Effects of Air Pollution on
Public Attitudes and Knowledge
Oregon State University
Corvallis
June 1972
-------
ACKNOWLEDGMENTS
This publication is the final report of a three-year study which was
conducted under the terms of contract number CPA 70^-117, Air Pollution
Control Office, Environmental Protection Agency. The research began
May 23, 1969 and terminated June 30, 1972.
Many people have contributed to the completion of various phases
of this work. Pamela Bodenroeder joined the project in September 1969
as assistant leader and has worked tirelessly on all phases of the study.
Assisting her from time to time were Dorothy Bowers, Betty Lou Brooks,
Diana Coulombe, Alice Doyle, Julia Fisher, Ann Ginn, Gay Greger, Alice
Henderson, Linda Henderson, Kathleen Hulburt, Margaret Kent, Cheryl Miller,
Kathleen O'Dwyer, Penny Schmidt, Lora Seleen, Gretchen Wagner, Debora
Wahlquist and Nancy Young. Marian Richardson and Joyce Driscoll completed
many of the arduous typing and other secretarial tasks.
The nature of the research problem interested professionals from
different disciplines and the author greatly profited from an association
with many who contributed to his understanding of the complexity and subtlety
of air pollution processes and effects. A number of individuals at Oregon
State University generously gave of their time to advise on the many
technical aspects of this study. These were R. M. Alexander, Air and Water
Resources Center; Richard Boubel, Mechanical Engineering; William G. Brown,
Albert N. Halter and Herbert H. Stoevener, Agricultural Economics; Lyle
Calvin and Norbert Hartmann, Statistics; David Chilcote, Farm Crops; and
E. Wendell Hewson and Wesley Tufts, Atmospheric Sciences. As well,
Richard E. Hatchard, Columbia-Willamette Air Pollution Authority; Michael
D. Roach, Mid-Willamette Air Pollution Authority; and Thomas E. Waddell,
Environmental Protection Agency, provided most useful professional help
and advice.
-------
EFFECTS OF AIR POLLUTION ON
PUBLIC ATTITUDES AND KNOWLEDGE
Robert G. Mason
Professor of Sociology
Oregon State University
June 1972
-------
CONTENTS
Acknowledgments i
Summary, Conclusions and Recommendations 1
Introduction 9
Theoretical Framework for Public Attitudes and Knowledge 10
Public Attitudes and Governmental Action
Environmental Quality as a Public Issue
Enforcement and Community Conflict
Attitudes and Definition of the Situation
Definition of Air Pollution Situations
An Air Quality Communication Model
Methods--Content Analysis and Attitude Change 29
Selection of Location and Time
Development of Operational Measures
Data for Hypothesis Testing
Analysis of Data
Results and Discussion--Content Analysis and Attitude Change 40
Results of the Content Analysis
Analysis of Visibility Observations
Analysis of Attitude Change Scores
Discussion of Results
Methods--Air Quality Communication Model 53
Development and Evaluation of Operational Measures
Execution of the Sample Survey
Analysis of Data
Results and Discussion--A Test of the Air Quality Communication Model . 67
Results of Accuracy of Air Pollution Perceptions
Empirical Test of the Communication Model
Empirical Test of the Reformulated Communication Model
Discussion of Results
Literature Cited 90
Appendix A: Review of Literature 95
Appendix B: Interview Schedules A and B 101
Appendix G: Frequency Curves of Clean Air Themes 122
Appendix D: Verbatim Responses of Individuals Holding Accurate
or Inaccurate Visibility Perceptions. 136
-------
SUMMARY, CONCLUSIONS AND RECOMMENDATIONS
The purpose of this research was twofold: First, a test was made of a
hypothesis concerning the relationship between mass media content and attitude
change. More specifically, the relationship stemmed from the hypothesis
that perceived changes in air quality in a community are defined or redefined,
in part, by the local mass media. As well, a definition or redefinition in the
media is sufficient for people to change their attitudes toward the topic.
Second, a test was made of a theoretical model concerning the communication
of air quality information to the public. Specifically, the model stipulated
that three variables were mutually interdependent and were considered
endogenous. These variables were attitude toward air quality, mass media
exposure to information about air quality and level of air quality knowledge.
Variables considered exogenous were levels of interpersonal discussion,
social influence, levels of formal and science education and use of the environ-
ment for recreation purposes.
