-—;, ‘11 12911111111111

. V 903 01008 8015

mm His}. “' 'AR 3'"
Michigan State

University

 

This is to certify that the

dissertation entitled

Perceptions and Behavioral Responses of
Anglers to the Presence of Toxic Chemicals
in Michigan Rivers

presented by

Edward Udd

has been accepted towards fulfillment

of the requirements for
Park and
Ph. D degree in Recreation Resources

'
Major p essor

Date //'//’ f5-

ucn' Am_.__.._- _ A .r 1- .n
1

0-12771

PERCEPTIONS AND BEHAVIORAL RESPONSES OF ANGLERS TO THE
- PRESENCE OF TOXIC CHEMICALS IN MICHIGAN RIVERS

o By

Edward Udd

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Park and Recreation Resources

1985

ABSTRACT

PERCEPTIONS AND BEHAVIORAL RESPONSES OF ANGLERS TO THE
PRESENCE OF TOXIC CHEMICALS IN MICHIGAN RIVERS

BY

Edward Udd

Chemical substances which are known to be hazardous to
people are being released into the environment daily. The
living and working environments of many people are often
converted into areas of well documented health risks.
Recreation environments are not immune from toxic chemical
contamination. Anglers who consume fish caught from
contaminated waters may be endangering their health. The
Michigan Department of Natural Resources (MDNRJ has
recognized this hazard and has implemented an information
dissemination program targeted at anglers. The assumption
is that anglers will use the information to protect their
health. This expectation parallels the action of many
public agencies in their efforts to warn the public of the
dangers of other types of natural hazards. The primary
premise of such action is that the public will behave in a
predictable, rational manner when presented with correct
information about the hazard. Predictions for how people
will behave given correct information can be found in the
natural hazards literature. This study focuses upon the

applicability of those predictions to the recreation

UDD

environment contaminated by toxic chemicals. There were
three main findings in this study: 1) anglers did not
accurately perceive the level of toxic chemicals present in
the waters they were fishing and in the fish they were
catching. 2) anglers did not minimize their risk from toxic
chemicals by taking precautionary actions even when they did
perceive that toxic chemicals were present. 3) the
expectation that the public will make decisions that
parallel the economic rationality decision model was not
supported. The implications of the study are: 1) public
agencies need to explore new techniques for disseminating
information about toxic chemicals to anglers. 2) the
expectation that the public will respond in an economically
rational manner to correct information is misleading and
public agencies should not expect such results. 3) anglers
are being exposed to a health risk from contaminated fish

due, in part, to the ineffectiveness of the present

information dissemination program.

 

To Mary, with love.

ii

ACKNOWLEDGMENTS

As with all dissertations, this one was not completed
without considerable help from many peOple. My appreciation
for their thoughtfulness and consideration is deep and
sincere and I wish to take a moment to thank some of them at
this time.

First of all, I wish to express my appreciation to my
fellow graduate students both former and still trying. ‘No
one can understand the exasperation and frustration of the
graduate student process like a fellow graduate student.
Their empathy and support provided a helping hand whenever
it was needed. I would particularly like to thank the
following people: Dave Safronoff, Dave Feltus, Gene
Brothers, Tish Neumyer, Hideo Kikuchi, Bill Norman, Bob
Slana, Dave Klenosky and Marci Cobb. And a special thanks
to Cindy Deale who shared a good deal of my work and
efforts.

I would like to thank all the staff in the department
both former and present, and particularly Char, Betty,
Denise and Marcia. Their everyday kindness and helpfulness
made each day a little easier.

I would also like to thank the entire faculty as each

and every one of them has made the effort, at one time or

iii

another, to share some insights about a problem, to give a
word of encouragement or to simply share a word or two. I
would like to particularly mention Don Holecek, Dan Stynes,
Jim Bristor, Maureen McDonnough, Ted Haskell, Rick Paulsen,
Malverne Winborne and Lou Twardzik for their specific help,
attention and Opportunities presented.

I especially want to thank the members of my graduate
committee for their time and commitment in seeing this
dissertation through to its completion. I would like to
thank Dr. Martin Block for teaching me most of what I can
say I know about statistics and his help in putting together
a solid sampling design. I would like to thank Dr. Bob
Marty for reigniting my interest in learning when I did not
think I would ever get excited about taking a class again
and for his help in thinking through some major issues and
concerns.

I would also like to thank Dr. Dennis Propst for his
help in making this dissertation the best it can be and for
not settling for less than my best effort. His insightful
comments and suggestions have often led me to new heights of
travail and, consequently, to new plateaus of
accomplishment.

Lastly, I would like to thank Dr. Joe Fridgen who as
my committee chairman, advisor and friend has had the major
influence on my doctoral career. I cannot hope to repay all

the stress, frustration and persistence that went into the

iv

process of molding me into a doctoral candidate, but any
successes I have, small or large, will reflect Joe's faith
and belief in me.

This research project was funded in part by a grant
from the Center for Environmental Toxicology, Michigan
State University and the Agricultural Experiment Station,
Michigan State University, Project #1436-H, "Psychology of
Recreation and Tourism: The Relationship between the People

and the Resource," Joseph D. Fridgen, Principal Investigator.

 

TABLE OF CONTENTS

Title Page

List Of Tables 0.0.0.00.0.0...00.0.0...OOOOOCOOOOOCOOOO Vii
IntrOdUCtion OOOOOOOOOCOOOOOOOOOOOOOOIOOOOOCOOOOOOOOOOO 1-

Leisure 0......0......0.0.0.000...OOOOOOOOOOOOOOOOOOO 3

FiShing 00....0.00.0.0.0...0.000.000.0000.0.00.0.0... 10

Toxic Chemicals ..................................... 12
Natural Hazards ..................................... 18
Perception ....................................... 19
Decision Making .................................. 23
The Response Alternatives ........................ 26
Problem Statement ................................... 31
Hypotheses .......................................... 34
Methods ............................................... 40
Sampling ............................................ 40
Sample Size ...................................... 45
Limitations ...................................... 46
Credibility of Small Samples ..................... 46
Analysis ............................................ 48
Limitations of a Likert-type Scale ............... 49
Statistics .......................................... 50
Operationalization of the Hypotheses ................ 53
Results ............................................... 59
Reliability of Perception Scale ..................... 59
Control for Environmental Conditions ................ 60
Hypothesis 1 ........................................ 61
Hypothesis 2 ........................................ 64
Hypothesis 3 ........................................ 64
Hypothesis 4 ........................................ 66
Hypothesis 5 ........................................ 68
Survey POpulation Versus a Larger Scale Study ....... 69
Internal Profile of the Survey Population ........... 72
Description of Angler Behavior Towards
Toxic Chemicals ..................................... 73
Secondary Analysis .................................. 77
Discussion ............................................ 84
Perception of Toxic Chemicals ....................... 84
Behavioral Responses to Toxic Chemicals ............. 89
Mediating Variables ................................. 96
Conclusions .......................................... 102
Future Research Needs .............................. 104
Recommendations for Management
of the Study Rivers ................................ 108
Appendix A: Sampling Locations ....................... 112
Appendix B: Perception Scale Items ................... 113
Appendix C: Questionnaire ............................ 114
References ........................................... 125

vi

LIST OF TABLES

Title Page

Rating of the Beauty of the Surrounding
Land Across the Three Study Conditions ................ 62

Toxic Chemical Perception Scale Scores Across
Three Levels of Toxic Chemical Contamination .......... 63

Planned Consumption of Fish Versus Location ........... 65
Coping Behavior Versus Location ....................... 67

Accurate Versus Inaccurate Perception
Groups on Age, Education and Income ................... 70

Distance Anglers Travelled to
ReaCh StUdy Sites OOOCOOOOOOIOIOOOOOOOOO0.0.00.0.0.0... 74

Sex of Anglers Across Study Conditions ................ 74
Age of Anglers Across Study Conditions ................ 75
Race of Anglers Across Study Conditions ............... 75
Education Levels Across Study Conditions .............. 76
Annual Family Income Across Study Conditions .......... 76

Familiarity with Toxic Chemicals
Across StUdY conditions 00.0.0.0...OOOOOOOOOOOOOOOOOOOO 78

How Anglers Detect the Presence
0f TOXiC Chemicals OOOOOOOOOOOOOOOOOOOOOOOO0.0...00.... 78

Where Anglers Find Out About Toxic Chemicals .......... 79

Anglers Who Read MDNR Brochure
Across Study Conditions ............................... 79

Summary of Stepwise Multiple Regression

Analysis on Perception versus Age, Income,
Education and Miles from Home ......................... 83

vii

INTRODUCTION

The relationship between people and the environment is
dynamic. As the environment changes, people adapt. When
the environment is not suitable, people modify the
environment. Oftentimes these modifications are for the
better. Sometimes these modifications lead to unfortunate
side effects such as the destruction of habitats for
wildlife, air pollution, unrestored and exhausted strip
mines, acid rain or toxic wastes and spills. These impacts
are unfortunate in themselves, but they can also affect the
people who create the problem, as well as others.

This study looks at one relationship between peOple and
the environment: anglers and toxic chemicals. Anglers are
simply those people who fish for sport. Toxic chemicals are
those substances which have been defined by State and
Federal government agencies as presenting a public health
threat. The relationship is important because anglers may
be exposed to hazardous levels of toxic chemicals when they
consume contaminated fish.

Governor Blanchard in his State of Michigan message
(1985) recognized this problem: “u. Hundreds of chemicals
and heavy metals have been found in the Great Lakes, raising
concern about this little understood pollution problem.

Fish can concentrate these contaminants by many thousands of

times, enabling us to detect themJ' “L34).

Toxic Chemicals raise such levels of concern because
they are known to be a cause of cancer and of birth defects,
as well as lesser health problems such as nausea and severe
headaches (Burmaster, 1982; MDNR, 1984b; Collins, 1978;
Schweitzer, 1983; Holdgate, 1982; Eckholm, 1982; Logue
et al., 1981). The Michigan Department of Natural Resources
and the Michigan Public Health Department recognize the risk
of toxic chemicals to anglers and have jointly published a
warning about toxic chemicals in certain Michigan rivers as
well as the Great Lakes. This warning is printed on the
Michigan Fishing Guide (MDNR, 1984a) available to anglers
when they purchase a fishing license.

Limited research exists on the relationship between
anglers and toxic chemicals. In this study, findings from
the natural hazards literature are used to predict how
anglers will respond to the presence of toxic chemicals in
the waters that they are fishing. These predictions are
based on the assumption that anglers will respond to the
presence of toxic chemicals in the same way that people
respond to such natural hazards as: air and water pollution,
floods, earthquakes and tornados. Some differences exist
between the "typical" natural hazard situation and the
threat that toxic chemicals pose to anglers. These

differences will be discussed later in this chapter.

However, one difference is of enough concern to discuss
at this point; due to the inherent nature of toxic chemical
contamination, the chemicals are not detectable by any of
the human senses in the levels they are usually found in the
environment. For this reason it cannot be automatically
assumed that peOple can "perceive" the presence of the
hazard: toxic chemicals. Thus, this study also examines the
question of the perception of toxic chemicals by anglers as
well as their behavioral responses.

The remainder of this chapter will review the litera-
ture as it pertains to leisure and fishing experiences,
toxic chemicals in the environment, and natural hazards.
These literature bases establish the foundation for the
approach taken in this study, help articulate the problem
statement and provide the background for the hypotheses
which appear at the end of this chapter.

Leisure

Even though "leisure" time has existed almost from the
dawn of human creation, debate still continues today as to
how to define leisure. "Lurking in the shadows of most all
research into whatever aspect of leisure is the question:
What is leisure?" (Harper, 1981, p. 113). iLeisure can be
defined as a specific type of activity. By definition then,
such activities as boating, skiing, backpacking, walking

etc. as a group make up the term "leisure”.

However, such an approach has some difficulties. What
is leisure for one person may be work, or worse, for
another. For example, a couple may both be cross-country
skiing. For one the experience is totally enjoyable. The
other, however, may only be participating to please the
first or because the first forced the other to go. The
”experiences” of the two are totally different even though
they are participating in the exact same activity.

Another conceptualization of leisure, and the one used
in this study, is that leisure is an experience rather than
an activity or a type of behavior (Harper, 1981; Gunter,
1979). An experience, unlike an activity, has a cognitive
component in addition to the behavior itself. For example,
two individuals may be participating in the identical
activity, but the psychological outcome of that activity may
be entirely different for each of the individuals (Driver
and Brown, 1978). Using this definition, two individuals
who are fishing under identical conditions may be receiving
two entirely different types of leisure experiences as
mentioned above» One may feel that the fishing experience
is an excursion into very primitive natural conditions,
while the other may feel hemmed in by the signs of other
anglers and civilization. Both are participating in the
same activity, but each is receiving a unique leisure

experience.

In addition to leisure as an experience it is important

to keep in mind that leisure is voluntary in nature. ”Of

 

all the characteristics of the leisure experience, there is
probably none moree'obvious' than the assumed freedom to
choose the desired activity in the use of one's time”
(Gunter, 1979, p. 16).

This point is emphasized because anglers are partici-
pating in a leisure experience. The "freedom" character-
istic of leisure means they are relatively free to choose
where to fish, when to fish and even whether to fish. If
they are forced to fish they are not participating in a
leisure experience. For this study this simply means that
anglers are free to fish or £125 to fish in waters that
contain toxic chemicals. This freedom of choice, of
course, depends on anglers knowing where toxic chemicals
are to be found. If they cannot perceive the presence of
these chemicals then they cannot make knowledgeable choices
in attempting to avoid contaminated areas. The ability of
anglers to perceive the presence of toxic chemicals is of
central concern in this study.

A system which evolved from the concept of leisure as
an experience is the Recreation Opportunity Spectrum Land-
Use Classification system (ROS). Driver and Brown (1978)
describe how various types of settings lead to Opportunities
for specific types of recreation experiences. In their

study, settings range from primitive to urban in nature.

The basic tenet of the ROS system is that specific types of
settings are most appropriate for specific types of
recreation experiences. Thus, the environment, which
includes social and managerial factors as well as physical
factors, plays some role in what peOple will experience from
their leisure activities.

Clark and Stankey (1979) used the ROS system to
describe how Lake Kachess in Washington evolved over time
from an area initially offering primitive leisure
experiences to an area offering urban leisure experiences.
The reason for this evolution was due to managers responding
to changing use patterns at the lake. For example,
initially a wagon road was built to provide access to the
lakes The improved access led to more people using the
area. Improvements were made to accommodate this increase
in use which, in turn, led to greater use and increased
development. What was once an area suitable for primitive
use became an area suitable for urban-type use. As the
environment changed so did the type of use.

The sensitivity of peOple to changes in their leisure
environments gives impetus to this study. Toxic chemicals
are not normally found in hazardous amounts in undisturbed
environments. The presence of toxic chemicals might lead
to changes in recreation experiences in the affected

environment.

Hazards in the recreation environment present a unique
managerial problem. Some recreation activities involve
certain risks that are part of the recreation experience
(Cheron, et al., 1982). One way of removing all risk from
the sport of, say, rock climbing would involve the removal
of all vertical lepes and all rocks. Without question,
such actions would dramatically change the rock climbing
experience. The regulation of recreationists, even for
their own protection, goes against the voluntary nature of
leisure and recreation. "Recreation and visitor regulations
are inherently contradictory. Recreation is a voluntary,
pleasurable, rewarding activity, based on free choice, while
regulations are designed to restrict free choices" (Lucas,
1982, p. 148). .A balance needs to be struck between freedom
and regulations created to protect the safety of
recreationists.

When a hazard is readily apparent and avoidable by the
type of recreationists who use an area, regulation need
only be minimal. However, ”u. When a real hazard is
perceived as being minimal or nonexistent, a much more
intensive management program is required? (Jubenville,

1978, p. 150). In other words, recreation managers need to
regulate recreationists when recreationists cannot perceive
real hazards that are present. As already mentioned,
peOple rely on their senses to detect environmental

hazards, but their senses cannot directly perceive the

presence of toxic chemicals. This presents a situation
where some regulation by recreation managers is desirable.
From a legal standpoint, a hazard such as toxic
chemical contamination constitutes a hidden danger. The
public cannot be expected to know of the presence of toxic
chemicals without some notice or warning from the managing
agency. The minimum legal duty of a recreation manager is
to provide the public with adequate warnings about the
presence of hazards that they would normally not be expected
to know about (Marty, 1982, pu94). Failure to do so may
lead to the recreation agency being liable for any
subsequent injuries to anglers who fished on their lands.
When choosing to regulate, recreation managers need to
determine the degree of intervention, from both legal and
practical standpoints, necessary to alert the public. Lucas
(1982) states that this leads to a choice between "soft” and
"hard” management. Soft management would choose the type of
regulation that has the least impact on the recreation
experiences of the public being regulated. One example of
this type of management would be the publication of a
warning on a pamphlet that would receive wide distribution.
This is similar to the approach taken by the Michigan
Department of Natural Resources with its publication of
warnings about toxic chemicals in its Michigan Fishing Guide

(MDNR, 1984a).

 

While it may not be legally mandated, a "hard"
management approach may be desirable when people continue to
behave in a manner which endangers their safety. .A hard
management approach might include removing the hazard
completely. At this point in time, that would entail the
removal of all contaminated fish from contaminated waters.
Such a solution, besides being prohibitively expensive, also
removes a key element from the fishing experience: catching
fish. ILess extreme measures might include: blocking access
to hazardous sites, creating fines for hazardous acts,
increasing the patrolling of a hazardous area, increasing
the visibility of a hazard using high density signing
methods or closing down the hazardous site(s). In any case,
the ultimate goal of the manager is to reduce the risk from
the hazard, but each has some impact on the fishing
experience.

In summary, leisure is a voluntary experience that is
influenced by the recreation environment. When known
hazards are present in these environments that are difficult
or impossible for recreationists to detect, recreation
managers may wish to intercede on the behalf of visitors.
This action must balance the desirability of preserving the

freedom of the visitor, maintaining the essence of the

leisure experience and protecting the safety of the visitor.

