Environmental Education, I 7, 35-40.

Axelrod, L. J. (1994). Balancing Personal Needs with Environmental Preservation:
Identifying the values that guide decisions in ecological dilemmas. Journal of
Social Issues, 50(3), 85-104.

Babbie, E. (1990). Survey Research Methods. (2nd Ed.). Belmount, CA: Wadsworth.

Ball-Rokeach, S. J ., Rokeach, M. & J .W. Grube (1984). The Great American Value Test:
Influencing Behavior and Belief Through Television. New York: Free Press.

Bamberg, S., Kuhnel, S. M. & P. Schmidt (1999). The Impact of General Attitudes on
Decisions. Rationality and Society, 11(1), 5-25.

Bandura, A., Adams, N. E., Hardy, A. B. & G.N. Howells (1980). Tests of the Generality
of Self-Efficacy Theory. Cognitive Therapy and Research, 4, 39-66.

218

Bardwell, L. (1991). Problem Framing: A Perspective on Environmental Problem
Solving. Environmental Management, I 5, 603-6 12.

Barrett, M. E. (2000, November). Structural BMPs.: The Solution to Water Quality
Problems? Paper presented at the Water Quality And Quality Issues In Coastal
Urban Areas, American Water Resources Association, Miami, FL.

Bawden, R. (1995). Systemic Development: A Learning Approach to Change (pp. 67).
Western Sydney: Centre For Systemic Development, University Of Western
Hawkesbury.

Bazerman, M. H., Messick, D. M., Tenbrunsel A.E. & K.A. Wade - Benzoni (1997).
Environment, Ethics, and Behavior. San Francisco: New Lexington Press.

Beckwith, J. A. & S.M. Ray] (2002). The Role of Knowledge and Responsibility in the
Environmental Behavior of Mid-Westem University Students. Environmental
Knowledge & Behavior, Article Under Review.

Bengston, D. N. (1994). Changing Forest Values and Ecosystem Management. Society
and Natural Resources, 7, 5 15-5 34.

Biawas, A. B. (1973). Socio-Economic Consideration in Water Resources Planning.
Water Resources Bulletin, 9, 746-754.

Biawas, A. K. & RW. Durie (1971). Sociological aspects of Water Development. Water
Resources Bulletin, 7, 1137-1143.

Boerschig, S. & R. Deyoung (1993). Evaluation of Selected Recycling Curricula:
Educating the Green Citizen. Journal of Environmental Education, 24(3), 17-22.

Bosselrnann, Klaus (1997, October) Human Rights and the Environment: Redefining
Fundamental Principles? Paper Presented at Environmental Justice: Global Ethics
For The 21St Century, Melbourne, Australia

Boster, F. (1999). Communications 901 : Path Analysis Class Text.
Bright, A. D., & M.J. Manfredo (1997). The Influence of Balanced lnforrnation on
Attitudes Towards Natural Resource Issues. Society and Natural Resources, 10,

469-483.

Brown, G., & C. Harris (1992). The US. Forest Service: Towards the New Resource
Management Paradigm? Society and Natural Resources, 5, 231-245.

Burroughs, R. (1999). When Stakeholders Choose: Process, Knowledge, and Motivation
in Water Quality Decisions. Society and Natural Resources, 12, 797-809.

219

Buss, D. M., Criak, K. H. & K.M. Dake (1986). Contemporary Worldviews and
Perceptions of the Technological System. in V. T. Covello, J. Menkes, & J.
Mumpower (Eds), Risk Evaluation and Management (pp. 93-130). New York:
Plenum.

Buss, D., Craik, K H. & KM. Dake (1985). Perceptions of Decision Procedures for
Managing and Regulating Hazards. in F. Homburger (Ed.), Safety Evaluation and
Regulation (pp. 199-208). New York: Kager.

Carson, R. T. Louviere, J. J. & D.A. Anderson (1994). Experimental Analysis of Choice.
Marketing Letters, 5, 351-368.

Christianson, E. H. & T.A. Arcury (1992). Regional Diversity in Environmental
Attitudes, Knowledge, and Policy: The Kentucky River Authority. Human
Organization, 51(2), 99-108.