An empirical test of the mass media content - attitude change hypothesis
involved a content analysis of two daily newspapers, one each from two
communities. A panel design was employed to measure attitude change of a
random sample of residents in the two communities. These communities
were selected on the basis of known changes in air quality levels during the
late summer. The communities were Eugene and Salem, Oregon. Because
of its location at the base of the Willamette Valley, Eugene has experienced
poor natural ventilation whereas Salem, located in the mid-Willamette Valley,
has had good natural ventilation. Therefore, a contrast in visibility would be
expected to occur when there is air pollution in the Valley. Personal inter-
views (for measuring attitudes) were timed in 1970 and in early 1971 to take
advantage of this known change in air quality. The source of air pollution
was agricultural field burning and the measure of air quality employed was
level of visibility. Visibility observations were completed during interviewing
periods to determine if changes in visibility occurred as anticipated. The
-------
recording unit for the content analysis was the theme and the unit of analysis
was the article. Fifteen themes were developed and pretested. A total of
3, 054 newspaper articles from an 11-year period (1960-1970) was analyzed.
Frequency of themes, type of article and source of pollution mentioned were
the data employed for determining if definition or redefinition had occurred.
Attitudes toward a'.r quality were measured by interviewing a random sample
of residents in each community twice, once when visibility was high and once
when it was low. Adjusted attitude change scores were employed to determine
if a change in air quality was associated with an observed change in attitude.
Results of the panel, interview indicated that observed changes in attitudes
were produced primarily by errors of measurement associated with attitude
change scores and not by changes in air quality. Air quality did, however,
change in the anticipated directions, accorded to analysis of visibility
observations.
Results of the mass media analysis suggested that the air quality situation
in Eugene was defined in late 1966 or early 1967; in Salem, definition likely
occurred in August 1969- An air pollution episode occurred in Eugene in
August 1969. In response, the state governor banned field burning for one
week and placed control of future burning in the hands of a state environmental
control agency. This action may well have contributed to defining the situation
for Salem and for the entire state as well. One can conclude that the findings
are weak, but positive, support for the definitional hypothesis concerning the
relationship between mass media content and attitude change.
An empirical test of the communication model involved a personal inter-
view survey of a random sample of individuals in the two communities
employed for testing the mass media - attitude change hypothesis - Fugene
and Salem, Oregon. Operational measures of variables in the theoretical
model were based on summated response scores of items asked in a standard-
ized interview schedule. Items employed for each variable were extensively
pretested in a third community and subjected to an item analysis. Reliability
estimates were also calculated. The personal interview survey was completed
-------
in 1970 and in c:arly J971 within planned time- limits. Completion rates were
well within acceptable levels. Significance of variables in the model was
tested by two-stage least squares.
Results of the test of the communication model showed an error of specifi-
cation in the model. One exogenous variable, interpersonal discussion, should
have been considered endogenous. A reformulated model was developed to
include this consideiation. A new exogenous variable, belief that air pollution
was serious in the community, was postulated exogenous to level of attitude.
It was also anticipated that different models may be operating for individuals
who accurately or inaccurately perceived changes in visibility which had
actually occurred. A test for pooling data of accurate and inaccurate sub-
groups disclosed that data could not be pooled for the two subgroups in Salem.
The model was analyzed separately for each subgroup. For Eugene, the
community with an air pollution episode in 1969, a confounding of the per-
ception measure was discovered which precluded attempts to pool data on
the basis of accuracy of perceptions. A relatively large number of respond-
ents compared pollution levels of 1970 with those of 1969, rather than with
the previous interviewing periods in the panel. The communication model
was tested in three instances. One was for Eugene respondents. Another
was for respondents who held inaccurate perceptions in Salem. A third was
for respondents also in Salem who held accurate perceptions.
One can conclude that different models were functioning in each instance
and suggest that each model may represent different stages of the same.
underlying communication process. Initially, when changes in air quality
are not likely to impinge on the individual's consciousness, attitudes are
Data from the previous model were used to test the reformulated model.
This test is less meaningful than if new data had been acquired since the
equations were developed after data from the previous model had been
examined. Such selection tends to capitalize on relationships which might be
high because of sampling fluctuations. Thus, caution should be employed in
viewing results of tests of the reformulated model.
-------
4
formed primarily by interpersonal discussion and on relatively low levels of
knowledge. Selectivity in mass media exposure and interpersonal discussion
is likely to be operating. As well, knowledge levels are formed partly on
this selective mass media exposure. Later, when changes in air quality do
impinge on the individual's consciousness, a lack of sufficient information
in the channels of communication is likely. Those with high knowledge levels
are likely to seek out information from both local mass media and inter-
personal sources. Selectivity in information exposure has been abandoned
and attitudes are based primarily on increased levels of knowledge. Finally,
after a continued period of exposure to deteriorated air quality, an inter-
dependence has been established between mass media exposure and level of
knowledge. A sufficiency of information is available in the local mass media
and knowledgeable individuals seek out and use this information. Level of
knowledge is an important basis for determining level of attitudes which, in
turn, is a predictor for level of interpersonal discussion. Those who are
involved in interpersonal discussion are also greater users of the mass media.