10

Fishing

Fishing is a leisure experience that is at once simple
and complex. (An angler can be defined as anyone who
participates in the act of fishing for non-profit motives.
This definition captures the simple essence of fishing.

However, in line with the definition of leisure as an
experience, in this study fishing is considered an
experience also. In this study, an angler is someone who
uses fishing as a vehicle to obtain a variety of recreation
experiences. This is consistent with the definition of
leisure given above where a recreation activity is not
equated with a single recreation experience.

Bryan (1979) describes four types of anglers:

1. Occasional - ”those who fish infrequently because
they are new to the activity and have not
established it as a regular part of their leisure,
or because it simply has not become a major
interest."

2. Generalist - anglers ”u. who have established the
sport as a regular leisure activity and use a
variety of techniques.”

3. Technique specialists - "anglers who specialize in
a particular method, largely to the exclusion of
other techniques."

4. Technique-setting specialists - ”highly committed
anglers who specialize in method and have distinct
preferences for specific water types on which to
practice the activity." (p. 33)

These anglers differ from one another in the type of
equipment they use, their desire to catch large numbers of

fish, the type of fish they are seeking and the type of

environment in which they are willing to fish.

 

11

Manfredo et a1. (1978) used cluster analysis to
identify six types of wilderness anglers based on the
psychological outcomes they derived from their experiences.
For example, the members of the type five group were
characterized by their emphasis on experiencing nature and
the change of pace that fishing afforded them (p. 293).

The other groups differed in the amount of emphasis they
placed on various psychological outcomes.

It is clear that there is no single, stereotypical
angler. Generally speaking, however, anglers are united by
the fact that they go fishing to attempt to catch fish.
Success in catching those fish is not essential to the
overall enjoyment of the fishing experience. "It would be
incorrect to suggest that, in the long run, the size and
number of fish caught are not important to fishing
enjoyment. In many instances, however, other factors
significantly influence fishing enjoyment" (Moeller and
Engelken, 1972, p. 1254). ”It seems apparent that the
nature of the fishing experience goes far beyond the actual
taking of fish.uf (Hendee et a1, 1977,EL4).

Driver and Knopf (1976) suggest that "...It is
unlikely that many fishermen would fish waters in which
they knew no fish existed. However, many fishermen have
had beautiful trips without having caught a fish, and
perhaps without having wet a line. Therefore, the satis-

faction derived from fishing is dependent on the existence

12

of fish, but goes considerably beyond the actual taking of
the fish" (p. 21).

In the study by Hendee et al. (1977), which examined
use of alpine lakes in the state of Washington, about forty
percent of the anglers studied did not catch any fish at
all. In another study, across a 135 day fishing season in
South Carolina on a stream stocked with fish, 30 percent of
the groups sampled caught no fish at all (James et al.,
1971).

In the Moeller and Engelken (1972) study, the single
most important factor that influenced the fishing experience
was water quality. Out of eight factors rated, the size of
fish and number of fish rated fourth and sixth respectively.

In summary, anglers can be characterized as being
composed of many types who seek a variety of recreation
experiences from fishing. These anglers are united by the
goal of catching fish, but success at this task is not
always necessary for enjoying the activity. Other charac-
teristics, such as water quality, may play an equally
important role in leading to a satisfactory fishing
experience.

Toxic Chemicals

 

The following three quotes give an indication of the

scope of the toxic chemical problem:

 

and controlling the thousands of chemicals and toxic
waste sites that may harm human health'(BJanchard,
1985).

”In 1980 a special U.S. government committee on toxic
substances characterized the broad challenge: the

staggering in view of the number of substances whose
risk should be evaluated, the rate of growth in both
number and volume of chemicals, the various routes by
which humans and the environment are exposed, possible
synergistic or combined effects of the substances, and
the effects that they cause - acute and chronic,
immediate and delayed” (Eckholm, 1982, p. 108).

"Recent concerns in the United States, Canada, and
Europe over the increasing number of environmental
chemicals that are believed to pose serious problems
for human health have led to a flurry of national laws
and regulations for the control of chemicals. These
laws and regulations are designed to address ecolo-
gical concerns as well, but actual and potential
health hazards have received by far the greatest
attention in recent years" (Schweitzer, 1983, pu22).
Toxic chemicals are not a problem peculiar to the
state of Michigan. They are a problem across the country
and around the world.

It is easy to make a strong case that toxic chemicals
are hazardous to human health. As already mentioned in the
introduction toxic chemicals are known to cause nausea and
severe headaches as well as cancer and birth defects (MDNR,
1984b; Collins, 1978; Schweitzer, 1983; Holdgate, 1982;
Eckholm, 1982; Logue et al., 1981). What remains a problem

is determining what level of toxic chemicals is acceptable

in the environment.

14

This is a particular problem with water because of the
many variables involved in measuring water quality (Ott,
1978). What is perfectly acceptable under one set of
conditions is unacceptable under another set of conditions.
In addition, there is presently no national standard for
what constitutes a hazard for most toxic chemicals (Ott,
1978). "The task of maintaining water quality".u.is.n
"complicated by increasing amounts of evidence that minute
doses of organic substances could injure human health.
Refinements in detecting minute quantities of
pollutants"... have... "helped uncover the evidence"
(Holdgate, 1982, ;»163). The levels of toxic chemicals that
experts suggest are needed to create a hazard continues to
decrease. As our measuring techniques continue to improve,
these low levels of toxic chemicals are being detected in
the environment.

An illustration of the problem of toxic chemicals can
be drawn by looking at Polychlorinated Biphenyls (PCBs).
PCBs have the ”u. properties of nonflammability,
stability, resistance to acids, alkalis and other caustic
chemicals and low volatility under prolonged heating".
(Hesse, 1976, gx127). Some of the uses of PCBs were in
”u. dielectric fluids in capacitors, plasticizer appli-
cations, transformer fluids, hydraulic fluids and

lubricants, and heat transfer fluids” (Hesse, 1976, px127).

15

The very prOperties that made PCBs so attractive to
industrial use are the prOperties that make PCBs such a
problem when they are spilled into the environment. The
chemical structure of PCBs is essentially the double
bonding of two phenol rings with several sites available
for bonding with chlorine atoms. This chemical arrangement
is very stable and resistant to any breakdown.

In addition, there are 209 possible chlorobiphenyl
isomers (Mieure, 1976). This simply means that there is not
just one chemical structure for PCBs, but 209 possibilities.
To counteract the presence of PCBs in the environment using
other chemicals to neutralize their hazardous effects is
very close to impossible. If it could be done, a total of
209 antidotes would have to be created, each with their
known and unknown side effects. There is no proven
technology for the removal of PCBs from water and, hence,
the emphasis is on preventing PCBs from entering waste water
(Kopp, 1976).

Because of possible health effects from PCBs, stricter
monitoring of these chemicals is in effect. For example,
back in 1976, it was noted that ”unbecause of the potential
effects of PCBs on fish and consumers of fish, the Great
Lakes Fishing Laboratory in 1977 will incorporate analysis
of PCBs with its routine program of pesticide analysis in

Lake Michigan fishes" (Wilford, 1976, p. 177).

16

Polychlorinated Biphenyls present a threat to
consumers of fish because PCBs "bioaccumulate" (Train,
1976). Bioaccumulation means that some chemical may exist
in trace or even undetectable amounts in water, but fish
living in that water build up high levels of the chemical
in their body tissues over time. Water that is safe to
drink may contain fish that are hazardous to eat because of
the bioaccumulation of toxic chemicals.

In short, toxic chemicals present a hazard in the
environment. The extent of that hazard is unclear, but it
appears that the risk is increasing and is present in
waters around the world.

The purpose of this paper, however, is not to examine
the health effects that can be attributed to toxic chemicals
nor the difficulty in detecting their presence and
establishing that a hazard exists: rather, the purpose is
to examine the relationship between anglers and toxic
chemicals. This relationship can be examined only by being
able to define the level of toxic chemicals present in the
waters and fish being studied. Without such a definition,
it is impossible to talk about anglersd.responses and
perceptions because the hazard is not accurately described.

The definition of the level of toxic chemicals present
in Michigan waters is made possible because the Michigan
Department of Natural Resources, in COOperation with the

Michigan Public Health Department, has identified which

17

waters in Michigan contaminated by toxic chemicals
constitute a threat and which do not:

"Persons frequently eating fish caught in Michigan

waters should be aware of the potential risk of

exposure to chemical contaminants which may be present
in some species caught in some locations.

Contaminants of concern include Mercury, PCB, PBB,

DDT, Dieldrin, Chlordane, Toxaphene and Dioxin. One

or more of these chemicals have been detected in fish

and serve as the basis for the advisory for various

bodies of water" (MDNR, 1984a, p.7).

The warning goes on to explain which waters contain toxic
chemicals in sufficient amounts to make the consumption of
any fish unsafe; which waters have certain species of fish
which are unsafe if consumed in large amounts or are unsafe
if consumed by high risk people such as expectant women or
children under six; and which waters can be assumed to be
safe because they receive no mention. The specific rivers
used in this study will be discussed in Chapter Two.

In summary, toxic chemicals pose a threat not only in
Michigan, but throughout the world. The threat occurs
because toxic chemicals endanger human health and those
chemicals are present in the environment. .Anglers
represent a group at risk because toxic chemicals are found
in the fish they catch and potentially consume. The state
of Michigan has recognized this problem and has identified
those waters which are potentially hazardous due to toxic

chemical contamination. The waters identified in this

warning will be used to define the relative levels of toxic

18

Chemical contamination in this study.
Natural Hazards

A natural hazard can be defined "."as an interaction
of people and nature governed by the coexistent state of
adjustment in the human use system and the state of nature
in the natural events systenflmu. "Certain of the hazards
are created by man through his alteration of land and water
or by his invasion of risky areas; others are exacerbated
by his efforts to reduce the risk" (White, 1974, p. 4).
Stated differently, people adjust to disturbances in the
natural environment. These disturbances may have natural
causes such as earthquakes, tornados or floods. They may
also be due to peOples' disturbance of the environment such
as with air and water pollution. Or, these disturbances
may be due to peOple choosing to be in environments that
are potentially hazardous such as building a house on a
fifty year floodplain or recreating in an area known to
contain toxic chemicals.

Early investigations of natural hazards revolved
around the question of why people persisted in occupying
floodplains (White, 1974). The driving force behind this
early research was that the federal government had invested
over $5 billion in flood control efforts starting in 1936

and ending in 1957 only to find a net increase in flood

losses during the same time period (White, 1974). Early

 

19

research efforts soon branched out from simply studying
floods to other types of natural hazards as well.

The accumulation of research led to Hewitt and
Burton's (1971, p.140) description of four possible beha-
vioral responses to a natural hazard:

. PeOple avoid the natural hazard.

. People use mechanical circumvention to deal with
the hazard (i.e. the use of a boat to escape from
a flood).

3. People coge with the hazard by changing whatever

is causing the damage.
4. People adapt to the hazard by accepting what is
happening due to the hazardous event.

1
2

These four responses form the basis of much of the
behavioral prediction research in natural hazards. These
four behavioral responses also form the basis for the
predictions made in the hypotheses of this study. However,
as a precursor to these behavioral responses, it is first
necessary that pe0ple perceive the presence of the hazard.

Perception

 

At this point it is important to discuss perception as
a key concept before the presentation of the research
paradigm for this study. With the study of many natural
hazards this is an insignificant issue. For example, in the
study of a tornado it is not necessary to first ask if the
victims of the event noticed the tornado. When a tornado is
present it is obvious and leaves little doubt that it is

present.

20

However, because toxic chemicals are not directly
perceivable it is first necessary to establish if people
can ”perceive" their presence. "It seems likely that the
process of choice does not begin until after a first
threshold of awareness of actual or anticipated loss is
reached" (Burton et al., 1978, p.89). If the threat does
not exist in the mind of the angler then questions of
behavioral responses cannot be made. This makes intuitive
as well as logistical sense. If something is not
perceivable then it does not exist unless knowledge of its
presence is obtained from some indirect source.

Perception as a field of study is well established and
a vast body of literature exists. The oldest school of
thought deals with perception as strictly a physiological
process (Kolers, 1983). This process was characterized by
the reception of real, physical stimuli and a physiological
reaction by the recipient to that physical stimuli. This
conceptualization led to the study of perception that did
not involve any cognitive processing (Kolers, 1983).

More recently, a second school of thought has
developed that has dealt "u.with perceptual experiences
less realistic than those of the traditional topics”
(Kolers, 1983, p. 130). In essence, this school of thought
deals with perceptual processes that "distort" reality.
For example, when a visual stimulus is of a very short

duration, say thirty to sixty milliseconds, our ability to

21

distinguish temporal differences is impaired (Haber and
Hershenson, 1973, pp. 134-138). Visual stimuli of less

than this duration still exist in reality, but are

generally beyond our ability to determine the duration of
their existence. This principal is what makes moving
pictures work. The frames of the movie are moved faster
than our ability to detect each individual frame of the
movie. What we see is different than what is really present
and can be distorted in many other ways as well.

A third school of thought deals with perception as an
image that peOple carry with them (Kolers, 1983). This
point of view leads to the acceptance that some sort of
cognitive process occurs after the reception of physical
stimuli which leads to a reconstruction of what the
perceiver actually imagines the stimuli to be. In effect,
"u. all our modes of experience, perceptual, affective and
normative, are brought into action at the same time" (Levy-
Leboyer, 1982,;L 46).

This notion leads to an even more complex and difficult
conceptualization of perception. Not only is it important
to study the physical reception of physical reality it is
also important to allow for distortions in that reception
and to allow for cognitive processing of the stimuli which
may lead to an image that is quite different from reality.

These differences are due to the evaluation process by the

22

perceiver of the stimuli as a result of the cognitive
processing.

Given this complexity studying perception is very
difficult. "The only legitimate approach to it must be
that of focusing the spotlight on one of the phases of the
perceptual-evaluation process, without forgetting the
modulating effects of other aspects” (Levy-Leboyer, 1982).

In this study the focus of study for perception deals
with the final product of the entire perceptual process:
what is the level of toxic chemicals that anglers perceive
in the fish and in the water? Taking this approach does
not mean that the whole series of cognitive processes that
led to these perceptions is unimportant, rather, studying
the final perceptual image is the necessary point of study
to determine whether anglers can perceive the alternative
choices available to them.

At this point one can begin to examine the rela-
tionship between anglers and toxic chemicals. Anglers do
not directly perceive the presence of toxic chemicals
simply because there is no directly perceivable physical
stimulus. However, they may carry an image of the level of
toxic chemicals in the environment as a result of indirect
stimuli, such as the media, friends and acquaintances,
official publications and other sources.

In addition, they may also attribute the presence of

toxic chemicals in the environment to other stimuli that

23

they can directly perceive. Barker (1970) found such a
relationship while examining water pollution of beaches in
the Toronto area. The pollutants in the water could not be
directly perceived, but peOple attributed the level of
pollution to the presence of stimuli that they ggulg
perceive such as algae or floating debris.

If one chooses water conditions that are roughly
equivalent in sensory characteristics it would be expected
that these characteristics would not aid individuals in
their evaluation of the presence or non-presence of
pollutants. In this study, rivers which were roughly
equivalent in sensory stimuli were chosen as study sites.
Toxic chemicals further parallel the water pollution
example by not being directly perceivable.

Decision Making

 

The decision making process does not begin until the
available alternatives pass a perceptual threshold and
become perceived alternatives (Burton et a1, 1978). In this
study, once the hazard is perceived as present then anglers
have a whole array of alternatives available to choose from
to respond to the hazard.

Before those alternatives are presented, it is
necessary to examine the decision making frameworks of
anglers. The agency needs to understand, at least in a

rudimentary fashion, the decision making frameworks of the

24

public if they hope to influence their behavior. It is also
necessary to examine the decision making framework of the
managing agency since public agencies often try to influence
the behavior of the public.

The two main decision making frameworks of concern here
deal with economic rationality and bounded rationality.
Economic rationality, as might be expected, conceptualizes
decision making as a benefit/cost analysis by individuals
(Einhorn and Hogarth, 1981). Bounded rationality can be
conceptualized as meaning that each individual uses his/her
own individualistic decision making model that may or may
not lead to an optimal decision (Einhorn and Hogarth, 1981).

Like perception, decision making involves extensive
cognitive processing that is complex. And again, it is not
the focus of this study to attempt to explain all these
complexities as they relate to toxic chemicals and anglers.
The key issue is that the two decision making models lay the
groundwork for how public agencies expect the public to
behave and provides an alternative explanation of how the
public behaves in decision making situations.

In the case of natural hazards, such as toxic
chemicals, it has been shown repeatedly that public
agencies assume that the public utilizes the economic
rationality model in decision making (Slovic et a1, 1974).
The economic rationality model assumes that an individual

will be aware of all costs and benefits associated with all

25

the available alternatives. It is then assumed that the
decision maker will choose the alternative that will lead
to the highest benefit/cost ratio (Payne, 1982).

There is considerable debate whether bounded
rationality is really just an extension of the economic
rationality model or a separate model in its own right
(Einhorn and Hogarth, 1981). This paper differentiates
between the two model types because of the ample evidence
that peOple generally are not capable of assessing
benefit/cost ratios accurately because of less than perfect
information (Payne, 1982, Slovic et al, 1974). The bounded
rationality model predicts that individuals will make
decisions that they feel benefit themselves the most, but
that these decisions may not be the optimal decisions
predicted by the economic rationality model.

The general situation with natural hazards is that
public agencies attempt to protect the public using the
assumption that the public makes use of the economic
rationality model (Slovic et a1, 1974). Using this sort of
decision framework, public agencies expect the public to
respond as if it had perfect information and in a way to
best protect their health. If the public does participate
in a manner paralleling economic rationality then specific
behaviors can be anticipated such as those described above

for responses to natural hazards. The bounded rationality

26

model suggests that such rational, predictable behavior
cannot be anticipated and expecting the public to behave in
such a manner is in itself not rational.