Clean Water Act. (1972). 33 U.S.C. 1251-1387.
Cohn, J. (1992). A Power Primer. Psychological Bulletin, 112, 155-159.

Collins, H. M. & J. P. Pinch (1982). Frames of Meaning: The Social Construction of
Extraordinary Science. London: Routledge and Kegan Paul.

Conner, M. & C.J. Arrnitage (1998). Extending the Theory of Planned Behavior: A
Review and Avenues for Further Research. Journal of Applied Social Psychology,
28, 1429-1464.

Converse, P. E. (1964). The Nature of Belief Systems in Mass Publics. in D. E. Apter
(Ed), Ideology and Discontent (pp. 206-261). New York: The Free Press.

Crosby, F., Bromley, S. & L. Saxe (1980). Recent Unobtrusive Studies of Black and
White Discrimination and Prejudice: A Literature Review. Psychological Bulletin.
87, 546-563.

Dake, K. (1992). Myths of Nature: Culture and the Social Construction of Risk. Journal
of Social Issues, 48(4), 21-37.

Dake, K. (1991). 'Orienting Dispositions in the Perceptions of Risk: An Analysis of
Contemporary Worldviews and Cultural Biases. Journal of Cross-Cultural
Psychology, 22, 61-82.

Dillman, D. A. (2000). Mail and Internet Surveys. (2nd Ed.). New York, NY: John Wiley
& Sons, Inc.

Douglas, M. (1975). Implicit Meanings: Essays in Anthropology. london: Routledge &
Kegan Paul.

220

Douglas, M. (1990). Risk as a Forensic Resource. Daedalus, 119(4), 1-16.
Douglas, M. (1986). How Institutions Think. Syracuse, NY: Syracuse University Press.

Douglas, M. (1985). Risk Acceptability According to the Social Sciences. New York:
Russell Sage Foundation.

Douglas, M. & A. Wildavsky (1982a). Risk and Culture: An Essay on the Selection of
Technological and Environmental Dangers. Berkley: University Of California
Press.

Douglas, M. E. (1982b). Essays in the Sociology of Perceptions. London: Routledge &
Kegan Paul.

Dunlap, R. & R. Scarce (1991). The Polls & Poll Trends: Environmental Problems and
Protection. Public Opinion Quarterly, 55, 713-734.

Dunlap, R. E. & K Van Liere (1978). The New Environmental Paradigm. Journal of
Environmental Education, 9, 10-19.

Dunlap, R. E., Van Liere, K. D., Mertig, A. G. & R.E. Jones (2000). Measuring
Endorsement of the New Ecological Paradigm: A Revised NEP Scale. Journal of
Social Issues, 56(3), 425-442.

Eckersley, R. (1992). Environmentalism and Political Theory: Towards an Ecocentric
Approach. Albany: State University of New York.

Eden, 8. E. (1993). Individual Environmental Responsibility And Its Role In Public
Environmentalism. Environment And Planning, 25, 1743-1758.

Edgell, M. C. R. & D.E. Nowell (1989). The New Environmental Paradigm Scale:
Wildlife and Environmental Beliefs in British Columbia. Society and Natural
Resources, 2, 285-296.

Farley, J. U., Lehmann, D. R. & M.J. Ryan (1981). Generalizing from "Imperfect"
Replication. Journal of Business, 54, 597-610.

Feather, N. T. (1979). Value Correlates of Conservatism. Journal of Personality and
Social Psychology, 3 7(9), 1617-1630.

Festinger, L. (1962). A Theory of Cognitive Dissonance, (2nd Ed.) California: Stanford
University Press.

Fishbein, M. & I. Ajzen (1975). Belief, Attitude, Intention and Behavior: An Introduction
to Theory and Research. Reading, MA: Addision-Wesley.

221

 

Fiske, S. T. & S.E. Taylor (1990). Structures of Social Life: The Four Elementary F arms
of Human Relations. New York, NY: The Free Press.