The effects of changes in air quality appear to be subtle, cumulative but
nonetheless real, when studied in terms of cumulation of knowledge and
attitude formation. The channels of communication play different roles at
each stage of the communication process. An understanding of these roles
should enhance the efforts of those interested in speeding up the formation
of positive attitudes and the level of knowledge of people toward this topic.
The effects of exogenous variables -- social influence, science education,
use of the environment for recreation and belief that air pollution is a serious
community problem -- occurred in their predicted direction for all models.
One effect of exogenous variables noted was to determine the percent change
of each exogenous variable on the percent change in attitudes. Results showed
that the social influence variable would produce the greatest change in air
quality attitudes for Eugene respondents and for Salem respondents who held
accurate air pollution perceptions. In these instances a 1% change in social
influence would change attitudes toward air quality by 10% in Eugene and by 7%
in Salem. Effects of ocher exogenous variables were small. However, a
-------
change in use of the environment for recreation was by far the greatest pro-
ducer of attitude change among Salem respondents who held inaccurate air
pollution perceptions. A 1% change in use of the environment would produce
a 16% change in attitudes. Efforts to sensitize people who hold inaccurate
air pollution perceptions should include attempts to persuade them to use the
environment more for recreation. This practice is likely to produce more
positive attitudes toward air quality and, perhaps, a more accurate perception
of visibility restrictions. It is concluded that it would be difficult to produce
more favorable air quality attitudes among respondents holding accurate than
among respondents holding inaccurate visibility perceptions. Social influence
is considered a more difficult attribute to change than it is to persuade an
individual to use the environment more for recreation.
Moreover, social influence and interpersonal discussion concerning air
pollution play a rcle in attitude formation, primarily at the initial stage of
the communication process when perceptions are not likely to be accurate.
The importance of interpersonal discussion is likely to be reduced once
knowledge levels and attitudes are well established and after perceptions
concerning air qualify become accurate. As well, the positive relationship
between science education and knowledge, found at all stages of the communi-
cation process, suggests that a knowledge gap exists between the highly- and
poorly-educated segments of the community. Over time, this gap likely will
become larger, not smaller. Thus, any program to employ the mass media
to reach large numbers of people with information concerning air pollution is
likely to reach the already well-informed. These people will be the primary
beneficiaries of such an effort. Finally, the belief that air pollution is
serious -- based on the individual's experience living or working in the
community -- appears to have a bearing on attitudes at all stages of the
communication process. A pre-condition for the establishment of a successful
air pollution control program is that the citizens of an airshed recognize
the scope and magnitude of the problem. This pre-condition is satisfied when
they perceive odors and loss of visibility, for example, for what they are and
in terms which they believe are important to them.
-------
It seems reasonable to set down four recommendations concerning tht;
role of public attitudes and knowledge in air pollution control efforts:
1. Air quality goals and standards should be established on the basis
of their social feasibility as well as their economic and technical feasibility.
The establishment of goals and standards is defined as the concentration
of pollutants which the public believes it can accept without adversely affecting
its health or welfare. Few will argue the relevancy of economic or techno-
logical criteria. But, it is argued here that public acceptance of air pollution
goals and standards implies the existence of a social definition of air quality
which merits consideration as well. This social definition includes the role
of public attitudes and knowledge. It is not argued that air quality goals and
standards should mirror existing public attitudes or knowledge of the time.
Rather, public concerns should be considered in good faith prior to the
establishment of goals and standards. In many instances the public hearing
requirement for implementation of the Clean Air Act of 1970 may well be
sufficient, or result in, the "good faith" requirement stipulated above. In
other instances, it may well be that public concerns would be more adequately
considered by alternative methods.
2. Emission control programs should be based, in part, on the existence
of an informed and supportive public.
The technology associated with emission control programs is likely to
become more, rather than less, complex. The public is likely to be exposed
to a number of new ideas constantly and, if accepted, some of these new ideas
may help while some may hinder the improvement of control programs.
Professionals who administer emission control programs should be prepared
to state their cases clearly and publicly and, if possible, to secure public
support for the changes they wish to make and for those they wish to ignore.
An informed public does not necessarily mean a supportive one; yet, it is
difficult to imagine a supportive public that is totally uninformed or is mis-
informed. It is quite likely that young people entering the adult population
-------
are far mon- sophisticated in comprehension of the technical aspects of air
pollution, for example, than older citizens. Greater sophistication of the young
likely stems from mere intensive formal science education and from greater
attention to environmental matters in the mass media, particularly news-
papers. This demographic implication should not be ignored by professionals
in charge of emission control programs in their planning of information and
education activities.
3. Air pollution control authorities should become skilled in the use
of applied behavioral, science research in carrying out their emission control
programs.