Such a scenario has been tested extensively in the
natural hazards literature and a discussion of the relevant
literature follows the presentation of each hypothesis. In
short, while decision making is complex and involves many
cognitive processes this study focuses on the expectation of
specific behavioral responses that result from adopting the
economic rationality model. The bounded rationality model
is presented as an alternative model for explaining
behavioral responses.

The Response Alternatives

The above information provides a skeleton for creating
a "decision tree” for the response to toxic chemicals by
anglers. A decision tree simply represents the choices
available to a person confronted with a specific decision.
It is generally more detailed than the one presented here
because it tries to represent all the cognitive processes
that are involved before the actual alternatives are
confronted. In this study, the focus of attention is on the
behavioral responses of anglers to toxic chemicals and not
on the cognitive processes of decision making. Hence, those
processes are not depicted.

Figure 1 represents a possible decision tree for

anglers confronted with toxic chemicals. This decision tree

27

Figure 1

Decision Tree for Anglers Confronted by

Toxic Chemicals

 

 

Assessment of the hazard

 

 

/

Hazard perceived as being
present

 

l f \

 

 

Avoid fishing at \1

that location

 

\

 

v

Hazard not perceived

(as being present

 

1

 

Continue to fish as
usual

 

I

 

 

f

Continue to consume,
prepare and cook
fish as usual

 

 

or c0ping

 

Participate in on—site avoidance

 

l

 

 

Consume fish caught

 

 

A X

 

Avoid.consuming
fish caught

 

 

 

 

Minimize risk by
using alternate
preparation
procedures

 

 

l

 

 

 

Minimize risk Minimize risk by
by using using alternate
alternate cooking procedures
preparation

 

Minimize risk by
using alternate
cooking procedures

 

 

procedures

 

 

 

 

 

28

is modeled using the four behavioral responses provided by
Hewitt and Burton (1970). The model also recognizes the
importance of perception in the study of toxic chemicals and
uses part of the decision tree model described by Svenson
and Fischhoff (1985).

Figure 1 starts at the point where the initial
decision about the presence of toxic chemicals is being
made. This ignores all the information gathering processes
that went into the formation of this perception. This is
not to downplay the importance of those processes, rather
it is necessary to focus in on specific points to create a
study that is researchable.

It should also be noted that decision making theory
only plays a role, in this study, insofar as creating
expectations within a public agency for specific behavioral
responses by the public. In this case, the expectation of
the agency is that the public will take precautions to
protect themselves once they perceive the presence of toxic
chemicals.

If toxic chemicals are not perceived as being present
it would be expected that there would be no change in
behavior. In this study that would be represented by
peOple fishing as usual. If toxic chemicals ege perceived
as being present then peOple have the option of responding

or not responding.

29

Anglers have four response choices to toxic chemicals:
avoidance, mechanical circumvention, c0ping or adaptation.
Mechanical circumvention is not included in the model as
such circumvention is not possible with toxic chemicals at
this time.

Anglers can avoid the threat from toxic chemicals in
fish by not fishing at contaminated locations, by not
consuming the fish they catch or by reducing their
consumption of contaminated fish. Anglers can COpe with
toxic chemicals in fish by modifying their preparatory and
cooking behaviors to minimize the amount of toxic chemicals
they consume. Lastly, anglers can choose to continue to
fish as usual.

Figure 1 represents the decision tree of anglers
without any mention of optimality. There is no inherently
"right" path of choice. The problem for this study arises
because a government agency (Michigan Department of Natural
Resources) is attempting to change the behavior of anglers
who are confronted by a hazard (toxic chemicals). The
preferred path, as defined by the MDNR (1984a), is that
anglers modify their fish consuming behavior when fishing on
waters contaminated by toxic chemicals.

This directly parallels the initial impetus for all

natural hazards research: the government creating programs

to reduce the risk from natural hazards by attempting to

 

3O

modify the behavior of the public at risk (White, 1974).
It has been found repeatedly that the desired modifications
in behavior may not occur.

This discrepancy is discussed at length by Slovic et
al. (1974). The discussion revolves around the differences
in the decision making environments in response to risk
between managers and the managed public. The first
environment (management) can be described as often
following the tenets of the maximization of expected
utility. The second environment (public) can be described
as following the decision making framework of bounded
rationality.

The points of contrast to be considered are those
dealing with rationality and probability. The economic
rationality model "u. assumes that the rational decision
maker wishes to select an action that is logically
consistent with his basic preferences for outcomes and his
feelings about the likelihoods of the events upon which
those outcomes depend" (Slovic et al., 1974, p. 188). "The
theory of bounded rationality, on the other hand, postulates
that decision makers do not think probabilistically and that
they try to avoid the necessity of facing uncertainty
directly. Likewise they avoid the problems of evaluating
utilities and comparing incommensurable features. The goal

of the decision maker is assumed to be the achievement of a

31

satisfactory, rather than a maximum, outcome” (Slovic et
al., 1974, p. 190).

In short, the expected utility model assumes that the
public will make decisions in a rational fashion that will
lead to an action that represents the highest probability
of maximizing personal utility. Bounded rationality
suggests that peOple do not always act in a rational
fashion and that they have difficulty in accurately deter-
mining the real probabilities of an event.

The managerial actions of the MDNR mimic the expected
utility model. The public is provided with information
about the presence of toxic chemicals and the appropriate
behavioral responses. Given this correct information it is
assumed that anglers will behave in a manner to reduce
their risk from toxic chemicals.

In summary, when confronted with the presence of toxic
chemicals anglers have several response choices. The MDNR
is trying to guide those choices in a direction that
minimizes the risk of toxic chemicals to anglers. Two
decision making models, that of expected utility and
bounded rationality, are presented as representing the
decision making processes of the MDNR and the public.

Problem Statement

 

As with all environments, recreation environments are
not immune from common problems of civilization such as: air

pollution, acid rain, radioactive materials or toxic

32

chemicals. As with other hazards created by humans, when
these hazards present a real danger to recreationists
precautionary measures must be taken. ‘When a hazard, such
as a short metal stake hidden by tall grass, is present, it
is the duty of the managerial agency to warn the public from
this hidden hazard or risk being held liable for any
accidents. In a similar way, when a hazard, such as toxic
chemicals is in the recreation environment, it is the duty
of the managerial agency to warn the public or risk being
held liable for any accidents.

It has been found that managing agencies, in general,
attempt to protect the public from natural hazards by
adopting economic rationality as the model for expected
public decision making behavior (Slovic et al, 1974).
Possible problems result when the decision making
environments of managing agencies do not match the decision
making environments of the general public (Svenson and
Fischoff, 1985). A decision tree (see Figure 1) was created
to help illustrate the possible paths of decision based on
the perception that toxic chemicals exist in the
environment.

Toxic chemicals in the environment represent a problem
similar to the presence of other natural hazards in the

environment. Because of this similarity, the natural

hazards literature is used to provide predictions as to how

 

33

anglers would be expected to behave in the presence of
toxic chemicals. As with a great deal of natural hazards
literature, attempts to protect the angler in Michigan from
the dangers of toxic chemicals rest on the assumption that
the angler uses the economic rationality model in his/her
decision making process. This assumption leads to
providing the public with correct information with the
expectation that this information will lead to Optimal
decisions in minimizing the risk from toxic chemicals.

This study first looks at the question of whether
anglers can differentiate between sites containing
different levels of toxic chemicals. The ability of the
public to perceive these differences is essential to the
initial choice process. If no alternatives are perceived,
then there is no choice. The level of toxic chemicals may
vary, but if anglers cannot perceive any differences then
they do not have any alternatives to choose from and they
cannot respond to the presence of a stimulus with which
they are not aware.

Of those anglers who do perceive the presence of toxic
chemicals, the desired behavioral response is that they
will choose the alternative that will best protect their
health. This is similar to the expectation that once a
person is informed that a stake is hidden in the grass that
they will avoid injuring themselves on that stake. This

study examines the behavioral responses of anglers to the

34

perceived presence of toxic chemicals in their fishing
environment.

To accomplish this purpose, the behavioral predictions
made in the natural hazards literature are tapped. Since
the relationship between toxic chemicals and anglers has
not been previously studied in the natural hazards
literature, the hypotheses for this study have assumed that
anglers will behave in a manner consistent with the
economic rationality decision making model. Being aware of
the presence of toxic chemicals should lead to behaviors
which will minimize the risk from those toxic chemicals.

In essence, this study examines whether the natural
hazards literature provides a theoretical structure for
examining toxic chemicals as a hazard in recreation
environments. The hypotheses for this study follow.

Hypotheses

 

Hypothesis 1

 

Anglers will perceive real differences in the level of
toxic chemicals present in their fishing environment when
real differences exist.

Before anglers can respond to the presence of a hazard
they need to be able to perceive its presence. Three
different conditions of toxic chemical contamination were

used in this study to examine this issue. This hypothesis

represents the desired outcome of an information program

 

35

developed by the Michigan Department of Natural Resources.

The study of these initial perceptions is not unusual
in the natural hazards literature. For example, Mitchell
(1974) found "... in a series of studies of air pollution
in British cities, it was concluded that many types of
pollutants are not easily detected by the human senses
until they reach chronic proportions“ (p. 329). Similarly,
Barker (1970) found that for beach users that "In response
to the question.VHow can you tell if the water is
polluted?‘ more than half (56%) evaluated water quality on
the basis of appearance u. another 15% thought that
odour was significant u. Only a few beach users felt
that pollution could not be seen while 23% of the cottagers
stated that pollution was invisible and that impurities
could be identified only by scientific analysis" (p. 40).

Accurate perceptions about the presence of toxic
chemicals are needed before the economic rationality model
can be said to be working. This hypothesis simply states
that anglers will have access to the information about the
real levels of toxic chemicals present in the waters they
are fishing.

Hypothesis 2

For those who perceive the presence of toxic chemicals,
as the real level of toxic chemical contamination increases
the number of anglers who consume the fish they catch will

decrease.

36

When confronted with the presence of toxic chemicals it
would be expected that anglers would not consume them. The
best scientific evidence available shows that the
consumption of such chemicals is dangerous. A decision
maker who understands this would not be expected to consume
fish containing hazardous levels of toxic chemicals.

In similar instances where poisonous substances were
present in animals to be consumed considerable avoidance
behavior occurred. The following two economic studies
deal directly with avoidance of game contaminated by toxic
chemicals. In the first, Shulstad and Stoevener (1978)
examined the avoidance cost of mercury contamination in
pheasant in Oregon. They determined that the net loss to
the state was $1.35 million. They concluded "... it is
especially important that, through continued research,
information flows be improved so that the decisions of both
private and public decisionmakers are as well-based as
possible" (p.48).

Swartz and Strand (1981) studied the avoidance cost of
imperfect information. In this study oysters from the James
River were contaminated by kepones and the harvest of those
oysters was prohibited. The news of the contamination
deflated the market for ell oysters in the area even though

the oysters that reached the market place were from

uncontaminated sources.

 

37

The importance of these studies is that toxic chemicals
do lead to avoidance behavior by the public. Avoidance is
one of the four possible behavioral responses outlined
earlier by Hewitt and Burton (1971). It also suggests that
peOple are responding in a manner consistent with
protecting their health.

Hypothesis 3

 

For those who perceive the presence of toxic chemicals,
as the real level of toxic chemicals increases so will the
number of anglers who participate in COping behaviors.

Again, the direction of this hypothesis is predicted by
the desired results of the MDNR information program.

Coping behaviors form another one of the categories
mentioned by Hewitt and Burton (1971). It is different
from avoidance behavior in that peOple participate in a
hazardous event or activity, but they take active measures
to minimize their risks. For example, a person may choose
to live on a floodplain but builds their house on stilts to
minimize the damaging effects of floods.

Sims and Baumann (1983), in a review of the literature
found little evidence to support a causal relationship ”u.
between education and awareness and between awareness and
behavior..." (p. 167). However, the authors point out that
there are exceptions to their findings. They conclude ”The
question then is, What influences an individual to adopt a

measure that will reduce his or her vulnerability? The

38

answer is that little is known; past research is suggestive
and offers at best, some general directions or guidelines"
(p. 184).

Hypothesis 4

 

Anglers who have had negative experiences with
consuming fish contaminated by toxic chemicals in the past
will participate in COping behaviors at a higher rate than
anglers who have not had such experiences.

Hanson et al. (1979) studied the long term effects that
a devastating tornado in Flint, Michigan had on the coping
behaviors of people who had experienced the tornado and who
still lived in the area. The main finding of the study was
that the effect on coping behavior was striking even after a
period of over twenty years.

Baum et al. (1983), in an extensive discussion about
the differences between natural hazards and technological
catastrOphes, suggest that technological hazards may have
particularly long lasting psychological consequences. Their
rationale was that the duration of the ill effects of such
hazards are unknown.

Sims and Baumann (1983) in their review of the
literature disagree with the above conclusions. "The facts

are that hazard experience is only sometimes efficacious in

 

evoking subsequent preventive measures and then only under

 

39

certain circumstances and Often only for short periods"
(p.171).

Hypothesis 5

 

Anglers of high socioeconomic status will be more
accurate in their perceptions of toxic chemicals than
anglers of low socioeconomic status.

Mitchell (1974) found that age and socio-economic
status were most closely correlated with awareness of
various air pollutants. Mitchell felt that older, higher
status peOple were most likely to read or pay attention to
media sources. Hence, this group of individuals would be

most likely to receive the messages of any educational

efforts.

 

METHODS

This chapter is divided into three sections. The first
deals with sampling, the second deals with analysis and the
third deals with the Operationalization of the hypotheses.
The sampling section is given particular attention as the
population being sampled is not drawn from a list or a
definitively described population.

Sampling

Moser and Kalton (1972) describe two types of
populations in sampling: target populations and survey
pOpulations. The target pOpulation deals with the
population of interest and the one that the researcher would
most like to represent in the sample. The survey pOpulation
deals with the population that is actually obtained in the
study and is usually at least somewhat different than the
target pOpulation. The ideal is to minimize the gap that
falls between the two types (Moser and Kalton, 1972).

The target population for this study is represented by
all anglers who fish in Michigan. This population was
broken down into three groups. One group was those
anglers who fished in rivers containing levels of toxic
chemicals which made all fish unsafe to eat. The location
for this group corresponded to those rivers which received

the highest level of warning published on the 1984 Fishing

 

Guide (MDNRa, 1984). The second group was those anglers who

40

41

fished in rivers where the number of fish considered unsafe
to eat was not all inclusive. The location for the second
group corresponded to the next level of warning published in

the same 1984 Fishing Guide. The third group was composed

 

of those anglers who fished on rivers where all fish were
considered safe to eat. The location for this group
corresponded to those rivers which were not listed in the
1984 Fishing Guide.

The actual survey population used, however, was much
more limited than the target population described above.
The survey population is described in more detail later in
this section.

Several sampling methods were considered for this
study. Simple random sampling was the preferred
alternative. A list of anglers who purchase fishing
licenses was available. However, there is no list of el_
peOple who fish in Michigan. This was due to two reasons.
The first is that not everyone who fishes obtains a fishing
license. The second is that spouses may fish off one
license.

It was important to have the peOple interviewed
directly confronted with the conditions of the study. This
was done to minimize the problems associated with recall of
the stimulus of interest over uncontrolled amounts of time.
Such problems are not easily addressed by a mail-out survey.

A mail-out survey would also be expected to have a lower

42

return rate than personal interviews in addition to the time
and memory problems.

Cluster sampling was also considered for this study,
but was rejected because of the difficulty of identifying
identical clusters on the chosen sample sites. In general,
cluster sampling is used in urban settings to minimize the
time necessary to obtain a random sampling. Its usefulness
in this study would have been limited.

In the end, a stratified, systematic sampling design
was chosen. Stratification is used to ensure that various
elements of the target population are adequately represented
in the survey population. In stratification "."An attempt
is normally made to ensure that each subgroup or stratum in
the population is constructed in such a way that the
individuals or elements within a stratum resemble one
another" (O'Muircheartaigh, 1978, p. 16). "Each
suprpulation or stratum constitutes a group to which the
researcher may wish to generalize results" (Drew, 1980, p.
189).

Three types of rivers were needed for this study.
Randomly drawing three rivers in Michigan would not
necessarily lead to the selection of the types of rivers
desired. Further, the study sites were chosen on the basis
that they were free, public access sites located within

fifty miles of moderate size urban areas (populations >

43

50,000L. This was done to minimize the economic barriers
that fees and distance would place on anglers who wanted to
fish.

th case is widely

The '“.procedure of taking every i
used by professional samplers and is called systematic
sampling. Because of its simplicity and usefulness in
complex sampling situations, systematic sampling is probably
used far more frequently than simple random sampling)‘
(Sudman, 1976, p. 52L. ”In general this type of sampling

u. differs from simple random sampling in that it does not
give all possible samples of size n from the population of
size N an equal chance of selection" (Moser and Kalton,
1972, p. 83).

Generally Speaking, systematic sampling is used when
lists are available of the target population. However, the
method is also used in field studies. The problem with
using this method is that it assumes that people appear at a
location in random order (Bailey, 1978, p. 78).
“.uSystematic sampling can suffer from periodic
fluctuations in the population list but this is not a common
problem. If the periodic structure of the population is
known, a systematic sample can usually be designed to
capitalize on it." (O'Muircheartaigh, 1978, p. 17).

Sudman (1976, p. 52) suggests that only two things are
needed for systematic sampling: a random start and a

sampling interval. When a list is not available Bailey

44

(1978) suggested that if peOple descend upon a location in a
linear fashion then a random start could be used and an
interval chosen that would allow the completion of the
study.

The choice of an interval is designed to minimize the
chance for autocorrelation in the sample (Sudman, 1976, p.
57). This means that one event should not have any
relationship to a neighboring event. For example, when
taking surveys in the field one should not take more than
one survey per group unless group interactions are a primary
focus of the study.

Preliminary field work identified three rivers which
met the conditions of the sampling scheme: the Saginaw, the
Kalamazoo and the Grand. The Saginaw represented the high
toxic chemical contamination condition, the Kalamazoo
represented the middle condition and the Grand (where
sampled) represented sites with no warnings about toxic
chemical contamination.