Folger, R. (1996). Distributive and Procedural: Multi-Faceted Meaning and
Interrelationships. Social Justice Research, 9, 371-394.

Fortner, R W. & A.E. Lyon (1985). Effects of a Cousteau Television Special on Viewer
Knowledge and Attitudes. Journal of Environmental Education, 16(3), 12-20.

Fulton, D. C., Manfedo, M. J ., & J. Lipscomb (1996). Wildlife Value Orientation: A
Conceptual and Measurement Approach. Human Dimensions of Wildlife, I , 24-
47.

Furby, L. (1986). Psychology and Justice. in R. C. Cohen (Ed.), Justice: Views from the
Social Sciences (pp. 153-203). New York, NY: Plenum Press.

Furnham, A. F. (1988). Lay Theories: Everyday Understanding of Problems in the Social
Sciences. Tarrytown, NY: Pergamon.

Gallup, R (1989). Environment Regaining a F oothold on the National Agenda (Report
285). Princton, N.J.: Gallup Organization.

Geller, S. E. (1995). Integrating Behaviorism and Humanism for Environmental
Protection. Journal of Social Issues, 51(4), 179-195.

Gregory, R., Keeney, R. & D. Von Winterfeldt (1992). Adapting the Environmental
Impact Statement Process to Inform Decision Making. Journal of Policy Analysis
and Management, I, 58-75.

Guagnano, G. A. (1995). Locus of Control, Altruism and Agentic Disposition.
Population and Environment: A Journal of Interdisciplinary Studies, 17(1), 63-
77.

Hamilton, J. P. (1986). Environmental locus of Control as a Function of the Perceived
Importance of Environmental Problems and Environmental Knowledge. Journal
of Interpretation, 1 1(2), 15-31.

Harris, D. H. (1977). The Human Dimensions of Water-Resources Planning. Human
Factors, 19(3), 241-251.

Herberlein, T. A. (1972). The Land Ethic Realized: Some Social Psychology

Explanations for Changing Environmental Attitudes. Journal of Social Issues,
28(4), 79-87.

222

Hines, J. M., Hungerford, H. R. & A.N. Tomera (1986). Analysis Synthesis of Research
on Responsible Environmental Behavior: A Meta-Analysis. Journal of
Environmental Education, 18(2), 1-18.

Homer, P. M. 8: L. R. Kahle (1988). Astructural Equation Test of the Value-Attitude-
Behavior Hierarchy. Journal of Personality and Social Psychology, 5 4, 638-646.

H00per, J. R., & J. M. Neilsen (1991). Recycling as Altruistic Behavior: Normative and
Behavioral Strategies to Expand Participation in a Community Recycling
Program. Environment and Behavior, 23(2), 195-220.

Hrubes, D., Ajzen, I. & J. Daigle (2001). Predicting Hunting Intentions and Behavior: an
Application of the Theory of Planned Behavior. Leisure Sciences, 23(4), 165-178.

Hungerford, H. R. & T.L. Volk (1990). Changing learner Behavior Through
Environmental Education. Journal of Environmental Education, 21(3), 8-21.

Hwang, Y.-H., Kim, S.-I. & J.M. Jeng (2000). Examining the Causal Relationships
among Selected Antecedents of Responsible Environmental Behavior. Journal of
Environmental Education, 31(4), 19-25.

Jackman, M. R. (1978). General and Applied Tolerance: Does Education Increase
Commitment to Racial Integration? American Journal of Political Science, 22,
302-324.

Jorgensen, B. S. & G. J. Syme (2000). Protest Responses and Willingness to Pay:
Attitude Toward Paying for Storrnwater Pollution Abatement. Ecological
Economics, 33(2), 251-265.

Katz, P. A. (1976). The Acquisition of Racial Attitudes in Children. In P. A. Katz (Ed.),
Toward the Elimination Of Racism (Pp. 125-154). New York: Pergamon.

Kempton, W., Boster, B. & J. Harley (1995). Environmental Values in American Culture.
Cambridge, MA: MIT Press.