An objective assessment of the strengths and weaknesses of emission
control programs is considered a necessary condition for the subsequent
improvement or for the survival of the program itself. Assume that a
social definition, including public attitudes and knowledge, of control efforts
is an important consideration. It follows that this definition should be studied,
analyzed and interpreted to establish the reference points for coping with the
public issues which are likely to develop over the implementation of future
control programs. It is here that well-planned and professionally conducted
objective assessments of public attitudes, for example, will have a bearing.
Professionals who administer emission control programs, among others,
should become familiar with the strengths and weaknesses of public opinion
surveys, for example. These professionals should be able to determine
what such surveys can do and cannot do and where they most effectively play
a role in evaluating an emission control program. How effectively they are
used depends on the insight, skill and judgment of individual administrators.
As well, these administrators should have sufficient "back-up" expertise
available to evaluate the results of public attitude surveys which may be
profferred by some interest group. The quality of such surveys may render
the results specious or genuine, for example.
-------
4. Additional research should be conducted to provide a more complete
understanding of human reactions and behavior toward air pollution situations
and control efforts.
An example is the gathering of additional data to verify the formulation
of the communication models developed in this study. Specifically, models
should be viewed as stages of an underlying communication process which
are embedded in a particular air pollution situation. The situation itself
should be studied caiefully prior to theory construction. The role of economic
benefits and externalities of a polluter as well as feasibility of controls
should be considered not only for the specification of variables in the model
but also for the selection of communities or situations employed for empirical
tests. In addition, the ability of pollutants to impinge on the consciousness
of individuals should be examined prior to theory construction for the same
reasons. Finally, the model, or models, should be extended to other
pollution situations where the role of economic benefits, externalities,
feasibility of controls and ability of pollutants to impinge on the individual's
consciousness are likely to differ. With sufficient research one may
determine the generalizations required to explain some aspect of human
behavior toward the phenomenon of envi ronmental pollution.
-------
INTRODUCTION
The role of public knowledge and attitudes in the administration of air
pollution control policies and programs has received, at best, cursory
attention among professionals in the air quality field. One of the few to
recognize that these variables may indeed play a role is Schueneman [41] ,
who has pointed out the contribution of an informed and active public to a
successful emission control effort. The process of setting air quality
standards and enforcing emission restrictions has often relegated the under-
standing of people to a residual priority. More often, attention is focused
on the preparation of air quality criteria, goals and standards; on the
development of emission goals and standards; on the stipulation of design
standards; and, finally, on the establishment of measures of tests for
pollutants, equipment and the ambient air.
Granted that the establishment of standards is important, is highly
technical -- requiring a specialized competence -- it is assumed nevertheless
that, to be effective, these standards must be accepted with a modicum of
public understanding and support. Establishment of standards implies that
the decisions and actions taken by government to conserve (or not to conserve)
air quality are based on valid knowledge about people and involve a degree of
judgment about human behavior.
This report presents the results of testing a public information model
in which attitudes toward and knowledge about air quality play a role. In
this report, first, trie linkage between public attitudes and government action
is discussed. Seccnd, environmental conditions which may have a bearing
on variables in the model are considered. Finally, a theoretical model con-
cerning the acquisition and processing of air quality information, by people,
is formulated and tested.
Numbers in brackets correspond to literature citations, page 90.
-------
10
THEORETICAL FRAMEWORK FOR
PUBLIC ATTITUDES AND KNOWLEDGE
Public Attitudes and Governmental Action
It is perhaps axiomatic, as Key [23] has noted, that all governments must
concern themselves with public opinion; that they do not maintain their author-
ity without some willing acceptance and conformity from most of those whom
they govern. He considers popular government a special type in which one
base for authority is the seeking of popular opinion. Here, government gives
weight to the opinicns of its citizens in its decisions. Persons outside the
governmental establishment have a right to be heard. The legitimation of
the 'right to be heard1 becomes an ethical, if not a moral assumption under-
lying popular government and is operationalized into consistent patterns of
action by those who hold power. Thus, the linkage between public opinion
and governmental action rests not in a mirroring of opinion and action but in
the expectation that public preferences and mass interests are considered in
good faith. It is also assumed that a two-way flow of communication is estab-
lished between the individual and his government and from government to its
citizenry for either party to obtain consent --by conveying information or by
persuasion itself.
Moreover, as it seeks to form opinions favoring its policies, American
government enjoys no monopoly on information or access to persuasion.
Competing policies are urged on the public by a variety of sponsors -- special
interest groups, opposition political parties and competing centers of political
and economic power, for example.