Two different sampling conditions were present at the
sites used. One situation had anglers fishing along the
banks of the river. The other sampled anglers who used
public boat ramps. The sampling period of six hours was
equally divided between these two conditions.

Preliminary field work also indicated that most people
fished during the evenings and on weekends. A sampling

schedule was drawn up that sampled every Wednesday, Friday,

45

Saturday and Sunday from July lst through Labor Day during
'the summer of 1984. On Wednesdays and Fridays the sampling
periods were between the hours of 3 and 9 PM. On Saturdays
the six hour sampling periods were randomly drawn from the
hours of 9 AM to 9 PM. On Sundays the six hour sampling
periods were randomly drawn from the hours of noon to 9 PM.
These times were based on the heaviest use by anglers on the
chosen sites.

The location of the sample sites to be sampled each day
were drawn on a random basis. The sites used are shown in
Appendix A.

Sample Size

 

A review of the natural hazards literature showed
studies with as few as 67 respondents. Bailey (1978, p. 84)
suggested that 100 is the minimum acceptable to many
researchers. ‘No previous studies were found that dealt with
the perception of toxic chemicals by anglers, so an estimate
of the sampling error was not calculated. Consequently, a
target of 150 interviews per stratum was set for a total of
450 interviews to allow for nonresponse and interviewer
errors. iHowever, only 318 interviews were collected across
all three conditions: 100 from the Saginaw, 105 from the
Kalamazoo and 113 from the Grand.

Communication with other researchers who had

interviewed anglers during the summer of 1983 indicated that

 

46

1 to 3 interviews per hour could be performed. It became
clear during pretesting of the questionnaire that the
collection of interviews would sometimes be lower than this
rate. An additional interviewer was employed to insure an
adequate rate of interviewing during the interviewing period
between July lst and Labor Day, 1984.

Limitations

 

The sample sites for this study were deliberately
located to ensure that anglers of all socioeconomic levels
would have equal access to the study sites. However,
anglers who exclusively used private marinas were excluded
from the sample population.

Also, anglers who used remote fishing locations and not
heavily used public access sites were not included. In
addition, anglers who exclusively use fishing sites outside
of the study area were not included. Rural anglers, as
well, (iJL those who fish more than 50 miles from the urban
centers of the study) were not likely to be included in the
survey pOpulation.

Credibility ef Small Samples

 

 

Sudman (1976) uses a subjective scale that gives an
indication of the credibility of a small sample study. In
creating the sampling plan extensive use was made of the
scale. Due to limitations in time and money not all the

criteria were Optimized, but an effort was made to address

them all. What follows is a brief discussion of each of

 

47

these criteria and the sample plan used for this study.

Three criteria are mentioned by Sudman (1976) as
important to generalizability. The first deals with
geographic spread. It was not possible to address the
entire angler population of Michigan nor even the entire
angler population of the Lower Peninsula. However, it was
possible to have study sites that represented different
geographic locations in the lower half of the Lower
Peninsula. In addition, care was taken so that surveys were
collected from more than one location on each river.

The second criterion deals with a discussion of the
limitations of the sample. This has been outlined above and
was recognized before the study began.

The third criterion deals with the use of special
populations. Only the population of interest should be
included in the sampling plan. This meant that it is not
efficient, nor desirable to survey peOple who do not
properly belong in the population of interest. In this
study only those anglers who were confronted by contaminated
conditions were included in the sample pOpulation. Since
this was also the target pOpulation, the sampling scheme was
very efficient.

Three other areas of concern are also mentioned by
Sudman: sample size, sample execution and use of resources,

such as time, money and personnel. Sample size has already

48

been discussed. Sample execution deals with the actual
collection of the data and the quality of that collection
process. The response rate of 318 out of 338 contacts
suggests the sampling was executed prOperly. The use of
resources is very subjective. If a list of anglers who met
the criteria of the study had been available efficiency
could have been improved by minimizing the sampling cost per
response.

Analysis

 

To measure the perceptions of toxic chemicals by
anglers a previously constructed scale was sought. Because
of the scarcity of work in this area no such scale was
available in the literature. To measure these perceptions
three subscales were created relating to the presence of
toxic chemicals in the waters being fished, the presence of
toxic chemicals in the fish in the waters being fished and
the threat to health from toxic chemicals.

In accordance with the concept that it is necessary to
first perceive that a hazard is present before it elicits a
response, a series of items were created that fell into one
of the above three categories. It was decided to use a six
point forced choice response set to eliminate the problem of
the undefinable middle point and also to gain impressions
from respondents who might otherwise feel unqualified to
respond to the items. The procedures used to calculate

reliability were taken from Nunnally (1978).

49

The scale, along with other items in the questionnaire
were pretested during June, 1984. The initial reliability
coefficients for each of the subscales were: .73 for the
subscale pertaining to toxic chemicals in the water; .63 for
the subscale pertaining to toxic chemicals in the fish; .56
for the subscale pertaining to threats to health: and .79
when the three subscales were combined.

The low reliability coefficient of the fish and health
subscales resulted in the addition of more items to those
subscales to increase the reliability. However, the
additional items led to reliability coefficients lower than
those already obtained. The results of these additions were
unproductive and a decision was made to use the scales as
they were with the expectation that an increased sample size
would lead to an improved reliability coefficient. The
final reliability coefficients are found in the Results
section. The final items used in this scale can be found in
Appendix B.

Other items used were constructed following procedures
common to survey design and can be found in Appendix C
attached to this document.

Limitations 2: e Likert-type Scale

 

A Likert-type multi-item scale is one of the least
SOphisticated of many types of ordinal scales available.

With a possible range of scores of only one to six,

50

the standard deviation found within groups has to be very
small to allow any significant differences to be found.
Alternative methods, such as rank-orderings or paired
comparisons (Nunnally, 1978, p. 46), generally allow more
precise measurements. In addition, these scales may be
quantified with further mathematical manipulations.

The justification for using a Likert-type scale is that
it gives respondents a task that is easier and more common
than that of paired comparisons or rank-orderings. Further,
when high levels of precision are not required, such as in
this study, a Likert-type scale provides a relatively
efficient means of making crude categorizations.

The use of parametric statistics on such a scale
violates the general assumption of the data being gathered
on an interval or ratio level of measurement. However,
parametric statistics are often used anyways as a means of
exploring relationships that might not otherwise be possible
using the available nonparametric statistics. When this is
done, it is assumed that the underlying distribution of the
scale scores used is normal. Care should be taken to state
the level of measurement used so that readers may decide for
themselves the validity of such an analysis.

Statistics

 

The statistical procedure used in analyzing the data
included: multiple regression, one-way analyses of

variances, t-tests and chi-squares. The level of

51

measurement and sampling procedures conformed to the basic
assumptions of the statistical tests employed.

The t-test was used with data that was measured on an
interval scale or higher. Age, income and education were
all measured in this fashion. The chi-square test was used
when two variables, measured on a nominal scale were
compared to a variable measured on an interval or higher
scale.

The One-Way Analysis of Variance was used between
groups measured on an ordinal scale to take advantage of the
highest powered statistical test available. In this case,
the use of a parametric ANOVA was more likely to show any
differences between groups, if any, than a nonparametric
ANOVA. If the parametric ANOVA showed significant
differences the test would be replicated using a
nonparametric Kruskal-Wallis ANOVA.

A discussion by Blalock (1979) carries this issue a bit
further. A Kruskal-Wallis ANOVA is considered by Blalock
(1979) to have a power efficiency of ninety-five percent
compared to a parametric ANOVA when large samples are used
(p. 367). "If we refer to the power efficiency of one of
these non-parametric tests as 95 percent, we mean that the
power of the non-parametric test using 100 cases is
approximately the same as that of the t-test using 95 cases

if the model used in the t test is correctF (Blalock, 1979,

52

p. 253).

The power of a test is the probability that the test
will show significant differences when those differences
actually exist. By definition, a less powerful test is more
likely to have a type II error: not showing significant
differences when significant differences actually exist.
Using a more powerful test, again by definition, means that
one is less likely to make a type II error. Using the more
powerful parametric ANOVA means that one is minimizing the
chance of a type II error at the risk of violating the
assumption of a known underlying distribution. However,
with large groups, such as in this study, the assumption of
normality is generally relaxed as the data will approach
normality as the sample gets larger (Blalock, 1979, p. 253).
This study, as already stated, made use Of the more powerful
parametric statistics.

Multiple regression was used in a secondary analysis
of selected demographic variables. Multiple regression was
used in this analysis due to a lack of a parallel test using
nonparametric statistics. Again, the large sample size
meant that the distribution of the data approached normality
and the general robustness of the multiple regression
analysis to violations of many assumptions made the test
apprOpriate (Blalock, 1979, p. 387). The analysis included
a check for multicollinearity, nonlinearity, residual

outlyers and normality of the residual distribution.

53

Operationalization of the Hypotheses

 

Hypothesis 1:

 

Anglers will perceive real differences in the level of
toxic chemicals present in their fishing environments when
real differences exist.

The scale develOped to measure the perception of toxic
chemicals has been discussed above. This scale
acts as an indicator of the level of perceived toxic
chemicals present in the waters studied.

The real differences in the level of toxic chemicals
present are represented by the three levels of warning
published by the MDNR: all fish are unsafe to eat: certain
fish are unsafe to eat, should not be consumed by high risk
groups and/or should not be consumed in large quantities; no
warning exists about the presence of toxic chemicals at this
time (MDNR, 1984a).

The resultant Operationalized hypothesis is: The
perceived level of toxic chemicals, as indicated by the
scale scores, will be significantly different between, and
will correspond with, the three different conditions of
toxic chemical contamination.

Hypothesis 3;

 

For those who perceive the presence of toxic chemicals,
as the real level of toxic chemicals contamination increases

the number of anglers who consume the fish they catch will

54

decrease.

The three levels of toxic chemical contamination have
been discussed above.

Those anglers who do not consume the fish they catch
are considered to be participating in an avoidance behavior.
Figure 1 shows this as one of the choices in the decision
tree. A precondition of being able to make this choice is
also present in Figure 1: anglers need to have determined
that toxic chemicals are present before this choice occurs.
Anglers who do not perceive the hazard as being present are
assumed to not be responding to the threat posed by toxic
chemicals.

This precondition is measured using the scale described
above. Recall that the scale is a six point Likert-type
scale. The midpoint of the six point scale is 3.5. Any
score above this midpoint indicates at least some agreement
that toxic chemicals are present. Any score which falls
below this midpoint indicates at least some disagreement
that toxic chemicals are present.

For this hypothesis only those anglers who have average
scores above 3.5 are considered in the analysis. These
anglers are considered to at least slightly agree that toxic
chemicals are present on the sites they are fishing.

This criterion is followed across all three study
conditions including the condition where there is no warning

about the presence of toxic chemicals. Swartz and Strand

55

(1981) looked at the effect of misinformation on behavioral
responses to a hazard. They found that the perception of a
hazardous situation, even though false, could lead to
significant behavioral changes. Those anglers who
"perceive” the presence of toxic chemicals where those
chemicals are not present would be expected to behave in the
same way as if toxic chemicals really were present.

The Operationalization of this hypothesis is that the
number of anglers who avoid toxic chemicals by not eating
the fish they catch will be significantly and positively
correlated with the real level of toxic chemicals present.

Hypothesis 3

 

For those who perceive the presence of toxic chemicals,
as the real level of toxic chemicals increases so will the
number of anglers who participate in COping behaviors.

Again the conditions of the study relating to the
levels of toxic chemicals has been discussed extensively.

The precondition that anglers need to be able to
perceive the presence of toxic chemicals before they can
respond to their presence is the same as for Hypothesis 2.
The precondition is measured in the same way as for
Hypothesis 2.

Figure 1 illustrates two types of coping behaviors that

are used in this study. These two types of coping behaviors

are: alternative preparation procedures of the fish caught

 

56

and alternative cooking procedures. Alternative preparation
procedures were indicated by asking anglers how they
prepared the fish they caught. This question was asked
before toxic chemicals were mentioned. Those anglers who
responded that they skinned, cut off excess fat or filleted
the fish they caught were considered as making an effort to
COpe with the presence of toxic chemicals in the fish they
caught. These procedures paralleled those recommended in

the Michigan Fishing_Guide (MDNR, 1984a). Alternative

 

cooking procedures were defined as baking, broiling or
smoking the fish caught.

An angler who participated in fly one of these
alternative procedures was considered as COping with the
presence of toxic chemicals in the fish caught. Coping, for
this study, is defined as a behavior to reduce the risk of
toxic chemicals while not avoiding the risk entirely.
Alternative preparation and cooking procedures reduce the
risk, but they are not as effective as not eating those
fish.

The Operationalization of this hypothesis is: The
number of anglers who participate in alternative preparation
and cooking procedures of the fish they catch is
significantly and positively correlated with the real levels

of toxic chemicals present in the areas being fished.

57

Hypothesis 4

 

Anglers who have had negative experiences with
consuming fish contaminated by toxic chemicals in the past
will participate in coping behaviors at a higher rate than
anglers who have not had such experiences. COping behaviors
are measured in the same way as described for Hypothesis 3.

To indicate past negative experiences with toxic
chemicals anglers were asked if they, or anyone they knew,
had been made sick from fish they believed were contaminated
by toxic chemicals during the last year. Those who answered
that they had had such an experience were considered as the
group with negative experiences from toxic chemicals. All
other anglers formed the no negative experience group.

The Operationalization of this hypothesis is: Anglers
who have been made sick from fish they believed were
contaminated by toxic chemicals, or knew someone who had been
made sick during that same time period, will participate in
COping behaviors at a significantly higher rate than anglers
who have not had such negative experiences.

Hypothesis 5

 

Anglers of high socioeconomic status will be more
accurate in their perceptions of toxic chemicals than
anglers of low socioeconomic status.

Accuracy in perception was measured making the
following assumptions about the toxic chemical perception

scale: Anglers who fished under the two conditions where

58

toxic chemicals were present were considered accurate in
their perceptions if their averaged scale scores were
greater than the midpoint value of 3.5. Anglers who fished
under the condition where no toxic chemicals were present
were considered accurate if their averaged scale scores were
less than the midpoint value of 3.5.

Anglers who agree to any extent that toxic chemicals
were present when toxic chemicals were indeed present and
disagreed to any degree when toxic chemicals were not
present were considered as accurate in their perceptions.
Anglers who do not meet this criteria were considered
inaccurate in their perceptions. Age, income and education
were used as indicators of socioeconomic status.

The Operationalization of this hypothesis is: Age,
income and/or education will be significantly higher for
anglers who are accurate in their perceptions of toxic
chemicals than for anglers who are inaccurate in their

perceptions of toxic chemicals.

RESULTS

The overall research paradigm for this study is
summarized in Figure l in the Introduction. This model is
divided into two major parts: the perception of toxic
chemicals and the behavioral responses to the presence of
toxic chemicals.

The study was designed to determine if toxic chemicals
in waters fished by anglers exceed the ”threshold of
awareness" mentioned by Burton et al. (1978) where toxic
chemicals actually exist. When toxic chemicals are
perceived to be present, this study looks at how anglers
respond to the perceived presence. Lastly, two mediating
variables, past experiences with toxic chemicals and socio-
economic status, are examined to determine if they are
associated with behavioral responses to toxic chemicals.

Reliability of Perception Scale

 

Perception of toxic chemicals is measured by a scale
constructed specifically for this study. This scale is
composed of nine items shown in Appendix B. These nine items
had a reliability coefficient of .91 with inter-item
correlations ranging from .35 to .75.

This scale was formed by combining two subscales. One
was composed of five items that asked about the presence of
toxic chemicals in the waters being fished. This subscale

had a reliability coefficient of .83 with inter-item

59

6O

correlations ranging from .35 to .65.

The other subscale was composed of four items that
asked about the presence of toxic chemicals in the fish.
This subscale had a reliability coefficient of .80 with
inter—item correlations ranging from .41 to .75.

In addition to the above two subscales another subscale
composed of four items was created. This subscale had four
items and asked questions that dealt with the perceived
threat to health from toxic chemicals. The reliability
coefficient for this subscale was .50 with inter-item
correlations ranging from .07 to .46. Because of the low
reliability of this subscale it was not used in the analysis
of the data.

According to Nunnally (1978, pp. 245-246) an acceptable
reliability coefficient for studies where great precision is
not required would be about .70. The scale used for the
analysis of the data exceeded this minimal level of
reliability.

Control for Environmental Conditions

 

In selecting study sites, it was impossible to control
for all details of the environment where reSpondents were
located. To provide some estimate of equivalency across the
study sites several ten-point questions were asked of
respondents pertaining to the surrounding land area and the

waters being fished.

61

The characteristics for the surrounding land areas
were: beautifulness, naturalness and cleanliness. The
characteristics for the waters being fished were: clarity,
cleanliness, fragrancy, freshness and overall water quality.

The only characteristic that was significantly
different across the three conditions of the study was the
beauty of the surrounding land area (Table 1).

Hypothesis 1

 

Anglers will perceive real differences in the level of
toxic chemicals present in their fishing environments when
real differences exist.

Anglers were asked to give their degree of agreement
with nine items concerning the presence of toxic chemicals
in the waters and in the fish. It was expected that anglers
would show the greatest degree of agreement that toxic
chemicals were present at the most contaminated site, the
least at the noncontaminated site and some intermediary
level of agreement for the moderately contaminated site.

In Table 2 a six would indicate strong agreement that
toxic chemicals were present in the fish and water. A one
would indicate strong disagreement that toxic chemicals were
present.

In this study, there were no significant differences in
the perceived levels of toxic chemicals across the three

conditions. The null hypothesis cannot be rejected.