Kramer, R. M., McClintock, C. G. & D.M. Messick (1986). Social Values and
Cooperative Response to a Simulated Resource Conservation Crisis. Journal of
Personality, 54, 576-592.

lane, R. (1973). Patterns of Political Belief. in J. Knutson (Ed.), Handbook of Political
Psychology (pp. 83-116). San Francisco: Jossey—Bass.

Lefcourt, H. M. (1982). Locus of Control - Current Trends in Theory and Research. (2nd
Ed.). Hillsdale, NJ: Erlbaum.

223

 

Lind, E. A., Kulik, C. T., Ambrose, M. & M. Park (1995). Procedural Justice and
Arbitrated Dispute Resolution: The Role of Fairness Judgments in Individual and
Corporate Decisions. Administrative Science Quarterly, 38(2), 224-252.

Lind, E. A. (1992, March, 1992). The Fairness Heuristic: Rationality and "Relationality"
in Procedural Evaluation. Paper presented at the fourth international conference
of The Society For The Advancement Of Socio-Economics, Irvine, CA.

Lind, A. E. & T.R. Tyler (1988). The Social Psychology of Procedural Justice. (Vol. 9).
New York, NY: Plenum.

Lovrich, N. P. J., Pierce, J. C., Tsurutani, T. & T. Abe (1986). Policy Relevant
Information and Public Attitudes: IS Public Ignorance a Barrier to Nonpoint
Pollution Management? Water Resources Bullein, 22, 229-236.

Luzar, J. E. & KJ. Cosse' (1998). Willingness To Pay or Intention To Pay: The Attitude-
Behavior Relationship in Contingent Valuation. Journal of Socio-Economics,
27(3), 427-444.

Lynne, G. D. (1995). Conservation Technology Adoption Decisions and the Theory of
Planned Behavior. Journal of Economic Psychology, 1 6, 581-598.

MacCoun, R. J ., Lind, E. A., Hensler, D. R., Bryant, D. L. & P.A. Ebener (1988).
Alternative Adjudication: An Evaluation of the New Jersey Automobile
Arbitration Program. Santa Monica, CA: RAND Corporation: Institute of Civil
Justice.

Madden, T., Scholder Ellen, P. & I. Ajzen (1992). A Comparison of the Theory of
Planned Behavior and the Theory of Reasoned Action. Personality and Social
Psychology Bulletin, 18(1), 3-9.

Manfrado, M. J. & H.C. Zinn (1996). Population Change and its Implications for Wildlife
Management in the New West: A Case Study of Colorado. Human Dimensions of
Wildlife, 1, 62-74.

Mars, J. (1982). Cheats at Work: An Anthropologv of Worlgolace Crime. London: Allen
& Unwin.

Maslow, A. H. E. (Ed.). (1959). New Knowledge in Human Values. New York: Harper.
McConahay, J. M. (1986). Modern Racism, Ambivalence, and the Modern Racism Scale.

in J. F. Dovidio & S. L. Gaertner (Eds), Prejudice, Discrimination, and Racism
(pp. 91-124). Orlando, FL: Academic Press.

224

McFarlane, B.L. & P.C. Boxall (2000). Factors Influencing Forest Values and Attitudes
of Two Stakeholder Groups: The Case of the Foothills Model Forest, Alberta,
Canada, Society & Natural Resources 13: 649-661

McKenzie-Mohr, D., Nemiroff, L. S., Beers, L., & S. Desmarais (1995). Determinants of
Responsible Environmental Behavior. Journal of Social Issues, 51(4), 139-156.

Messick, D. M. (1993). Equity as a Decision Heuristic. in B. A. Mellers & J. Baron
(Eds.), Psychological Perspectives on Justice: Theory and Applications (pp. 1 1-
31). New York: Cambridge University Press.

Messick, D. M. & C.G. McClintock (1968). Motivational Bases of Choice in
Experimental Games. Journal of Experimental Social Psychology, 4, 1-25.