American government may be characterized as functioning within some
vaguely defined limits of public opinion. This opinion is often unevenly dis-
tributed among people and its translation into support is often a matter of
conjecture at best. Still, governmental decisions must be and are made,
and the substance of action, in many instances, is related to the contours of
opinions that bear on an issue. Obviously opinion, in the sense of a mass
-------
11
opinion, haw no immediate bearing on many actions of government. These
actions represent a fait accompli simply because no general public opinion
about them exists. Actions on issues in which opinions are widely held are
another matter, for it is here that government often hesitates to act until the
weight of public opinion has been assessed. Even when it appears that a
majority favors new policies, government is hesitant to adopt a new course
of action on a major opinion-related issue. Information from reputable
pollsters, for example, that a majority favors a change in policy is neither
necessary nor sufficient to compel a political leader to change his position
on the issue. Reactions of people to new governmental actions are often
unpredictable and politicians understandably want to be sure that if they act
they will avoid widespread public hostility. Moreover, the social and
economic costs to those adversely affected by new actions may be sufficient
to guarantee inaction or delay.
To be sure, the appearance of a crisis may, at times, sweep political
leaders into some form of action. But the action is often inaction or delay
until the nature of the crisis is fully assessed. At other times, political
leaders act successfully on controversial matters with what appears to be
marginal public support. The rule, however, seems to be that major
governmental actions await development of predominantly favorable public
opinion.
It is in this vein that one might profitably examine environmental quality
as a public issue and attempt to extract some major variables that would be
of interest for study.
Environmental Quality as a Public Issue
Public concern about the environment is not new, as any veteran con-
servationist will tell you. Environmental quality as a public issue has waxed
and waned in America for more than 100 years. The issue has its historical
roots as a challenge to the widespread exploitation of the wilderness for
man's use. Nash [32] and Huth [20] have chronicled the adoption of a
-------
12
different relationship between man and nature by these environmentalists.
The dominant view expressed at the time (and still current) was that it is
man's right and duty tx> overcome nature for his own use; hence, emphasis
was placed upon the creation and use of technology. Early environmentalists,
on the other hand, organized a new ideology around the goal of preserving
tracts of land in their natural state for the esthetic, scientific or spiritual
values inherent in a wilderness preserve. The ideology implied restrictions
and cessations of commercial expansion. Modern environmentalists have
gone beyond the wilderness-saving goals of their forebears. A view has
been established that, as a consequence of the power created by science and
used by technology, the future of mankind in the modern world is in danger.
The value of scientific progress, technological change and economic growth,
for example, has been seriously questioned by Boulding [3] and Commoner [9],
to name a few. It is the codification and documentation of these two lines of
thought that have contributed to the current ideology of present-day ecologists.
4
Concurrent to the development of this new ideology has been the apparent
swift rise of environmental quality as a public issue. One heard almost no
mention of it five or six years ago in public forums. Today it ranks among
the top two or three issues facing the nation, according to public opinion
pollsters. This swift rise in public concern may stem from several sources.
First, mass media attention concerning the environment may well have played
a role in focusing the public's concern on this issue. One has only to count
the number of environmentally-related titles in the Readers' Guide to Periodical
Literature to note the striking increase in the number of these titles. More-
over, the number has increased positively each year, from 79 in 1950 to 228
in 1969. Admittedly, increased media attention does not guarantee increased
See The Roper Report [ 48] for an up-to-date summary of a national sample.
Problems of environmental quality ranked second in a survey of issues con-
fronting a. random sample of Oregon residents. The survey was conducted in
April 1972 by this author. Only problems concerning taxes and school financing
ranked higher. Fifty-eight percent of the respondents surveyed mentioned
environmental quality as one of the two or three most important problems
facing the state today. Sixty-one percent mentioned taxes and school financing.
-------
13
public concern. Yet, one sees convincing examples of mass media effects.
Tichenor _et ail. [ 52] have noted the impact of the publication of Rachel
Carson's Silent Spring [6.] on the subsequent pesticide controversy in the
early 1960's. Second, as Tichenor et aL [52], have noted, environmental
quality "reflects ideals which have been taught directly, specifically and
energetically in American schools in recent years". They assert that
messages concerning the environment have become important non-controversial
issues to young people. Compared to family life education, for example,
environmental quality has become a modern "motherhood" issue and far safer
for teachers to manage. Third has been the widespread activity of political
leaders on environmental matters. While one may argue that political action
is a reflection of public concern, the fact that political leaders have acted
with relative swiftness in attempting to cope with environmental problems has,
in itself, contributed to socially defining and legitimizing quality of the environ-
ment as an issue. The relative ease of passage of the Clean Air Acts in 1963,
1967 and 1970, for example, testifies to the support environmental quality
has among both conservative and liberal political viewpoints. Increased mass
media attention concerning political action about environmental matters tends
to generate still further support. One finds politicians of both political parties
vying for honors as guardians of environmental quality.