62

Table 1

Rating of the Beautyyof the Surrounding Land
Across the Three Study Conditions

 

Grand River Kalamazoo River Saginaw River
(No contamination) (Medium) (High)

mean 7.1 6.1 6.4

80d. 1.9 2.2 2.3

N 111 105 100

 

Note: One-way ANOVA; df = 2, 313: F = 6.978; alpha = .0011

1 - The surrounding land area is ugly
10 - The surrounding land area is beautiful

63

Table 2

Toxic Chemical Perception Scale Scores Across*
Three Levels of Toxic Chemical Contamination

 

 

Grand River Kalamazoo River Saginaw River
(No contamination) (Medium) (High)

mean 3.8 4.2 4.1

S.d. 1.2 1.2 1.4

N 100 95 93

Note: One-Way ANOVA; df = 2, 285; F = 1.85; alpha = N.S.

6 - strongly agree that toxic chemicals are present
1 - strongly disagree that toxic chemicals are present

64

Hypothesis 2

 

For those who perceive the presence of toxic chemicals,
as the real level of toxic chemical contamination increases
the number of anglers who consume the fish they catch will
decrease.

Two types of avoidance behavior were identified for
this study: avoidance of the fishing location and avoidance
of fish containing toxic chemicals by not consuming those
fish. Since this study took place on-site, those peOple who
avoided fishing at those locations were not part of the
sample population.

However, the threat from toxic chemicals in fish can be
avoided by not eating those fish. An angler can still
participate in fishing and, if the fish are not consumed,
that angler can avoid the threat posed by contaminated fish.

This hypothesis also presupposes that anglers can
perceive the presence of toxic chemicals. Only those anglers
who indicated at least some agreement that toxic chemicals
were present were included in this analysis.

Table 3 summarizes these results. The chi-square for
this analysis is nearly significant at the .05 level. The
trend depicted in Table 3, however, is not in the predicted
direction. The null hypothesis cannot be rejected.

Hypothesis 3

 

For those who perceive the presence of toxic chemicals,

as the real level of toxic chemicals increases so will the

65

Table 3

Planned Consumption of Fish Versus Location

 

(Percentages in parentheses)

 

 

Grand Kalamazoo Saginaw Row
River River River Totals
Plan to 28 38 26 92
eat fish (49.1) (66.7) (45.6) (53.8)
caught
Do not 29 19 31 79
plan to (50.9) (33.3) (54.4) (46.2)
eat fish
caught
Column 57 57 57 171
Totals (33.33) (33.33) (33.33) (100.0)
Note: Chi-square = 5.83; df = 2; alpha = .054

66

number of anglers who participate in COping behaviors.

The rationale for this hypothesis is the same as for
hypothesis 3. It was expected that anglers would
participate in coping behaviors in greater numbers when the
real level of toxic chemicals was greatest. Coping
behaviors were indicated in this study by those anglers who
participated in any one of several alternative methods of
fish preparation or cooking.

Again, it was a precondition for inclusion in the
analysis that anglers perceived the presence of toxic
chemicals in the waters they were fishing and/or the fish in
those waters.

Table 4 summarizes these results. The chi-square for
this relationship is highly significant, but the trend is in
the Opposite direction as that predicted. Anglers at the
most highly contaminated condition participate in coping
behaviors at a lesser rate than anglers in either of the
other two conditions. The null hypothesis cannot be
rejected.

Hypothesis 4

 

Anglers who have had negative experiences with
consuming fish contaminated by toxic chemicals in the past
will participate in some form of coping behavior at a higher
rate than anglers who have not had such experiences.

Past experiences with natural hazards has been shown to

lead to an increase in COping behaviors in some studies.

67

Table 4

CopingyBehavior Versus Location
(Percentages in parentheses)

 

 

Grand Kalamazoo Saginaw Row
River River River Totals
Cope with 38 45 ' 24 107
toxic (60.3) (66.2) (38.7) (55.4)
chemicals
Do not 25 23 38 86
cope with (39.7) (33.8) (61.3) (44.6)
toxic
chemicals
Column 63 68 62 193
Totals (32.6) (35.3) (32.1) (100.0)

Note: Chi-square = 10.80: df = 2; alpha = .0045

68

The hazard in this study is fish containing toxic chemicals.
It was expected that past negative experience(s) with such
fish would lead to an increase in the number of COping
behaviors by anglers.

Past experience with the negative effects of fish
contaminated by toxic chemicals was indicated by asking
anglers if they or anyone they knew had been made sick from
fish they believed were contaminated by toxic chemicals
during the past year. Coping behaviors were to be indicated
in the same fashion as represented in Hypothesis 3.

The number of anglers who said that they had had such
negative experiences was extremely low. Only 14 of the 312
anglers said they or someone they knew had been made sick
from fish they believed were contaminated by toxic
chemicals. The average number of instances of these
experiences for these 14 people was 1.2. Any further
analysis of this data, because of the low response, would be
misleading.

Hypothesis 5

 

Anglers of high socioeconomic status will be more
accurate in their perceptions of toxic chemicals than
anglers of low socioeconomic status. Age, income and
education were used as indicators of socioeconomic status.

Accuracy in the perception of toxic chemicals was
indicated by using the toxic chemical perception scale

scores. Anglers were judged "accurate” when their scale

69

scores showed at least some agreement with the actual
conditions that existed at each of the study sites. Anglers
were divided into two groups: those whose perceptions were
accurate and those whose perceptions were not accurate.

A t-test was run for each of the socioeconomic status
indicators as each was measured using a ratio scale of
measurement between accurate and inaccurate anglers to
determine if there were any significant differences.

Table 5 summarizes the results of this analysis. None
of the socioeconomic status indicators was significantly
different between the two types of anglers. The null
hypothesis was not rejected.

Survey POpulation Versus a Larger Scale Study

 

The Fisheries Division of the Michigan Department of
Natural Resources in COOperation with the Park and
Recreation Resources Department at Michigan State University
is conducting a large-scale study of anglers across the
state of Michigan (Mahoney, Jester and Kikuchi, in
preparation). The sample size for that study is 5155 and is
composed of all anglers who purchase fishing licenses in
Michigan. The preliminary results from that study are
presented here to provide a comparison between the two
survey populations.

To differentiate between the two studies, the study

being done by the Fisheries Division will be referred to as

on Age,

70

Table 5

Accurate Versus Inaccurate Perception Groups

Education and Income

 

 

(means)
Variable Accurate Inaccurate t-value probability
Age1 38.4 39.9 -.88 .19
s.d. 14.6 14.4
N 168 119
Education 11.8 11.5 1.0 .15
s.d. 2.7 3.1
N 167 120
Income3 29.6 22.6 1.1 .13
s.d. 61.6 15.6
N 153 106
1 - in
2 - in
3 - in $thousands

71

the MDNR study. The results of the sampling from the study
done for this dissertation will be referred to as the on-
site study.

In the MDNR study 94.5% of respondents were male. In
the on-site study 84.6% of respondents were male. The
practice of allowing spouses to fish off one license may
help explain the high number of males in the MDNR study;
presumably males are the primary purchasers of licenses.

The overall mean of the age of respondents in the MDNR
study was 43.6 with a median of 40.9. In this study the age
mean was 39.5 with a median of 35.9. Teenagers sixteen or
over were eligible to participate in the on-site study.
Part of the difference may be due to the inclusion of a
small number of anglers who were under eighteen. These
anglers are not required to purchase fishing licenses and
would not be part of the MDNR survey population.

In the MDNR study 81.5% of the respondents were white,
2.2% were black with 16.3% being of some other race or were
part of the missing value set. In the on-site study 61.3%
were white, 32.4% were black, 4.4% hispanic, 1.6% oriental
and .3% of some other race. Since the rivers chosen for
this study were close to urban centers the high
representation of minorities may be partially explained by
this factor.

In the MDNR study, education was measured using a

Ilominal coding scheme that is not directly comparable to the

72

on-site study. Of those who responded to the question,
53.9% had at least a high school diploma. The median for
the on-site study was 11.9 years of education which would be
indicative of an education level just slightly less than the
MDNR study.

In the MDNR study the median Of annual family income
was approximately $24,922. In the on-site study the median
was $19,013. If poor peOple are less inclined to purchase
fishing licenses, then that propensity might explain the
difference between the two survey pOpulations.

The differences encountered between the two sample
pOpulations may be accounted for by the differences in
sampling methods and the types of sites sampled. If
minorities and/or poor people are less likely to purchase
fishing licenses then the sample from this study is probably
more representative of anglers found at rivers near urban
centers than the MDNR study.

Internal Profile of the Survey Population

 

To give an overall perspective on the entire survey
pOpulation and the three suprpulations, a brief summary of
some demographics is presented at this point. Because most
of that information is repetitive it will presented
primarily in tabular form.

The first variable shown, miles of home from the sample
site indicates that anglers travelled a greater distance to

reach the Kalamazoo study sites than the other two study

73

sites (Table 6). This is most probably due to the study
sites on the Kalamazoo being located at a greater distance
from the selected urban centers than the Saginaw or Grand
river study sites.

Table 7 shows that the prOportion of males to females
was consistent across the study conditions. Age also,
shown in Table 8, was approximately the same across the study
conditions.

Table 9 indicates that there was a greater
representation of minorities on the Saginaw than on either
of the other conditions. Tables 10 and 11 indicate that
education and income levels were lowest on the Saginaw when
compared to the other two conditions.

Description of Angler Behavior Towards Toxic Chemicals

Several variables were included in the study that were
not part of the hypotheses of the study. Selected variables
are included herein for their value in furthering
understanding of the population sampled.

The variables which will be described include:
familiarity with toxic chemicals: knowledge about detecting
the presence of toxic chemicals; knowledge about where to
find out about toxic chemicals; and the numbers of anglers

who read the warning on the MDNR brochure.

74

Table 6

Distance Anglers Travelled to Reach Study Sites
(in miles)

 

 

 

Grand River Kalamazoo River Saginaw River
(No contamination) (Medium) (High)
median 4.2 24.9 6.2
mean 41.0 57.7 24.0
s.d. 233.8 118.9 77.3
N 111 105 100
Table 7

Sex of Anglers Across Study Conditions

 

Grand River Kalamazoo River Saginaw River
males 83% 86% 85%
females 17% 14% 15%

N 112 106 100

 

75

Table 8

Age of Anglers Across Study Conditions

 

Grand River

Kalamazoo River

Saginaw River

 

 

 

 

 

median 34.5 36.7 36.2
mean 37.9 39.8 41.1
s.d. 13.3 14.2 16.1
N 110 106 100
Table 9

Race of Anglers Across Study Conditions

Grand River Kalamazoo River Saginaw River
Black 21.4% 25.5% 52.0%
Hispanic 3.0% 4.7% 6.0%
Oriental 1.6% 0.0% 1.0%
White 71.3% 69.8% 41.0%
Other .3% 0.0% 0.0%
N 112 106 100

 

76

Table 10

Education Levels Across StudyyConditions

 

 

 

Grand River Kalamazoo River Saginaw River
median 12.1 11.9 11.7
mean 12.5 11.6 10.9
s.d. 2.7 2.8 3.4
N 110 105 100
Table 11

Annual Family Income Across Study Conditions
(in Sthousands)

Grand River Kalamazoo River Saginaw River

 

median 22.0 20.0 15.9

N 101 . 101 83

 

77

Table 12 indicates that anglers were generally
unfamiliar with toxic chemicals, but there were no
differences across the three study conditions.

The single greatest response when anglers were asked
how they would detect the presence of toxic chemicals in the
environment was "don“tlknow". The second most pOpular
response was "would get sick”. Table 13 summarizes these
results across sites.

When asked where they would seek information about
toxic chemicals if they wanted to know more, almost half the
anglers indicated they would seek such information from the
Michigan Department of Natural Resources. A substantial
number said they did not know. Andra smaller percentage
mentioned the Public Health Department. These results are
summarized in Table 14.

Lastly, anglers were asked if they had read the MDNR
brochure. Of those who responded to this question, about
half the anglers said no or that they did not get one and
about ten percent said that they had read the brochure.
Table 15 gives these results.

Secondary Analysis

 

As a result of the initial inspection of the results,
an additional analysis of the data was prepared. This
additional analysis consisted of a multiple regression of
the perception scale scores on the variables of age, income,

education and miles of angler from home.

78

Table 12

Familiarity with Toxic Chemicals Across Study Conditions*

 

Grand River Kalamazoo River Saginaw River
mean 2.5 2.4 2.4
s.d. .9 1.0
N 109 105 97

 

5 = extremely familiar; 1 = not at all familiar

 

Table 13

How Anglers Detect the Presence of Toxic Chemicals

Grand River Kalamazoo

Saginaw Total

 

Mentioned Don“t Know DonFt Know

most (24.1%) (26.7%)
often
Second Dead Fish MDNR
most (10.2%) Tests
mentioned Appearance
of water
(6.7%)

N 108 105

Don‘t Know Don“t Know
(23.5%) (24.7%)

Get Sick Get Sick
(13.3%) (3.7%)

98 311

 

79

Table 14

Where Anglers Find Out About Toxic Chemicals

 

 

 

Grand River Kalamazoo Saginaw Total
Mentioned MDNR MDNR MDNR MDNR
most (46.8%) (50.5%) (48.5%) (48.6%)
often
Second Public Don't Know Don't Know Don't Know
most Health (13.3%) (21.6%) (14.8%)
mentioned (11.9%)
Third Universities Universities Public Public
most (12.4%) (5.7%) Health Health
mentioned Library (7.7%)

(6.2%)
N 109 105 97 311
Table 15

Anglers Who Read MDNR Brochure Across Study Conditions

 

Grand Kalamazoo Saginaw Total
Read 54.1% 49.5% 45.9% 50.0%
brochure
Did not 34.9% 38.1% 42.9% 38.5%
read
brochure
Did not 10.1% 11.4% 10.2% 10.5%
receive
brochure
Don't .9% 1.0% 1.0% 1.0%
Know

N 109 105 98 312

 

80

The above variables were chosen to determine if
identifiable personal characteristics might benefit in
predicting the perception of toxic chemicals. Such
predictions would enhance management efforts as groups of
people could be targeted for maxinuunlnanagement focus.

An unforced stepwise multiple regression was run to
determine the Optimal linear solution. Scatterplots were
done for each of the independent variables and the
perception scale scores to insure that those relationships
were not curvilinear. The results of these scatterplots did
not give any indication of such a curvilinear relationship.

A basic assumption of regression models is "u. that
the random error terms are either uncorrelated random
variables or independent normal random variables” (Neter and
Wasserman, 1974). Two tests were run to check this
assumption. The first consisted of a scatterplot of the
residuals. Less than three percent of the residuals were
more than two standard deviations away from the mean and
none were greater than three standard deviations away from
the mean. In addition, no regular pattern, such as a sine
wave or skewness, showed up in the scatterplot. Pictorially
the distribution of the residuals appeared random and in
compliance with this basic assumption.

To further check for autocorrelation, a Durbin-Watson
test was run and a value of 1.99 was obtained. The upper

limit for the D statistic with n = 253 and the number of

81

independent variables equalling four, is 1.74 for alpha
equal to .05 (Neter and wasserman, 1974, p. 816). This
gives clear evidence that the residual terms were not
significantly correlated.

Multicollinearity refers to the correlation of the
independent variables in a regression equation with each
other (Pedhazur, 1982, pp.232-244). If these correlations
are very high they may lead to results which, though correct
are meaningless. For example, when two independent
variables are perfectly correlated, it can be shown that
there are an infinite number of possible regression
equations that perfectly fit the data (Neter and Wasserman,
1974, p. 340). There is, however, no present consensus as
to what constitutes unacceptable intervariable correlation
(Pedhazur, 1983).

Thus, an examination of a correlation matrix by a
researcher may lead to inaccurate conclusions about the
presence or nonpresence of multicollinearity. To counteract
this problem, the results of this analysis were taken to a
statistician for an objective evaluation. The results of
that consultation were that no significant multicollinear
effects were affecting the regression equation (Crabb,
1985). The range of simple correlations between the
independent variables ranged from -.25 to .28. There were

relationships between variables, but not enough to affect

82

the outcome of the analysis (Crabb, 1985).

The actual stepwise regression analysis using the
perception scale scores as the dependent variable and age,
income, education and miles from home as the independent
variables led to the inclusion of only three of those four
variables: age, miles and income. The variable education
did not contribute significantly to the overall explained
variance. The resulting regression equation, using beta
instead of B, was:

PSS = 4.87 - .25A - .17M + .151

where PSS = Perception Scale Score
A = Age
M = Miles from home
I = Income

The multiple R for the equation was .34 with an R2
value of .11. The variance accounted for by miles, age and
income was small, but significant. A summary table of the

results appears in Table 16.

Perception versus Age,

83

Table 16

Summary of Stepwise Multiple Regression Analysis on
Education and Miles from Home

Income,

 

Variable* R2 Pearson's r Beta F**
Age .06 -.25 .25 16.1
Miles .09 -.18 .17 8.4
Income .11 .14 .15 5.7
Education .11 .12 .01 0.0

 

**

In order of inclusion into the

F
F
F
F
(B

1

6.64
4.60
3.78
3.32

alock,

(df=1,infinite)
(df=2,infinite)
(df=3,infinite)
(df=4,infinite)
1979, p. 615)

alpha
alpha
alpha
alpha

stepwise equation.

.01
.01
.01
.01

DISCUSSION

 

The basic foundation of this study is the expectation
of public agencies that the general public will use the
economic rationality model in its decision making process.
This led to a series of hypotheses, borrowed from the
natural hazards literature, to determine if anglers indeed
respond in a rational manner to the presence of toxic
chemicals in the fishing environment. This section examines
and attempts to explain the results of those hypotheses
first looking at the issue of perception, then behavioral
responses and lastly, mediating variables.

Perception of Toxic Chemicals

 

Mitchell (1974), when discussing the ability of people
to detect several types of air pollutants, spoke of the need
for these pollutants to be present in ”chronic prOportions"
to be perceived. Apparently, in this study the presence of
toxic chemicals did not reach chronic prOportions. Anglers
did not show any significant differences in perceptions of
the levels of toxic chemicals across three levels of toxic
chemical contamination.

If toxic chemicals are not perceived to be present, the
choice process cannot begin (Burton et al., 1978). Figure 1
illustrates that anglers who do not perceive the presence of
toxic chemicals would not be confronted with the need to
choose from alternative responses to toxic chemicals.