Meyer, Gary D. (1996). Meeting Public Expectations - Judicial Review of Environmental
Impact Statements in the United States: Lessons for Reform in Western Australia?
Murdock University Electronic Journal Of Law, 3(2)

Michigan Information Center (MIC) (1999). County Estimates for Median Household
Income for Michigan: 1998. State Of Michigan. Available: Http://www.
state.mi.uS/dmb/mic/census/econ/98hhincome.htm [2002, March 15].

Miller, C. A. (2000). The Dynamics of Framing Environmental Values and Policy: Four
Models of Societal Processes. Environmental Values, 9, 211-233.

Montada, L. & E. Kals (2000). Political Implications of Psychological Research in
Ecological Justice and Proenvironmental Behavior. International Journal of
Psychology, 35(3), 168-176.

Municipal Research & Services Center (MRSC) (2000). Best Management Practices
(BA/H’s) For Storm and Surface Water. Available: http://www.mrsc.org/
environment/water/water-s/S W-BMP.htm [2001, August 13*]

Nancarrow, B. E., Smith, L. M. & G.J. Syme (1996-7). The Ways People Think About
Water. Journal of Environmental Systems, 25(1), 15-27.

National Resource Conservation Service (NRCS) (1900). Sycamore Creek Watershed
Water Quality Plan. Michigan State University Cooperative Extension Service,
Agricultural Stabilization and Conservation Service, Ingham County, Michigan.

Newhouse, N. (1990). Implications of Attitude and Behavior Research for Environmental
Conservation. Journal of Environmental Education. 22(1), 26-32.

O'Riordan, T. & S. Rayner (1991). Risk Management for Global Environmental Change.
Global Environmental Change, 1(4), 91-108.

225

 

Parcels, G. S. (1984). Theoretical Models for Application in School Health Education
Research. Journal of School Health, 54(6), 39-49.

Peterson, R. S. (1994). The Role of Values in Predicting Fairness Judgments and Support
of Affirmative Action. Journal of Social Issues, 50(4), 95-115.

Phares, E. J. (1976). Locus of Control in Personality. Morristown, NJ: General Learning
Press.

Pierce, J. C. (1979). Conflict and Consensus in Water Politics. Western Political
Quarterly, 32(3), 307-319.

Pierce, J. C., Lovrich, N. P. J ., Tsurutani, T. & T. Abe (1989). Public Knowledge and
Environmental Politics in Japan and the United States. Boulder, CO: Westview
Press.

Pierce, J. C., Steger, M. E., Steel, B. S. & N.P. Lovrich (1992). Citizens, Political
Communication and Interest Groups: Environmental Organizations in Canada
and the United States. Westport, CT: Praeger.

Rasinski, K. A. (1987). What's Fair is Fair - Or is it? Value Differences Underlying
Public Views About Social Justice. Journal of Personality and Social Psychology,
53(1), 201-211.

Rayner, S. (19883). Muddling Through Metaphors to Maturity: A Commentary on
Kasperson Et Al., The Social Amplification of Risk. Risk Analysis, 8, 201-204.

Rayner, S. (1988b). The Rules That Keep US Equal: Complexity and the Costs of
Egalitarian Organizations. In J. G. Flanagan & S. Rayner (Eds.), Rules, Decisions
and Inequity in Egalitarian Societies (pp. 20-42). Aldershot, England: Avebury.

Rayner, S. (1986). Management of Radiation Hazards in Hospitals: Plural Rationalities in
a Single Institution. Social Studies of Science, 16, 573-591.

Reading, R. P., Clark, T. W. & S.R. Kellert (1994). Attitudes and Knowledge of People
Living in the Greater Yellowstone Ecosystem. Society and Natural Resources, 7,
349-365.

Roesner, L. A., Bledsoe, B. P. & R.W. Brashear (2001). Are Best Management Practices
Criteria Really Environmentally Friendly? Journal of Water Resources Planning
and Management, May/June, 150 - 154.

Rokeach, M. (1973). The Nature of Human Values. New York: Free Press.

Rokeach, M. (1979). Understanding Human Values: Individual and Societal. New York:
The Free Press.

226

Rotter, J. B. (1966). Generalized Expectancies for lntemal Versus External Control of
Reinforcement. Psychological Monographs, 80(1).