While the evidence seems clear that there is a public consensus con-
cerning environmental quality at the present time, it is not at all clear that
the public agrees what should be-done. What apparently has been rejected is
the recommendation of some neo-environmentalists who call for drastic
changes in life styles, commensurate with a re-ordering of national priorities
concerning the role of science, technology and economic growth in American
society. The dominant belief, as reported by Tichenor -j^aL [52] for example,
is that "technology got us into the environmental crisis and technology will
get us out". The adoption of this belief in the technological ethic appears to
be widespread and, to date, few seem to be willing to give up their automobiles,
restrict their use of electricity or accept control measures which threaten
their economic self-interest.
-------
14
Enforcement and Community Conflict
In the short run there appears to be more support for maintaining
economic security of a community -- even among those most knowledgeable
concerning pollution matters -- than for the enforcement of pollution control
measures which strike at the heart of economic self-interest. Support for
pollution control measures which imply unemployment or other forms of
economic hardship is the least within those communities directly threatened,
according to Tichenor e
-------
15
also receives most or all of his externalities. The other is where the com-
munity which receives none or few of the economic benefits of a polluter also
receives most or all of his externalities. Assume, in the first instance, that
pollution conditions are sufficiently severe to impinge on the public's con-
sciousness. One would expect conflict about enforcement to occur within
the community if >.t appeared that abatement would jeopardize the economic
health of the community. The parties to the conflict, initially, likely would
be among the community's elite. Higher knowledge and favorable attitudes
toward environmental quality are associated with high educational levels,
professional or semi-professional occupations and participation in community
organizations. These people are more likely, compared to those low in these
attributes, to initially perceive and assess the implications of a polluter in
terms of the community's long-term welfare. Those who hold pro-environmental
attitudes are also likely to hold pro-community attitudes and may see nothing
inconsistent with holding both sets. It isn't until a situation ari ses in which
these two attitudes interact or are brought into relation with one another that
the inconsistency becomes apparent and a re-evaluation is set in motion.
In the Tichenor study cited earlier, those interviewed responded to a set of
proposed environmental measures that had a bearing on the economic welfare
of a community. Support for the environment was most prevalent among the
well-educated in all communities. Furthermore, it is among the elite that
one would expect to find individuals who possess the skill to develop and
propose workable alternatives to continued pollution, to obtain access to
governmental enforcement machinery and, in general, to engage in a con-
frontation with a polluter and other community elites. Moreover, the chances
are also good that in many non-metropolitan communities intense conflict
over pollution abatement is not likely to break out or become public. For,
as Coleman [8] has'noted, participation in community organizations tends
to cement the community by providing channels of communication, both
among citizens and between citizens and the elite. These channels also
serve as sources of influence when conflict threatens to break out. Individuals
in these organizations hear both sides.of the controversy in face-to-face contact
-------
16
and are subject to persuasive cross-pressures. Conflict is more likely to
break out in communities where the communication and influence networks
are not available or are not working. The evidence also suggests that the
local mass media is likely to "manage" conflict by not reporting it. Olien,
^t &L , [35], in a study of 88 Minnesota communities, found that the com-
munity press tends to support community institutions by ignoring the dis-
ruptive aspects of public life.
One would not expect the same behavior of people in a community which
received none or few of the economic benefits of a. polluter but received all
or most of his externalities. Again, assume that pollution conditions are
severe enough to impinge on the public's consciousness. One would expect
conflict about enforcement to occur between communities if it appears that
abatement would jeopardize the economic health of the polluting community.
The conflict is carried to some higher political institution for resolution,
such as a regional or state environmental control agency. One would also
expect to find a high level of consensus within each community concerning
the issue. As well, the local mass media in each community would be
expected to manage the conflict by reporting the issue fully, emphasizing
each community's side of the controversy respectively.
^
In both of these instances it is argued that social conflict may serve
the interest of environmental quality. Conflict is viewed as a method by
which values concerning environmental quality are affirmed. In the long
run society itself is likely to benefit from this form of conflict. As Coser
[10] has noted, conflict is a process through which adjustment to new con-
ditions can be brought about. It is through conflict behavior that norms
are modified or created and the continuation of society is assured under
changed conditions. A rigid social system -- one not permitting conflict --
will block adjustments and will tend to maximize the danger of catastrophic
social breakdowns.
-------
17
The externality concept, as used here, cannot be applied to predict forms
of conflict for all air pollution situations. One can conceive of communities
which produce pollutants that mix with emissions from other communities
in the same airshed. Pollutants may or may not fall on communities in
proportion to each one's contribution. An assessment of social effects of
externalities is difficult at best. In this instance the community conflict
model may not be appropriate. One would expect the advancement of an
environmental ideology to become one basis of conflict rather than community
self-interest per se.