The decision making framework and problems discovered

84

85

in the natural hazards literature is summarized in this
quote from Sims and Baumann (1983):
'By far the most prevalent and enduring logic
determining policy in attempts to influence response to
the threat of natural hazards is this: The provision of
information on the hazard risk and education in proper
mitigation measures will result in the public under
threat being aware of both the nature of the risk and
the wisdom of protective measures".lt doean:
necessarily follow that because information is given it
is received or because education is provided there is
learning...” (p.187).
The results of this study do not support the contention that
providing the public with correct information leads to the
public being sensitive to the presence of toxic chemicals.
Recalling the economic rationality model described by
Slovic et al (1974), one of the assumptions of that model
is that peOple possess perfect information about the
alternatives available to them and the consequences of
choosing each of those alternatives. In a study pertaining
to familiarity with local parks, Stynes et al (1985) found
that people simply were not aware of all the parks available
to them. Stynes et al (1985) suggest that other variables
may be necessary to get at the complexity of how people
become aware of parks. In a similar way, anglers in this
study simply did not accurately perceive the presence of
toxic chemicals even though information about those
chemicals was available. Without an accurate assessment of

the degree of threat present, anglers do not possess perfect

information and the behavioral responses anticipated by

86

management will occur only by chance.

Such findings suggest a change in management strategy.
Lucas (1982) discusses the issue of ”soft" and "hard"
management regulations. As a rule the softest type of
regulation that is effective is most apprOpriate. Too much
regulation may be worse than no regulation at all.

However, the results of this study indicate that
greater efforts may be needed to make the presence of toxic
chemicals more "perceivable" to anglers. Such efforts as
warning signs or some form of physical patrolling of
hazardous sites might be apprOpriate. Radio or television
advisories similar to those given for air pollution in big
cities might be an innovative approach to heightening the
perception of toxic chemicals. Informal discussion with
anglers during the study hinted that warning signs were
something they look for to help them determine the safety of
a site.

The consequences of increasing regulations will surely
lead to some impact on recreation experiences. As already

discussed, leisure is voluntary in nature and freedom is a

 

central aspect of the experience (Gunter, 1979). Increasing
regulatory efforts will change the fishing experiences of
anglers on affected rivers. Further research is needed to
determine exactly what these impacts may be.

In addition, the lack of perception by anglers of the

87

presence of toxic chemicals in the waters they are fishing
points out a difference between ”typical" natural hazards
and toxic chemicals. "u. toxic chemicals are not
detectable by the human senses. Unlike floods, earthquakes,
fires or tornados there are no physical signs or damage to
aid in the identification of the hazard itself (Baum et a1,
1983). People must rely on specialized measuring
instruments or expert help to determine if toxic chemicals
are present in their environment" (Udd and Fridgen, 1985).

An illustration of how people determine that
undetectable hazards are present is presented by Barker
(1970). In that study Barker found that over fifty percent
of the peOple who participated in a study about water
pollution in the Toronto area felt that they could determine
if the water was polluted by using the sense of sight or
smell.

In this study the most common response to a question
about how anglers would detect the presence of toxic
chemicals was "don‘t know" (Table 13). The second most
common response was “would get sick? (Table 14). Neither
answer suggests that anglers are well-informed about toxic
chemicals.

In seeming contradiction, however, more than half of
the anglers in this study said that they would seek
information about toxic chemicals from either the MDNR or

Michigan Public Health Department. Both of these sources

88

would provide accurate information about the risk from toxic
chemicals. Apparently, the threat from toxic chemicals is
not perceived to be of sufficient magnitude for the anglers
of this study to seek more information.

In contrast to most natural hazards, the effects of
consuming toxic chemicals is of an uncertain magnitude and
of an uncertain duration. This difference is one noted by
Baum et al. (1983) between technological disasters and
natural hazards. With the scientific community in doubt
about the real threat from toxic chemicals, it cannot be
expected that the public will be well-informed about this
threat.

As already mentioned, there is considerable debate as
to the degree Of risk from toxic chemicals. Anglers may
make the decision that the risk from toxic chemicals is too
small to be of consequence to them. The risk that they
perceive may be much lower than the actual threat defined by
the MDNR. This would be consistent with the bounded
rationality model discussed by Slovic et al. (1974). The
differences in the decision making framework of the public
and the MDNR expected decision making framework may help
explain why anglers do not behave in the manner desired by
the MDNR (Svenson and Fischhoff, 1985).

Anglers who would use an economic rationality model of

decision making, as previously explained, would be expected

89

to follow an Optimal path that would minimize their risk
from toxic chemicals. However, the lack of perception of
the presence of toxic chemicals by anglers would preclude
the decision making process from even occurring and would
make the issue of which is the better model a moot point.

In summary, the differences in the levels of toxic
chemicals present in this study were apparently not detected
by anglers. This suggests that the provision of information
about toxic chemicals has not brought the presence of toxic
chemicals to the point of perceptability to anglers who
continue to fish in contaminated rivers. Alternate
management strategies may be necessary to raise the
visibility of the hazard of toxic chemicals to anglers.
Presently the perceived threat from contaminated fish
appears too low for the anglers of this study to actively
seek more information.

Behavioral Responses to Toxic Chemicals

 

Baker et al. (1980) conducted a study to determine the
impact on visitation to a nearby beach from the potential
construction of an offshore nuclear plant. Approximately 25
percent of the peOple they studied said they would avoid
such a beach. However, when they examined actual behavior
at a beach on a lake located near a nuclear plant they found
actual avoidance was only 5 to 10 percent.

Anglers who avoided fishing on the rivers in this study

due to toxic chemical contamination were not part of the

90

survey population. However, the anglers who were part of
this study were asked if they would stop fishing at the
study site if they knew toxic chemicals were present.
Forty-eight percent of the anglers said they would avoid
fishing if they knew toxic chemicals were present. If
anglers in this study behaved in a manner proportional to
the beachgoers in the study by Baker et a1 (1980), the
actual avoidance behavior of anglers towards toxic
contaminated waters would be expected to be from around 10
to 20 percent.

To estimate actual avoidance behavior, anglers in this
study were asked if they had avoided fishing in waters they
believed were contaminated by toxic chemicals during the
past year. Twenty percent of the anglers in this study said
that they had avoided such waters in the past. The three
water areas most often mentioned were the Kalamazoo River
(n=11), the Titabawassee River (7) and the Saginaw River
(6). Each of these rivers did indeed have significant
levels of toxic chemical contamination. The percentage of
anglers who said that they would avoid waters contaminated
by toxic chemicals is within the range expected from
extrapolating from the Baker et al (1980) study.

Actual avoidance behavior towards toxic chemicals might
have significant negative consequences on economies wholly

or partially dependent on fishing revenues. Specific

91

monetary losses to the economy have already been discussed
(Swartz and Strand, 1981; Schulstad and Stoevener, 1978). A
loss of 10 to 20 percent of anglers who might otherwise fish
at a river cannot be ignored economically.

It should be noted that the rivers used in this study
are far from being the most valuable sport fisheries in the
state of Michigan. However, toxic chemicals are not limited
to urban centers as noted in the Michigan Fishing geige
(MDNR, 1984a). The valuable Great Lakes sport fisheries
also are contaminated by toxic chemicals. How this
contamination is affecting the economies of the Great Lakes
states is a question for future research.

Besides not fishing at a specific location, anglers can
avoid the risk of consuming fish contaminated by toxic
chemicals by not eating the fish they catch. This is also a
unique characteristic of the hazard presented by toxic
chemicals. "A few representative samples of natural hazards
literature (Jackson, 1981; Hanson et al, 1979; Preston et
al, 1983; Hansson et al, 1982; Evans and Jacobs, 1981),
reveal that most of the subjects either live in the path of
the potential hazard or the threat is a part of their
everyday living environment (i.e. air pollution). With
anglers, toxic chemicals in fish only become a threat when
people enter a specific environment (where hazardous levels
of toxic chemicals exist), participate in a specific

recreation activity (fishing), and consume the fish they

92

catch ..." (Udd and Fridgen, 1985, p. 2). Anglers have at
their disposal a very simple means of protecting themselves
from contaminated fish: don't eat the fish.

The relationship between the level of toxic chemicals
and the consumption of fish was close to being significant
(alpha = .0541). However, an examination of Table 3 shows
that the relationship is not a linear one. Anglers who fish
under the moderate toxic chemical contamination condition
planned on consuming the fish they caught in greater numbers
than the anglers under the high and no toxic chemical
contamination conditions. Those anglers who were under the
greatest threat from toxic chemicals consumed fish at
approximately the same rate as those anglers under the no
threat condition.

One possible explanation for such behavior might
revolve around dissonance theory (Festinger, 1957). "With
respect to both natural and manmade environmental hazards,
there appears to be a pervasive tendency for inhabitants of
high risk areas to: (1) deny the threatening aspects of
their environments, and (2) markedly underestimate the
probability of future disasters or calamities" (Shippee et
al., 1980, p. 36).

The anglers in this study did not own the property from
which they fished. However, the anglers who fished on the

Kalamazoo made the greatest effort to reach their fishing

93

locations based on the distance they travelled (see

Table 6). Since these anglers came the greatest distance
they also had the most effort invested in reaching the site
they were fishing. Their denial might be based on the fact
that they would not make such an effort to come to a fishing
spot where the fish were not safe to eat.

It should be noted that the anglers in the above
analysis were only those who reported at least some
agreement that toxic chemicals were in the waters they were
fishing. So, the analysis bypassed the problem of lack of
perceiving the presence of toxic chemicals by selecting only
those anglers who did have such perceptions. Nevertheless,
such perceptions did not lead to reduced consumption:

'80 strong is the assumption of human rationality .u

that many studies deal only with the causal link

between the provision of information and heightened
awareness, assuming that if that is achieved,
behavioral change will inevitably follow.uthe evidence
supporting the causal link between hazard awareness and

protective responses is minimal.”' (Sims and Baumann,
1983, p. 168).

 

To further examine the relationship between information
and behavior, an analysis was done to determine if there was
any relationship between having read the MDNR brochure and
planned consumption behavior of fish contaminated by toxic
chemicals. Having read the MDNR brochure was not related to
the planned consumption of fish (chi-square = 1.08; df = 2;
alpha==.58).

In addition to avoidance behavior, anglers may reduce

94

the risk from toxic chemicals by using alternate preparation
or cooking procedures. Such behavior parallels coping
behaviors towards other natural hazards. For example, when
confronted with the knowledge that a flood is imminent,
people take precautionary behaviors to minimize their losses
and danger to themselves and their families (Hansson et al,
1982).

The relationship between coping behaviors and the level
of toxic chemical contamination was very strong (chi-square
significant at .0045; Table 4). However, the direction of
the relationship was in the Opposite direction as that
predicted. .Anglers at the most contaminated site
participated in coping behaviors at a lesser rate than
anglers under the less contaminated conditions. The anglers
in this analysis were only those who at least slightly
agreed that toxic chemicals were present in the waters they
were fishing.

When studying people's responses to the potential
threat from earthquakes in San Francisco, Jackson (1981)
suggests "u. If earthquakes are among the most destructive
of natural disasters, then a "rational model" would suggest
that all peOple living in recognized zones of high seismic
risk would be aware of the hazard, accept the possibility of
damages and take all possible precautions to minimize the

impact of future events. We have seen, however, that

95

response to earthquake hazard shares none of these expected
characteristics" (p. 408).

While toxic chemicals in fishing waters do not
represent a threat of the magnitude of a major earthquake, a
"rational model" would suggest that anglers would take some
actions to reduce their risk. However, the evidence of this
study suggests that this is not the case. Anglers not only
do not limit their consumption when toxic chemicals are
present, they appear, in this study, to participate in fewer
precautionary behaviors at the most hazardous site.

Again, dissonance theory might be a useful area of
future research. Anglers who knowingly consume contaminated
fish at the most hazardous sites are experiencing the
greatest degree of dissonance. They are exposing themselves
to the greatest level of threat, yet they continue to
consume the fish they catch. Denial and/or underestimating
the threat as discussed by Shippee et al. (1980) would be
consistent with the findings of this study. The result of
this denial is that some anglers are exposing themselves to
a risk identified by the Michigan Department of Public
Health as a threat to human health (MDNR, 1984a).

The limitations of the survey pOpulation should be
recalled at this point. The survey pOpulation did not
include those anglers who have stOpped fishing at
contaminated sites. It may be that the anglers who continue

to fish at contaminated locations are not representative of

96

peOple who fish in general. Anglers in this study have
lower incomes when compared to state anglers in general (see
p. 72). Further, the distance travelled to the fishing
sites is very short for most anglers in this study. These
two findings suggest that the anglers of this study do not
have as many fishing alternatives available to them as state
anglers in general. They may be fishing on “substandard"
sites simply because the Option of going somewhere else is
not available to them.

Alternatively, fishing for some of the anglers in this
study may be a necessity rather than recreational. These
anglers may be dependent upon fishing available sites, even
though they are contaminated, to supplement their food
supply. The necessity of this supplemental source of food
may outweigh any potential risk from toxic chemicals. This
necessity, in turn, reduces the voluntary nature of the
experience that generally characterizes fishing experiences.

In any case, if the population of this study is a
special case or representative of the angling pOpulation in
general, some anglers are at risk from toxic chemicals.

”u. When a real hazard is perceived as being minimal or
nonexistent, a much more intensive management program is
required" (Jubenville, 1978, p. 150).

Mediating Variables

 

Past experience with natural hazards is sometimes a

97

variable that can predict behavioral responses of people
confronted with a natural hazard. Mitchell (1974), in a
discussion of natural hazards research, suggests that
personal experience with a hazard may lead to more coping or
precautionary behaviors against a similar recurrence.

Hanson et al. (1979) found that having endured a
devastating tornado led to an increase in precautionary
behaviors even over twenty years after the event. In
another study, Hansson et al. (1982) found that people who
had experienced floods in the past were more willing to pay
increased taxes to help with flood prevention programs.

The range of findings in the literature that deals with
the predictive value of past experience range from very
strong relationships to no relationships at all (Sims and
Baumann, 1983). No studies that explored this relationship
for toxic chemicals and peOple were discovered.

The main finding of this study was that few anglers had
ever experienced negative consequences from the consumption
of fish contaminated by toxic chemicals. Only 14 of the 312
(< 5%) anglers who responded to this item reported that they
themselves or someone they knew had been made sick from
consuming fish that they believed were contaminated by toxic
chemicals.

It can be concluded from these findings that negative
consequences from the consumption of fish contaminated by

toxic chemicals represents a rare event so far. Slovic et

98

al (1974) discussed the difficulty people have in estimating
the probability of a rare event. Given the very small
probability of being made sick from fish contaminated by
toxic chemicals it would not be expected that anglers will
accurately assess the probable outcomes of choosing among
the various alternatives available to them.

The small probability of having experienced toxic
chemicals as a hazard corresponds to the probability of
having experienced any particular natural hazard (Baum et
al., 1983). However, the low numbers of anglers in this
study who had experienced such an event made this variable
of little value in predicting anglers' responses to toxic
chemicals. This is in contrast to its value in helping
predict responses to some other types of natural hazards
(Sims and Baumann, 1983).

The last hypothesis of this study dealt with the
question of whether anglers of greater age or higher
socioeconomic status would be more accurate in their
perception of toxic chemicals. Mitchell (1974), in a review
of a series of studies on air pollution in British cities
found that age and socioeconomic status were most closely
correlated with awareness of air pollution.

This study divided anglers into two groups: those who
were ”correct" in their perceptions of toxic chemicals in

the waters they were fishing and those who were “incorrect".

99

No significant differences were found when these two groups
were compared on the basis of age, income or education.

This finding does not help in identifying those anglers
who are able to accurately assess the level of toxic
chemicals present in the waters they fish. More
problematically, these findings do not help to identify
those anglers who are inaccurate in their perceptions of
toxic chemicals in the waters they are fishing.

Given the other results of this study, even those
anglers who perceive the presence of toxic chemicals in the
waters they are fishing do not necessarily respond in a
manner to protect their own health. The problem goes beyond
simply raising the level of awareness of anglers about toxic
chemicals.

If awareness and the possession of correct information
does not lead to anglers taking protective measures, other
means of protection may be necessary. Not forbidding
anglers from fishing on contaminated waters will lead to
some anglers being exposed to health threats associated with
toxic chemicals. However, such actions should not be taken
without consideration of the potential impacts of those
restrictions.

Because present knowledge about the extent of the
threat from toxic chemicals in fish is unknown this may be a
serious problem or only a relatively minor problem. Future

research which identifies the real probabilities associated

100

with consuming fish contaminated by toxic chemicals may
provide a guide to what extent anglers need protection.

Lastly, the secondary analysis using multiple
regression looked at the predictive powers of several
demographic variables and found a weak, but significant
relationship between the perception of toxic chemicals by
anglers and age, income and distance travelled to reach the
fishing site. For age and miles the relationship was
negative. The older anglers were, the lower the level of
toxic chemicals they reported. A possible explanation for
this comes from informal discussion with the anglers. A
cunnmon response was ”I've been fishing here for twenty years
and nothing has happened to me yet". The perceived
probability of being injured by toxic chemicals is much
less than the indisputable fact that the angler is still
alive. The long time frame that typifies the effects of
toxic chemicals undoubtably makes any eventual consequences
even more remote. (And, again, the scientific community has
had tremendous difficulty tracing health problems back to
the consumption of contaminated fish.

The distance travelled to the fishing site is also
negatively correlated with the perceived level of toxic
chemical contamination. Cognitive dissonance may play a
role in the following scenario: the farther an angler

travels to reach a fishing spot the greater the commitment

101

of time and money. It is difficult to internally
rationalize all that time and effort to reach a site that
the angler believes is contaminated by toxic chemicals. One
way to minimize this dissonance is to cognitively minimize
the dissonance (Festinger, 1957). Minimizing the presence
of toxic chemicals in the fishing environment would help
rationalize why the angler continues to fish at a
contaminated location.