Rotter, J. B. (1975). Some Problems and Misconceptions Related to the Construct of
Internal Versus External Control or Reinforcement. Journal of Consulting and
Clinical Psychology, 43(1), 56-67.

Runes, D. (1983). Dictionary of Philosophy. New York, NY: Philosophical Library.

Samdahl, D. M. & R. Robertson (1989). Social Determinants of Environmental Concern.
Environment and Behavior, 21(1), 56-81.

Satterfield, T. 8: R. Gregory (1998). Reconciling Environmental Values and Pragmatic
Choices. Society and Natural Resources, 1 I , 629-647.

Schumann, H., Stech, C. & L. Bobo (1985). Racial Attitudes in America: Trends and
Interpretations. Cambridge, MA: Harvard University Press.

Schwartz, S. H. (1992). Universals in the Content and Structure of Values: Theoretical
Advances and Empirical Tests in 20 Countries. Advances on Experimental Social
Psychology, 25, l - 65.

Schwartz, M. & M. Thompson (1990). Divided We Stand: Redefining Politics,
Technology and Social Choice. Philadelphia: University of Pennsylvania Press.

Schwartz, S. H. & W. Bilsky (1987). Towards A Universal Psychology Structure of
Human Values. Journal of Personality and Social Psychology, 53(3), 550-562.

Schwartz, S. H. (1977). Normative Influences on Altruism. in L. Berkowitz (Ed.),
Advances in Experimental Social Psychology (Vol. 10, pp. 221-279). New York:
Academic Press.

Schwartz, S. H. (1970). Moral Decision Making and Behavior. in J. Macauley & L.
Berkowitz (Eds.), Altruism and Helping Behavior (Pp. 127-141). New York:
Academic Press.

Sears, D. O., & D.R. Kinder (1971). Racial Tensions and Voting Behavior in Los
Angeles. in W. Z. Hirsch (Ed.), Los Angeles.: Viability and Prospects for
Metropolitan Leadership (pp. 51-88). New York: Preager.

Seligman, C., Syme, G. J. & R. Gilchrist (1994). The Role of Values and Ethical
Principles in Judgments of Environmental Dilemmas. Journal of Social Issues,
50(3), 105-119.

227

Sheppard, B. H., Hartwick, J. & P.R. Warshaw (1988). The Theory of Reasoned Action:
A Meta-Analysis of Past Research with Recommendations for Modifications and
Future Research. Journal of Consumer Research, 15, 325-343.

Shultz, P. W. & S. Oskamp (1996). Effort as a Moderator of the Attitude-Behavior
Relationship: General Environmental Concern and Recycling. Social Psychology
Quarterly, 59, 375-383.

Sia, A., Hungerford, H. R. & A.N. Tommera (1985/86). Selected Predictors of
Responsible Environmental Behavior: An Analysis. Journal of Environmental
Education, 1 7(2), 31-40.

Sidanius, J. & F. Pratto (1993). The Inevitability of Oppression and the Dynamics of
Social Dominance. in P. Sniderman & P. Tetlock (Eds.), Prejudice, Politics, and
The American Dilemma (pp. 173-211). Stanford, CA: Stanford University Press.

Sivek, D. J. & H.R. Hungerford (1989). Predictors of Responsible Behavior in Members
of Three Wisconsin Conservation Organizations. Journal of Environmental
Education, 21(2), 35-40.

Slovic, P. (1987). Perceptions of Risk. Science, 236, 280-285.

Smith, D. G., Cragg, A. M. & G.F. Croker (1991). Water Quality Criteria for Bathing
Waters Based on User Perception. Journal of Environmental Management, 33,
285-299.

Smith, D. G. & R.J. Davies-Colley (1992). Perception of Water Clarity and Colour in
Terms of Sustainable Recreational Use. Journal of Environmental Management,
36, 225-235.

Smith-Sebasto, N. J. & R.W. Fortner (1994). The Environmental Action lntemal Control
Index. Journal of Environmental Education, 25(4), 23-29.

Smith-Sebasto, N. J. (1992). The Revised Perceived Environmental Control Measure: A
Review and Analysis. Journal of Environmental Education, 23(2), 24-33.