Moreover, while the effects of externalities may be important in sug-
gesting forms of community conflict, they are not likely to be as useful as an
alternative and related concept for predicting forms of conflict resolution.
This alternative concept is the use of economically and technically feasible
controls for air pollution emissions.
At times there is considerable interest shown by a polluter or an enforcer
(i. e. , governmental pollution enforcement agency) in reducing emission levels.
But the technology is not always available to guarantee lower rates --at least
sufficiently low to reach desired air quality goals. The availability of econo-
mically and technically feasible emission control devices may well determine
the level of emissions permitted. The level of control is considered a function
of the technological effectiveness of control measures, of control cost and of
damages averted by lowering pollution levels. Applying these elements, the
decision rule advanced by some economists is to expand air pollution control
so that the additional cost of one more increment of control is just equal to
the value of the additional damage averted. The point where the total net
benefits (totaj. benefits minus total costs) is a maximumi defines the optimal
allocation of resources to air pollution control. It is recognized that it is
difficult to quantify all the benefits or to quantify some of them so they are
commensurate with dollar values employed to measure costs. One can, however,
conceive of at least three instances when the presence or absence of feasible
See, for example, the recent article by Downing [13].
-------
18
emission control technology would have a bearing on the form of conflict
resolution.
The first is when both the polluter and enforcer agree that economically
feasible emission technology is available. Once a polluter is aware of
available feasible technology he agrees to comply with control standards.
It is a matter of establishing compliance schedules and testing and main-
taining control equipment. The second instance is when the enforcer
believes, but the polluter disagrees, that feasible emission control devices
are available. The disagreement arises because some of the pollution control
benefits are "external" to the decision process of the polluter. In this instance
one alternative for resolution commonly employed is to use some aspect of
the adversary process. Procedures of this process include informal and
formal agreements, binding judicial decisions without use of rules of evidence
or court tests where rules of evidence are employed. Decisions established
through some adversary procedure do not necessarily rule out subsequent
social conflicts, however. The ability of an enforcer to impose sanctions
upon a polluter by successful court (or administrative) decision, for example,
has the potential for considerable social upheaval as enforcers move to
impose emission restrictions which threaten either the short- or long-run
economic security of a community. Whether or not conflict breaks out may
be more a function of the perceived fairness of enforcement and the effect of
enforcement on the community's future than on mere economic well-being.
The third instance is when both the polluter and enforcer agree that feasible
controls are not available. In this instance one is likely to find, other than
a crash program of zesearch, attempts to modify the behavior of the public
or the polluter to adjust to severe pollution conditions. One sees, for
example, "smog alerts" employed in the Los Angeles and the New York -
New Jersey metropolitan areas and attempts to "manage" field burning in
Oregon's Willamette Valley. Such modifications of behavior are often viewed
as a temporary solution, based on the assumption that technology will develop
sufficient hardware to adequately control emissions. As with many temporary
-------
19
, modified behavior patterns tend to become ch-fined as Ihc Hiatus quo
until another environmental episode underscores their limitations, often
triggering further social conflict.
The notion of level of technology focuses on the role of some enforcer
as an agent for conflict resolution on the assumption that it is through him
that the laws or authority of government and use of available technology are
operationalized. Inadequacies of either the availability or application of law
or technology are necessary (but not sufficient) to generate a pollution episode.
An episode is defined as the occurrence of an event which threatens to produce
an unusually large loss such as damage to human health, property or status,
and in which the bureaucratic structure is unable to cope with or to controL
with existing laws and technology. A pollution episode stems from the con-
tamination of some aspect of the environment so it is unfit for intended use.
One effect of a pollution episode is to compel a political leader to affirm one
set of rival interests and values of competing groups. This may be operation-
alized by temporarily suspending or severely regulating the activities of a
polluter, the public, or both.
The social effects of externalities and availability of economically and
technically feasible controls may be summarized by examining their joint
distribution in the following six-fold table:
Availability of economic-
ally and technically feasible
emission controls
Pollution externalities
occur within a polluting
community
Pollution externalities
occur outside a
polluting community
1) Enforcer, polluter
agree, feasible
emission controls
available.
2) Enforcer agrees^ pol-
luter disagrees,
feasible emission con-
trols available.
3) Enforcer, polluter agree,
feasible emission con-
trols not available.
Minimal social con-
flict; compliance.
"Managed" social
conflict within a
community; adversary
procedure employed.
"Managed" social con-
flict within a community;
behavior management
employed.
Minimal social
conflict; compliance.
Social conflict be-
twee,n communities
public; adversary
procedure employed.
Social conflict between
communities public;
behavior management
employed.