Income, on the other hand, was positively correlated
with the perception of toxic chemicals. This is similar to
Mitchell's (1974) findings which suggest that high income
peOple are more likely to be exposed to information sources.
The vast volume of information about toxic chemical
contamination may lead to high income individuals at least
being suspicious of most surface water. Informal
conversations with anglers led to statements that many
people felt toxic chemicals were "everywhere” and just had
to be in the water they were fishing since they were
everywhere else. It may be that too much information in the
media leads to stereotyping the presence of toxic chemicals

by anglers.

CONCLUSIONS

Three levels of toxic chemical contamination were used
to examine differences in anglers in this study. Anglers
showed no differences in their perception of toxic chemicals
across those three conditions. They also showed no
differences in their behaviors to minimize the risk of
consuming fish contaminated by toxic chemicals across those
three conditions. If anything, fewer precautions were taken
under the most hazardous condition than for the other two
conditions. No strong predictive variables were found to
help identify accurate versus inaccurate 'perceivers” of
toxic chemicals.

The expectation that the provision of correct
information would lead to increases in awareness and to
precautionary behaviors to reduce the risk from toxic
chemicals was not supported by this study. The economic
rationality model discussed by Slovic et al. (1974) and
Kolers (1983) predicts choices based on accurate
information. Thus it is assumed that providing peOple with
accurate information will lead to predictable decision
making.

The bounded rationality model, also discussed by Slovic
et al. (1974) and Kolers (1983) suggests that people do not
make decisions based on perfect information and that they

will not always act in a manner that is predictable. Two

102

103

examples from this study help support that position. The
first comes from the inability of anglers to perceive the
presence of toxic chemicals. The second comes from the lack
of differences in behaviors between anglers who perceived
the presence of toxic chemicals.

The implication of this study is that the information
program instituted by the MDNR has not led to differences in
perceptions or behaviors under conditions of three types of
warnings about toxic chemical contamination. If the risk
from toxic chemical contamination is considered serious
enough, then new means of changing the behavior of anglers
should be explored.

Before any new programs are implemented, however,
economic analyses may be warranted to determine the impacts
on the local and state economies. It may be that the return
in added safety will not justify additional expenditures.

The problem becomes a balancing act between the
economic concerns, the rights of anglers to fish where they
choose and the possible health effects from not regulating
fishing activity. (All of these concerns cannot be maximized
simultaneously. At the time of this study, the rights of
anglers to fish where they choose is at a maximum. There
are no legal restrictions to prevent them from fishing at
sites which have hazardous amounts of toxic chemicals.

In summary, toxic chemicals in fish pose a health

threat to anglers. The decision tree model adapted from

104

Svenson and Fischhoff (1985) and from the natural hazards
literature provided structure for this research effort.
However, anglers did not choose those behaviors which would
minimize their risk from toxic chemicals in association with
the degree of actual risk present. Further intervention may
be necessary if the precautionary behaviors on sites
contaminated by toxic chemicals are to be increased
significantly.

Future Research Needs

 

Because of the limited nature of the survey pOpulation
of this study, a broader based study is necessary. Such a
study would determine whether the characteristics of the
anglers in this study are typical of all anglers or only
those of the sample population. If anglers in general do not
participate in risk-reducing behaviors then broad changes in
management strategies may be necessary. The implication of
this study is that the simple provision of printed
information in fishing regulations is not enough to
significantly alter behavior.

To explain anglers' responses to toxic chemicals an
examination and application of dissonance theory would
appear to be promising. In spite of knowing that toxic
chemicals are present in the water, anglers continue to
consume contaminated fish. Why they persist in this

behavior may be addressable through dissonance theory.

105

Smoking cessation studies might also prove productive
in providing study models for studying the relationship
between toxic chemicals and anglers. At present there are
no known addicting qualities attributed to toxic chemicals,
but the habit of fishing a particular site may be a greater
attractant than the deterrent posed by toxic chemicals.

Communication networks deserve study as well as they
may provide clues on how to increase the awareness of
anglers about toxic chemicals. The deliverance of the
information may have more credibility if it is delivered
through already existing communication networks.

The risk taking literature in recreation generally
revolves around the enhancement of an experience due to the
presence of some known risks. Toxic chemicals do not seem
to offer much in the manner of enhancing any kind of
experience, but the principles involved in the risk taking
literature might lead to a greater understanding about the
relationship between toxic chemicals and anglers.

As discussed earlier, informal discussion with anglers
suggested that some anglers felt toxic chemicals are
literally everywhere and therefore cannot be avoided. The
perceived control literature may lead to some understanding
of why such an attitude would exist.

Perceived control relates to the concept of whether
peOple feel they do or do not have personal control over

their environment and the events that occur in their lives

106

(Rothbaum et al, 1982). Individuals who feel they have a
great deal of control over their environment are said to

have an internal locus of control. Those who feel they do

 

not have a great deal of control are said to have an

external locus of control (Spector, 1982).

 

Individuals who have an internal locus of control u.
may actually seek situations in which control is possible"
(Spector, 1982, p. 483). Therefore, a possible research
question might be that one would expect to find more
internal locus of control anglers at low toxic sites and
fewer internal locus of control anglers at high toxic sites.
This would follow because one would expect anglers with an
internal locus of control to seek out safer environments
while one would not expect the same from external locus of
control anglers.

Another explanation of angler behavior may simply be
that anglers ”adapt“ to the presence of toxic chemicals.
Adaptation is one of the alternative behaviors suggested by
Hewitt and Burton (1971) and may explain why even when
peOple were aware of the presence of toxic chemicals they
did not take any precautionary actions to minimize their
risk. "Apparently, we can hear so much about a hazard that
it no longer frightens us" (Bell et al, 1978, EL35). News
of toxic waste dumps, spills, the Superfund, the expense of

cleanup and the widespread nature of the problem is

107

constantly in the news. Fewer precautionary behaviors due
to an overabundance of information would have exactly the
Opposite effect as that desired by public agencies.

In addition to exploring different theoretical
frameworks, research is also needed to provide more accurate
estimates of the real threats posed by toxic chemicals in
fish. The extent of the threat should serve as a guideline
for the amount of managerial effort to reduce the risk to
the public.

The impact of toxic chemicals on the recreation
experiences of anglers also needs attention. The results of
this study show that the behavior of anglers can potentially
be modified by the presence of toxic chemicals (see pages
89—90). These behavioral modifications certainly change the
fishing experience. Whether this effect is positive,
negative or neutral is unknown.

Toxic chemicals may also affect other types of
recreationists. For instance, boaters or swimmers may be
adversely affected by the presence of toxic chemicals both
physically and psychologically.

Lastly, there is a need to determine what is a
reasonable balance between visitor safety, leisure freedom
and economic consequences. All three variables interact and

a change in any of the three can affect the other two.

108

Recommendations for Management of the Study Rivers

The MDNR is in a difficult situation. Any action or
inaction will affect one of the variables mentioned in the
triangle of variables above. Increasing management
protection efforts may have spillover effects on
uncontaminated rivers (witness the effects on uncontaminated
shellfish in the study done by Swartz and Strand (1981)).
More specifically, increased management protection will
change fishing experiences and some anglers will be
antagonized.

However, the situation on the Saginaw River should be
of particular concern. All fish are considered unsafe to
eat in any amount. 'Yet anglers are consuming fish at that
site. At the very least, the MDNR should post notices at
reasonable intervals along the Saginaw to increase the
perceptability of the hazard to anglers.

Due to the possibility of vandalism of these signs, it
may be necessary to occasionally send someone out in the
field to warn anglers about the potential threat from toxic
chemicals. Ideally, attractive alternatives could be
identified where the fish are safe to eat.

However, because of the convenience of fishing on the
Saginaw, stiff resistance should be expected. If the threat
is discovered to be of sufficient magnitude it may be
necessary to switch to hard management efforts that may

carry the power of law on the Saginaw. Regardless Of the

 

109

method tried, some people will undoubtably continue to fish.

On the Kalamazoo the problem is not as severe because
the warning is only for certain fish and for large
consumption levels (> 26 pounds per year (MDNR, 1984a)).
However, for some reason, anglers on the Kalamazoo are the
most likely to consume the fish they catch even when they
are aware that toxic chemicals are present. It was
speculated that this might be due to their greater
commitment to the sites studied. If this is the case then
increased management efforts may be necessary to counteract
this commitment.

Efforts similar to those taken on the Saginaw might be
appropriate on the Kalamazoo. However, because of the more
dispersed nature of fishing on the Kalamazoo such efforts
would be more costly and difficult to implement. A higher
dependence upon increased signing rather than increased
patrols might be most efficient.

While the threat on the Kalamazoo is not considered to
be great, the idea that commitment may lead to an increased
liklihood of consumption needs further examination. Anglers
who travel a great distance or who spend a great deal of
money on fishing gear may be very committed to certain types
of fishing at certain locations (Bryan, 1979). This
commitment may act.as a barrier to accepting regulations

when toxic chemicals are discovered. This might be

110

particularly true for more pOpular fishing areas in the
northern part of the state and on the Great Lakes.

Lastly, the Grand River, where sampled, is considered
to have no hazardous levels of toxic chemicals. Yet, the
average scores on the perception scale for the Grand were
not significantly different from those of the other two
rivers. This implies that the Grand River is considered as
contaminated as both the Kalamazoo and the Saginaw. Such
beliefs may lead to unwanted avoidance behaviors. As
mentioned, Swartz and Strand (1981) found that imperfect
information may have the same consequences as correct
information. Care must be taken that waters that are not
contaminated are not confused with those waters which are
contaminated. The image of more than just the affected
river may suffer from the presence of toxic chemicals.
Further research might explore the effects of toxic chemical
contamination on the images of adjacent uncontaminated sites
as well as effects on regional and statewide images.

In summary, the two decision making frameworks
described early on in this study appear to be appropriate in
helping understand the differences in how management would
like anglers to behave and how anglers behave in reality.

No solution is offered here to resolving the differences.
However, the acknowledgement by management that people do
not make predictable decisions might lead to more productive

communication between managers of public lands and the

111

public.

Lastly, toxic chemicals pose a threat to the health of
anglers. The fact that this threat occurs in a recreation
setting does not diminish the threat. Further, with
continual improvements in our measuring techniques, toxic
chemicals will be discovered in more and more recreation
environments. The true sc0pe of the problem is only now
being discovered.

On one hand, the quality of life is degenerating and on
the other, the threat to life is increasing. At some point,
'the trends must be reversed. Managerial actions to protect
peOple from hazards created by people are only patchwork
efforts. Until peOple stOp creating the hazards, there is

no far-reaching solution.

APPEND I CES

APPENDIX A

Sampling Locations

112

Samping Locations
(Summer, 1985)

 

River Location

Saginaw Veteran's Park, Saginaw County, between
Saginaw and Bay City

Saginaw East riverfront park areas extending from
south of downtown to city limits, city of
Saginaw

Kalamazoo Morrow Pond, Public Access Site, between
Galesburg and Comstock, Kalamazoo County

Kalamazoo Allegan Dam, west end of Allegan Lake,
Allegan County

Kalamazoo North and south of bridge, New Richmond,
Allegan County

Kalamazoo Hacklander Public Access Site, 2 miles east
of Douglas, Allegan County

Kalamazoo Schultz Park, City of Douglas

Grand Francis Park, riverfront area and boat launch
on Opposite side of river, City of Lansing

Grand Riverfront Park, City of Grand Rapids

APPENDIX B

Perception Scale Items

113

Perception Scale Items*

 

The WATER area I am fishing in now contains toxic chemicals.

The FISH in the water I am fishing in now contain toxic
chemicals.

The WATER I am fishing in now contains dangerous levels of
at least one toxic chemical.

It is highly unlikely that the FISH in the water I am
fishing contain any toxic chemicals.

Compared to other water areas I fish, the level of toxic
chemicals in this WATER area is low.

The FISH in the water area I am fishing in now contain
dangerous levels of toxic chemicals.

Toxic chemicals are not present in this WATER.

Compared to other water areas I fish the level of toxic
chemicals in the FISH in this water area is low.

There are large amounts of toxic chemicals present in this
WATER.

Responses ranged from strongly agree to strongly
disagree on a six point scale.

APPENDIX C

Questionnaire

114

 

 

 

 

 

 

 

 

WEATHER DATE
ANGLER TYPE LOCATION
TIME RIVER BANK
BOAT
1984

ANGLER STUDY
Michigan State University
Park and Recreation Resources
FINAL VERSION

 

Excuse me. I'm from Michigan State University and I'm doing
a study on the perception of Michigan waterways by anglers.
Can I take a few minutes of your time to ask you some
questions?..... Thank you! First of all I need to tell you
that.I'm from the Park and Recreation Resources department
at Michigan State. If you have any questions or concerns
about what weflre asking please feel free to call the
department. Also, any information I collect will be kept
completely confidential. That means that we will use the
data we collect as a group, but we vmnflt be publishing any
one's questionnaire by itself. There will be no way that
anyone will be able to find out what you answered on the
questionnaire.

First we need some information about you as an angler.
1) When you fish, do you usually fish for a specific type

of fish?
yes no

 

 

If yes, can you tell me what species of fish?

 

2) What type of fish would you most like to catch today?
[NEED ONE TYPE OF FISH HERE FOR THE ANSWER

 

3) How many days each year do you fish?

 

4) How many days each year do you fish at this site.

 

5)

6)

7)

8)

9)

10)

115

Could you tell me how important it is to you to catch a
fish today?

extremely very moderately not very not at all
important important important important important

Could you give me an estimate on how much you have
invested in your fishing equipment? [total, not
including boat, if you own one]

$

 

Do you own a boat? yes no

 

If yes, what percentage of time do you estimate you use
your boat for fishing?

 

If yes, would you tell us how much your boat cost?
[total including motor and accessories]

$

 

How many times each year do go on overnight trips just
to fish?

 

When you go on vacation trips, would you say that
fishing is: (Please check one) [Ask until yes response]

the ONLY recreation activity you participate in

yes no

the MAIN recreation activity you participate in

yes no

not GENERALLY an activity you participate in
yes no

NEVER an activity you participate in
yes no

Thinking about your free time, what percentage of that
time do you spend fishing each year?

 

ll)

12)

13)

116

When you fish, who do you usually go with? (check one)
go alone
take (or go with) the family
go with friends
family and friends

other, please specify

 

 

Do you plan on eating any of the fish you catch today?

yes no depends on what I catch

 

 

If yes, can you tell me which types of fish you will be
eating?

 

 

If no, why not?

 

If no, what do you do with the fish you catch?

 

If depends, what type of fish would you eat?

 

Can you give me an estimate, in pounds, of the amount
of fish you catch and eat each year at this site.

 

If give an estimate, what type of fish is it that you
usually catch and eat?

 

14)

15)

16)

117

If you caught the fish you most wanted to catch today,
how would you prepare these fish to cook? [REFER TO
QUESTION 2 IF NEED A REMINDER].

I don't eat fish

skin the fish before cooking

scale the fish but leave the skin on

remove belly flap

fillet fish

other (please explain)

 

 

How do you usually cook these fish?

I don't eat fish _____ broil
_____ cook the fish whole _____ fry
_____ poach _____ smoke
eat raw

other (please explain)

 

 

At this point we would like you to think about the land
area surrounding the water you are fishing in today.

On a scale of one to ten with one being UGLY and ten being
BEAUTIFUL, with the numbers in between being somewhere
between ugly and beautiful, what number would you give this
land area? ,

 

With one being MAN-MADE and ten being NATURAL how would you
rate this land area?

118

With one being DIRTY and ten being CLEAN how would you rate
this land area?

 

Now we would like you to think about the WATER in front of
you.

On a scale of one to ten with one being MUDDY and ten being
CLEAR with the numbers in between being somewhere between
MUDDY and CLEAR, what number would you give the water?

 

With one being DIRTY and ten being CLEAN how would you rate
this water?

 

With one being FOUL and ten being FRAGRANT how would you
rate this water?

 

With one being STALE and ten being FRESH how would you rate
this water?

 

Lastly, in your estimation, on a scale of one to ten, with
one being POOR and ten being EXCELLENT, how would you rate
the overall quality of the water?

 

17) Would you please tell me what you feel is the worst
problem affecting the water quality Of this river?

 

 

119

I'm now going to read you a series of statements that all
relate to toxic chemicals. Toxic chemicals are such things
as PCB's, Lead or Dioxins that are sometimes found in the
every day environment. They all sound pretty close to each
other in meaning so please ask me to repeat them as often as
you need. I'll give you a card to help you remember the
responses. [GIVE CARD TO PERSON]. I'm going to ask you if
you strongly, loderately or slightly agree or disagree with
each statement. There are NO right or wrong answers for
these statements. We are very interested in YOUR opinions.
[ENCOURAGE PEOPLE TO RESPOND TO THESE QUESTIONS EVEN IF THEY
FEEL THEY ARE NOT FAMILIAR WITH TOXIC CHEMICALS. WE WILL
ASK THEM HOW FAMILIAR THEY FEEL THEY ARE WITH TOXIC
CHEMICALS LATER SO THEY SHOULDN'T WORRY ABOUT HURTING OUR
RESULTS] .

The WATER area I am fishing in now contains toxic
chemicals.

Strongly Moderately Slightly Slightly Moderately Strongly
Agree Agree Agree Disagree Disagree Disagree

It is not likely that I would get ill from consuming toxic
chemicals.

Strongly Moderately Slightly Slightly Moderately Strongly
Agree Agree Agree Disagree Disagree Disagree

The FISH in the water I am fishing in now contain toxic
chemicals.

Strongly Moderately Slightly Slightly Moderately Strongly
Agree Agree Agree Disagree Disagree Disagree

The WATER I am fishing in now contains dangerous levels of
at least one toxic chemical.

Strongly Moderately Slightly Slightly Moderately Strongly
Agree Agree Agree Disagree Disagree Disagree

It is highly unlikely that the FISH in the water I am
fishing in contain any toxic chemicals.

Strongly Moderately Slightly Slightly Moderately Strongly
Agree Agree Agree Disagree Disagree Disagree

120
Consuming toxic chemicals in any amounts will make a person
seriously ill.

Strongly Moderately Slightly Slightly Moderately Strongly
Agree Agree Agree Disagree Disagree Disagree

Compared to other water areas I fish, the level of toxic
chemicals in this WATER area is low.