Sniderman, P. M., Brody, R. & J .H. Kuklinski (1991). The Principle-Policy Puzzle: The
Paradox of American Racial Attitudes. in P. M. Sniderman, R. Brody, & P. E.
Tetlock (Eds.), Reasoning and Choice (pp. 58-69). Cambridge: Cambridge
University Press.

Sniderman, P. M. & P.E. Tetlock (1986a). Interrelationship of Political Ideology and

Public Opinion. in M. Hermann (Ed), Handbook of Political Psychology (2nd
Ed., pp. 62-96). San Francisco: Jossey-Bass.

228

Sniderman, P. M. & P.E. Tetlock (1986b). Symbiotic Racism: Problems of Motive
Attribution. Journal of Social Issues, 42(2), 129-150.

Steel, B., S, Soden, D. L. & R.L. Warner (1990). The Impact of Knowledge and Values
on Perceptions of Environmental Risk to the Great Lakes. Society and Natural
Resources, 3, 331-348.

Steel, B. S. (1996). Thinking Globally and Acting locally? Environmental Attitudes,
Behavior and Activism. Journal of Environmental Management, 47, 27 - 36.

Steel, B. 8., List, P. & B. Shindler (1994). Conflicting Values About Federal Forests: A
Comparison of National and Oregon Publics. Society and Natural Resources, 7,
137-153.

Stern, P. C., Dietz, L., Kalof, L. & A. Guagnano (1995). Values, Beliefs, and
Proenvironmental Action: Attitude Formulation Towards Emergent Attitude
Objects. Journal of Applied Social Psychology, 25 , 1611-1636.

Stern, P. C. Dietz, T., & L. Kalof (1993). Value Orientations, Gender and Environmental
Concern. Environment and Behavior, 25(3), 322 - 348.

Sutton, S. (1998). Predicting and Explaining Intentions and Behavior: How Well Are We
Doing? Journal of Applied Social Psychology, 28(15), 1377-1338.

Syme, G. J. & M.D. Fenton (1993). Perceptions of Equity and Procedural Preferences for
Water Allocation. Society and Natural Resources, 6, 347-360.

Syme, G. J. & B.E. Nancarrow (1992). Predicting Public Involvement in Urban Water
Management and Planning. Environment and Behavior, 24(6), 738-758.

Syme, G. J. & B.E. Nancarrow (1996). Planning Attitudes, Lay Philosophies, and Water
Allocation: A Preliminary Analysis and Research Agenda. Water Resources
Research, 32(6), 1843-1850.

Syme, G. J. & B.E. Nancarrow (1997a). Defining and Utilizing Lay Ethics and Equity in
Land and Water Allocation. Paper presented at the at the Environmental Justice
Conference, University Of Melbourne.

Syme, G. J. & B.E. Nancarrow (1997b). The Determinants of Perceptions of Fairness in
the Allocation of Water to Multiple Uses. Water Resources Research, 33(9),
2143-2152.

Syme, G. J ., Nancarrow, B. E. & J .A. Mccreddin (1999). Defining the Components of

Fairness in the Allocation of Water to Environmental and Human Uses. Journal
of Environmental Management, 5 7, 5 1-70.

229

Syme, G. J ., & KD. Williams (19930). The Psychology of Drinking Water: An
Exploratory Study. Water Resources Research, 29, 4003-4010.

Tarrant, M. A., Bright, A. D. & H.K. Cordell (1997). Attitudes Towards Wildlife Species
Protection: Assessing Moderating and Mediating Effects in the Value Attitude
Relationship. Human Dimensions Of Wildlife, 2(2), 1-20.

Tenbrunsel, A. E., Wade-Benzoni, K. A., Messick, D. M. & M.H. Bazerman (1997). The
Dysfunctional Aspects of Environmental Standards. in M. H. Bazerman, D. M.
Messick, A. E. Tenbrunsel, & K A. Wade-Benzoni (Eds.), Environment, Ethics
and Behavior (pp. 105-121). San Francisco, CA: The New Lexington Press.