Strongly Moderately Slightly Slightly Moderately Strongly
Agree Agree Agree Disagree Disagree Disagree

The FISH in the water area I am fishing in now contain
dangerous levels of toxic chemicals.

Strongly Moderately Slightly Slightly Moderately Strongly
Agree Agree Agree Disagree Disagree Disagree

Toxic chemicals are not present in this WATER.

Strongly Moderately Slightly Slightly Moderately Strongly
Agree Agree Agree Disagree Disagree Disagree

Compared to other water areas I fish the level of toxic
chemicals in the FISH in this water area is low.

Strongly Moderately Slightly Slightly Moderately Strongly
Agree Agree Agree Disagree Disagree Disagree

It is possible to get sick from eating fish that contain
toxic chemicals.

Strongly Moderately Slightly Slightly Moderately Strongly
Agree Agree Agree Disagree Disagree Disagree

There are large amounts of toxic chemicals present in this
WATER.

Strongly Moderately Slightly Slightly Moderately Strongly
Agree Agree Agree Disagree Disagree Disagree

FISH that contain toxic chemicals are always unsafe to eat.

Strongly Moderately Slightly Slightly Moderately Strongly
Agree Agree Agree Disagree Disagree Disagree

In general, the quality of the environment is improving.

Strongly Moderately Slightly Slightly Moderately Strongly
Agree Agree Agree Disagree Disagree Disagree

121
18) How familiar would you say you are with toxic
chemicals? [READ RESPONSES]

Extremely Very Somewhat Not very Not at all
Familiar Familiar Familiar Familiar Familiar

Getting away from those types of questions...

19) Have you or anyone you know been made sick from fish
you believe were contaminated by toxic chemicals?

yes no don't remember

 

 

If yes, how many times has this happened to you or
someone you know?

 

20) How would you determine if toxic chemicals were present
in the environment?

 

 

21) If toxic chemicals were present in the fish you were
nost.trying to catch today, would you keep those fish?
[REFER TO QUESTION 2 IF A REMINDER IS NEEDED]

yes no depends

 

 

If yes, would you eat those fish?

yes no

 

 

If yes, how would you prepare those fish to cook?
I don't eat fish
skin the fish before cooking
scale the fish but leave the skin on
remove belly flap
fillet fish

other (please explain)

 

 

22)

23)

122

If yes, how would you cook those fish?

I don't eat fish _____ broil
_____ cook the fish whole _____ fry
_____ poach _____ smoke
eat raw

other (please explain)

 

 

If no, what will you do with those fish?

 

If depends, what type of fish would you keep? [ALSO ASK
THE QUESTIONS FOR IF ANSWERED YES]

 

If you wanted or needed to find out about the possible
dangers of consuming toxic chemicals where would you
turn to find out this information?

 

If you were trying to avoid fish that contained toxic
chemicals at this site, which types of fish would you
avoid?

 

 

I'd like you to make one more rating now on the presence of
toxic chemicals in this water.

24)

25)

With one being NO TOXIC CHEMICALS are present in this
water and ten being HIGH LEVELS OF TOXIC CHEMICALS are
present, how would you rate the level of toxic
chemicals present in this water?

 

Did you get a chance to read the brochure that was
given to you with your fishing license?

yes no didn't get one don't know

123

If yes, did you read the warning about the presence of
toxic chemicals in some fish in some Michigan water
areas?

yes no don't remember

 

Lastly, we would like to get some information about your
trip planning and about you yourself.

26)

27)

28)

29)

30)

During the past year have you ever avoided fishing in
waters in Michigan because you believed they were
contaminated by toxic chemicals?

yes no don't remember

 

 

If yes, can you tell us how many bodies Of water you
have avoided because you believed they were
contaminated by toxic chemicals?

 

Would you please name these bodies of water?

 

 

If you knew that toxic chemicals were present at this
site would this prevent you from fishing here?

yes no

 

 

If no, why not?

 

 

What is your home city, county, state and zip code?

 

What was the destination of the trip you are on?

 

Approximately how far is this site from your home?

miles

 

124

Lastly, we need some general information about yourself...

31)

32)

33)

34)

35)

36)

37)

Sex (check one)

 

male _____ female
Age
years
Race
Black ___ Hispanic ___ Oriental ___ White
Other

Marital status?
married single divorced

other (please specify)

 

What do you do for a living?

 

Can you tell me how many years of school you completed?

 

Would you please give an estimate of your annual family
income before taxes?

$

 

OPTIONAL (PILL OUT YOURSELF):

NUMBER OF FISH CAUGHT BY RESPONDENT

 

TYPES OF FISH CAUGHT BY RESPONDENT

 

 

YOUR OWN IMPRESSION ON HOW GOOD THE RIVER IS TO FISH TODAY

(i.e.

are peOple generally successful in catching fish

today?)

 

REFERENCES

125

REFERENCES

Bailey, KgD. Methods ef Social Research. New York: The
Free Press, 1978.

Baker, E.J., West, S.G., Moss, D.J. and Weyant, J.M.
Impact of Offshore Nuclear Power Plants: Forecasting
Visits to Nearby Beaches. Environment and Behavior.
1980, 12(3), PP. 367-407.

Barker, MIL. IBeach Pollution in the Toronto Region. In
W.R.D. Sewell and 1. Burton (eds), Perceptions and
Attitudes i2 Resource Management. Ottawa: Dept. of
Energy, Mines and Resources, 1970, pp. 37-47.

 

Baum, A., Fleming, R. and Davidson, IuM. Natural Disaster
and Technological Catastrophe. Environment and
Behavior. 15(3),1983,;un 333-354.

 

Bell, P.A., Fisher, J.D. and Loomis, R.J. Environmental
Psychology. Philadelphia: W.B. Saunders Company, 1978.

 

Blalock, H.M. Social Statistics. New York: McGraw-Hill
Book Company, 1979.

 

Blanchard, J.J. Michigan State ef the State Message.
Lansing: Michigan Office of the Governor, 1985.

 

Brown, IhJ., Driver, BEL. and McConnell. The Opportunity
Spectrum Concept and Behavioral Information in Outdoor
Recreation Resource Supply Inventories: Background and
Application. In H.G. Lund, V.J. Labau, P.F. Ffolliott
and ELW. Robinson (eds), Integrated Inventories e;
Renewable Resources: Proceedings e: the Workshop.

USDA Forest Service General Technical Report RM-SS,
January, 1978, pp. 73-84.

Bryan, H. Conflict in the Great Outdoors. Birmingham: The
Birmingham PubliShing Company, 1979.

 

126

Burton, 1., Kates, R.W. and White, G.F. The Environment ee
Hazard. New York: Oxford Press, 1978.

 

Burton, 1. Cultural and Personality Variables in the
Perception of Natural Hazards. In J.F. Wohlwill and
ILH. Carson (eds), Environment and the Social
Sciences: Perspectives and Applications. Washington
ILC.: American Psychological Association, Inc., 1972,
pp.184-197.

 

 

 

Cheron, ELJ. and Ritchie, JJLB. Leisure Activities and
Perceived Risks. Journal ef Leisure Research. 1982,

 

Clark, ILN. and Stankey, CLH. The Recreation Opportunity
Spectrum: e_Framework for Planning! Management, and
Research. USDA Forest Service, General Technical
Report PNW-98, December, 1979.

 

 

 

Collins, RJL The Strategy for Cleaning Up Our Waters. In
T;Y. Toribara et al. (eds), Environmental Pollutants.
New York: Plenum Press, 1978, pp. 19-35.

 

Crabb, J. Statistics Consultation. Department of
Statistics. Kansas State University. Manhattan, KS,
9/17/85.

Drew, CAI. Introduction to Designing and Conducting

Research. St. Louis: The (AV. Mosby Company, 1980.

 

 

Driver, BJL and Brown, PJL The Opportunity Spectrum
Concept and Behavioral Information in Outdoor
Recreation Resource Supply Inventories: A Rationale.
In H.G. Lund, V.J. Labau, P.F. Ffolliott and D.W.
Robinson (eds), Integrated Inventories 9; Renewable
Resources: Proceedings ef the Workshop. USDA Forest
Service General Technical Report RM-55, January, 1978,
pp. 24-31.

 

 

 

 

Driver, B.L. and KnOpf, R.C. Temporary Escape: One Product
of Sport Fisheries Management. Fisheries. 1976,
1(2): pp. 21-29.

 

Eckholm, E.P. Qege _Te Earth. New York: W. W. Norton & Co.,
1982.

Einhorn, ILJ. and Hogarth, R.M. Behavioral Decision Theory:
Processes of Judgement and Choice. Annual Review efi
Psychology. Palo Alto: Annual Reviews, Inc, 1981,

Vol. 32, pp. 53-88.

 

 

127

Evans, G.W. and Jacobs, S.V. Air Pollution and Human
Behavior. Journal ef Social Issues. 1981, 37(1), pp.
95-125.

 

Festinger, L. A Theory 9; Cognitive Dissonance. Stanford,
CA: Stanford University Press, 1957.

 

Gunter, ELG. Properties of the Leisure Experience. In
Ibrahim, H. and Crandall, R. (eds) Leisure: 5
Psychological Approach. Los Alamitos: Hwong Publishing
Company, 1979, pp. 3-43.

 

 

Haber, ILN. and Hershenson. The Psychology ef Visual
Perception. New York: Holt, Rinehart and Winston,
Inc., 1973.

 

 

Hanson, S., Vitek, J.P. and Hanson, P.O. Natural Disaster:
Long-Range Impact on Human Response to Future Disaster
Threats. Environment and Behavior, 1979, 11(2), 268-
284.

 

 

Hansson, R.C., Noulles, D. and Bellovich, S.J. Knowledge,
Warning and Stress: A Study of Comparative Roles in an
Urban Floodplain. Environment and Behavior, 1982,
14(2),171-185.

 

Harper, W. The Experience of Leisure. Leisure Sciences.

Hendee, J.C., Clark, R.N. and Dailey, T.E. Fishing and
Other Recreation Behavior e3 High-Mountain Lakes :e
Washington State. USDA Forest Service, Pacific North
West Forest and Range Experiment PNW-304, November,
1977.

 

 

 

Hesse, J. L. Polychlorinated Biphenyl Usage and Sources of
Loss to the Environment in Michigan. In National
Conference e2 Polychlorinated Biphenyls. Environmental
Protection Agency, Washington D.C., March, 1976,
pp. 127-123.

 

 

Hewitt, K. and Burton, 1. The Hazardousness EE‘E Place: A
Regional Ecology ef Damaging Events. Toronto:
University of Toronto Press, 1971.

 

 

 

Holdgate, M.W.; Kassas, M. and White, G.G. (eds), The World
Environment Programnmh Dublin: Tycooly International
Publishing Limited, 1982.

 

 

 

128

Jackson, EJL. Response to Earthquake Hazard: The West Coast
of North America. Environment and Behavior, 1981,
13(4) , PP. 387-416.

 

James, G.A., Nelson, W.T. and Hopkins, M.L. Estimating
Recreational Use of a Unique Trout Stream in the
Coastal Plains of— South Carolina. USDA Forest Service
Research Note SE - 159, May, 1971.

 

 

 

 

 

Jubenville, A. Outdoor Recreation Management.
Philadelphia: VLB. Saunders Company, 1978.

 

Kopp, T. E. PCB Disposal, Reclaiming, and Treatment. In
National Conference on Polychlorinated Biphenyls.
Environmental Protection Agency, Washington D.CL,
March, 1976, pp. 108-123.

 

 

Kolers, IhA. Perception and Representation. Annual Review
ef Psychology. Palo Alto: Annual Review, Inc.,
Vol. 34, 1983, pp. 129-166.

 

 

Levy-Leboyer, C. Psychology and Environment. Beverly
Hills: SAGE Publications, Inc., 1982.

 

 

Logue, J.N., Melick, M.E. and Hansen, H. Research Issues
and Directions in the Epidemiology of Health Effects of
Disasters. Epidemiologic Reviews. 1981, 3, pp. 140-
162.

 

Lucas, TLC. Recreation Regulations - When Are They Needed?
Journal 2: Forestry. 1982, 80(3), pp. 148-151.

 

Mahoney, E., Jester, D. and Kikuchi, H. Fisheries Report.
Michigan Department Of Natural Resources, Fisheries
Division, (in preparation).

 

Manfredo, M.J., Brown, P.J. and Haas, G.E. Fisherman
Values in Wilderness. In Proceedings of the Western
Association of Fish and Wildlife Agencies and the
Western Division American Fisheries Society. San
Diego, July, 1978, pp. 276- 297.

 

 

 

 

 

Marty, R. Administrating Public Lands and Resources.
Unpublished manuscript, Michigan State University,
Department of Forestry, 1982.

 

Michigan Department of Natural Resources. Michigan Fishipg
Guide: 1983. Lansing: State of Michigan, 1984a.

 

129

Michigan Department of Natural Resources. Hazardous Waste
Management ie Michigan. Lansing: State of Michigan,
1984b

 

 

Mieure, J. P., Hicks, 0., Kaley, R. G., and Saeger, V. W.
Characterization of Polychlorinated Biphenyls. In
National Conference ep Polychlorinated Biphenyls.
Environmental Protection Agency, Washington DJL,
March, 1976, pp. 84-93.

 

Mitchell, J.K. Natural Hazards Research. In I.R. Manners
and M.W. Mikesell (Eds), Perspectives ee Environment.
Washington DJL: Association of American Geographers,
1974, pp. 311-341.

 

Moeller, GJL and Engelken, JJL What Fishermen Look For
in a Fishing Experience. The Journal ef Wildlife
Management. 1972, 38(4), pp. 1253-1257.

 

 

Moser. CLA. and Kalton, G. Survey Methods in Social

Investigation. New YOrk: Basic Books, Inc., 1972.

 

 

Neter, J. and Wasserman, W. Applied Linear Statistical
Models. Homewood, IL: Richard D. Irwin, Inc., 1974.

Nunnally, J4; Psychogetric Theory. New York: McGraw-Hill
Book Company, 1978.

 

CVMuircheartaigh, CLA. Statistical Analysis in the Context
of Survey Research. In CJL (YMuircheartaigh and C.
Payne (eds), The Analysis ef Survey Data: Volume ;
Exploring Data Structures. New York: John Wiley &
Sons, 1978.

 

 

 

Ott, W. R. Environmental Indices: Theory and Practice.
Ann Arbor: Ann Arbor Science Publishers Inc., 1978.

Payne, J.W. Contingent Decision Behavior. Psychological
Bulletin. 92(2),1982,;un 382-402.

 

Preston, V., Taylor, S.M. and Hodge, D.C. Adjustment to
Natural and Technological Hazards: A Study of an Urban
Residential Community. Environment and Behavior.
15(2), 1983, pp. 143-164.

Rothbaum, F., Weistz, J.R. and Snyder, S.S. Changing the
World and Changing the Self: A Two-Process Model of
Perceived Control. Journal ef Pegsonality and Social
Psychology. 42(1),1982, pp.5-37.

 

 

130

Schweitzer, GJL Toxic Chemicals: Steps Toward Their
Evaluation and Control. In, Environmental Protection:
The International Dimension. Totowa: Allanheld, Osmun
& Co”,Publishers,Inc.,1983,;nx22-24.

 

Shippee, G., Burroughs, J. and wakefield. Dissonance Theory
Revisited: Perception of Environmental Hazards in
Residential Areas. Environment and Behavior. 12(1),
1980, pp.33-51.

Shulstad, R.N. and Stoevener, H.H. The Effects of Mercury
Contamination in Pheasants on the Value of Pheasant
Hunting in Oregon. Land Economics. 1978, 54(1),
pp.39-49.

 

Sims, JJL and Baumann, DJL Educational Programs and Human
Response to Natural Hazards. Environment and Behavior.
15(2), 1983, pp. 165-189.

Slovic, P., Kunreuther, H. and White, CLF. Decision
Processes, Rationality and Adjustments to Natural
Hazards. In Gilbert F. White (ed) Natural Hazards:
Local, National, Global. New York: Oxford Press, 1974,
pp.187-205.

 

 

Sorensen, J.H. Knowing How to Behave Under the Threat of
Disaster. Environment and Behavior. 15(4), 1983,
pp. 438-439.

 

Spector, PJL Behavior in Organizations as a Function of
Employee's Locus of Control. Psychological Bulletin.

Stynes, D.J., Spotts, D.M. and Strunk, J.R. Relaxing
Assumptions of Perfect Information in Park Visitation
Models. Professional Geographer. 37(1), 1985, pp521-
28.

 

Sudman, S. Applied Sampling. New York: Academic Press,
1975.

 

Swartz, ILG. and Strand, I.E. Avoidance Costs Associated
with Imperfect Information: The Case of Kepone. Land
Economics. 1981, 57(2), FmL139-150.

 

Svenson, O. and Fischhoff, B. Levels of Environmental
Decisions. Journal ef Environmental Psychology. 1985,
5 (1) I 1313-55-68.

131

Train, R. B. Environmental Protection Prescription for
Health. In National Conference ee Polychloriated
Biphenyls. Environmental Protection Agency, Washington
D.C., March, 1976, pp.7-9.

 

Udd, E. and Fridgen, J.D. Anglers' Perceptions of Toxic
Chemicals in Rivers and Sports Fish. Paper to be
published in the Proceedings 9: the National Wilderness
Conference. Fort Collins, CO, July 22—25, 1985.

 

 

White, GJL Natural Hazards Research: Concepts, Methods,
and Policy Implications. 2n1CLF. White (ed) Natural
Hazards: Local, National, Global. New York: Oxford
University Press, 1974, pp. 3-16.

Wilford, W. A., Hesselberg, R. J. and Nicholson. Trends of
Polychlorinated Biphenyls in three Lake Michigan
Fishes. In National Conference ee Polychlorinated
Biphenyls. Environmental Protection Agency, Washington
D.C., March, 1976, pp. 177-181. ‘