Tetlock, P. E. (1986). A Value Pluralism Model of Ideological Reasoning. Journal of
Personality and Social Psychology, 50, 819-827.

Thibaut, J. & L. Walker (1975). Procedural Justice: A Psychological Analysis. Hillsdale,
NJ: Erlbaum.

Thompson, M. (1992). The Dynamics of Cultural Theory and Their Implications for the ,
Enterprise Culture. in A. Ross (Ed.), Understanding the Enterprise Culture .
Edinburgh: Edinburgh University Press.

Thompson, M., Ellis, R. & A. Wildavsky (1990). Cultural Theory. Boulder, CO:
Westview.

Thompson, M. & P. James (1989). The Cultural Theory of Risk. in J. Brown (Ed),
Environmental Threats: Perception, Analysis and Management (pp. 138-150).
london: Belhaven.

Thompson, M. (1988). Socially Viable Ideas of Nature. in E. Baark & U. Svedin (Eds.),
Nature, Culture, Technology: Towards a New Conceptual Framework. London:
Macmillan.

Tyler, T. R. & E.A. Lind (1992). A Relational Model of Authority in Groups. in M. P.
Zanna (Ed.), Advances in Experimental Social Psychology (Vol. 25, Pp. 1 15-192).
Orlando: Academic Press.

Tyler, T. R. (1988). What is Procedural Justice? : Criteria Used by Citizens to Assess the
Fairness of Legal Procedures. American Bar Foundation working Paper 8725

UNESCO. (1991). Our Common Future. Paris: UNESCO.
US. Census 1990 (1991). United States Census Bureau. Washington, DC.

US. Census 2000 (2001). United States Census Bureau. Washington, DC.

230

United States Environmental Protection Agency (USEPA) (1996, May 16, 200]).
Eflluent Trading in Watersheds Policy Statement. Environmental Protection
Agency [February 1996, April 12, 2002].

United States Environmental Protection Agency (USEPA) (1997). The Watershed
Protection Approach. Environmental Protection Agency. Available:
www.cpa.gov/owow/watershed/focuS/ref.htrnl [1997, January 27,1997].

United States Environmental Protection Agency (USEPA). (2000, 03/ 1 6/2001). Storm
Water Phase II Final Rule. Office of Water. Available:
www.cpagov/owm/sw/phaseZ [January 2000, April 2001]

van den Putte, B. (1993). On the Theory of Reasoned Action. Unpublished Doctoral
Dissertation, University Of Amsterdam, Amsterdam.

Van Liere, K. D. & RE. Dunlap (1978). Moral Norms and Environmental Behavior: An
Application of Schwartz's Norm Activation Model to Yard Burning. Journal of
Applied Social Psychology, 8(2), 174-188.

Van Liere, K D. & R.E. Dunlap (1980). The Social Bases of Environmental Concern: A
Review of Hypotheses, Explanations and Empirical Evidence. Public Opinion
Quarterly, 44, 181-197.

Vaske, J. J. & M.P. Donnelly (1999). A Value - Attitude - Behavior Model Predicting
Wildland Preservation Voting Intentions. Society and Natural Resources, 12, 523-
537.

Walster, E., Walster, W. G. & E. Berscheid (1978). Equity Theory and Research. Boston,
MA: Allyn and Bacon Inc.

Wenz, P. S. (1988). Environmental Justice. Albany: State University of New York Press.

Widegren, O. (1998). The New Environmental Paradigm and Personal Norms.
Environment And Behavior, 30, 75-100.

Wildavsky, A. (1991). The Rise of Radical Egalitarianism. Washington, DC: American
University Press.

Woolrnington, E. & J .S. Burgess (1983). Hedonistic Water Use and Low Flow Runoff in
Australia’s National Capital. Urban Ecology, Vol. 7, 215-227.

Wynne, B. (1989). Frameworks for Rationality in Risk Management: Towards the

Testing of Native Sociology. in J. Brown (Ed), Environmental Threats:
Perceptions, Analysis, and Management (pp. 33-47). London: Belhaven.

231