-------
20
One may well describe the type of social conflict and action concerning air
pollution effects by knowing the nature of the pollution externalities and the
availability of feasible emission controls. Suppose, for example, one wished
to study the effects of social conflict concerning air pollution abatement.
He would likely find these expressed in a situation where feasible controls are
not available; and where pollution externalities occur outside the polluting
community. Suppose, on the other hand, one wished to study the effect of
type of community power structure in resolving social conflicts brought about
by air pollution abatement. He would likely examine communities whose
pollution externalities occur within the community, and where feasible
emission controls are either not available or where the enforcer and
polluter disagree over the feasibility of available controls.
Social conflicts over air pollution effects and controls can, in themselves,
crystallize the attitudes of individuals and groups in affected communities.
Positive or negative attitudes are formed (or changed) from the interaction
between two or more groups over time. Norms, or direct guides to posi-
tive or negative action, are established as the end product of this interaction.
Relations among individuals belonging to these groups must become standard-
ized as norms toward other groups and toward situations before they are re-
flected as inter-group attitudes. The process of attitude formation or change
concerning pollution effects thus appears relevant to any study involving an
explanation of human behavior over this issue.
Attitudes and Definition of the Situation
It has frequently been asserted that a person's observed verbal or non-
verbal behavior is determined by his psychological structuring of events.
As well, this structuring is jointly determined by the interaction of both
external and internal attributes. Examples of external attributes are aspects
of objects or events in the environment and social influences such as inter-
personal suggestions and group participation and pressures. Internal attri-
butes include personal knowledge, interests, motives, set and socially-
derived factors such as attitudes, beliefs, internalized norms and values.
See, for example Sherif and Sherif [42].
-------
21
Presumably, if one had sufficient theory to tell how to combine the significant
attributes in their proper relationship with one another and, knowing the
scores of individuals on each attribute, one could then predict with a reason-
ably high level of certainty each individual's psychological structuring and
his observable behavior toward some event.
Moreover, if one had sufficient theory to specify the effects of a change
in external or internal attributes in terms of changes in psychological struc-
turing, then one could again predict with reasonable certainty changes in
each individual's observable behavior.
Individuals, however, may or may not be clearly conscious of their
psychological structure. They are often unable to explain their behavior
to anyone. To be sure, social scientists may be able, after reviewing inter-
view protocols, for example, to reconstruct or infer models which explain
an individual's behavior. Psychological structuring does not necessarily
play a role as a conscious antecedent to the individual's behavior. It is a
theoretical construct which mediates between external and internal attributes
and observable behavior.
Fxternal and internal attributes do, however, often impinge on the con-
sciousness of many and the resultant structuring is termed the individual's
"definition of the situation". This term is derived from W. I. Thomas' [49]
aphorism "if men define situations as real, they are real in their conse-
quences". This has led workers to investigate how individuals develop common
perspectives and communicate these to each other. The term implies no
one-to-one correspondence between an objective event (or situation) and
people's perceptions of that eventthat something intervenes between
events and people which makes possible a variety of interpretations to reality.
The same event, for example, can have different meanings for different
people and the degree of difference can be expressed as observable behavioral
differences. As well, objective events contain common elements through
time for different people (e.-_g- > many aspects of driving a car along one
smoggy freeway are the same as driving a car along another smoggy freeway).
-------
22
People who enter objective events carry similar backgrounds and attitudes
which make it possible for them to define events in similar ways. Uniqueness
of events, while an aspect of any situation, cannot be ruled as a barrier for
scientific study of them. What is required is specification of "something"
which intervenes between events, or situations, and people.
Stebbins [45 ] considers the definition of the situation a "more or less
conscious synthesis and personal interpretation of the interrelation of the
set of predispositions and the elements of the subjective situation". Three
concepts play roles in Stebbins1 formulation: objective situation, subjective
situation and action orientation. He defines an objective situation as the
immediate social and physical surroundings and the current physiological
state of the actor (L e^. , individual). The subjective situation is those
components of the objective situation which are seen by the actor to affect
any one of his action orientations. The term "action orientation" refers
to any purpose, trivial or important, which the actor has as he enters the
objective situation, such as driving down a smoggy freeway. The point is
that action orientations serve as guides to behavior as the individual enters
the objective situation. Stebbins asserts that action orientations guide the
selection of elements from the objective situation for the development of the
subjective situation. Subjective situations are postulated to have time limits;
action orientations play a crucial role in time calculations, since situations
exist as long as it takes the individual to manipulate them. Subjective situa-
tions, once identified, must be interpreted and given meaning In terms of the
individual's background. This interpretation, as noted above, is the definition
of the situation and its final formulation is affected by both predisposltional
and situational factors. Predispositlonal factors stem from the actor's past
definitions of situations, from former and future plans of action, from past
and present action orientations .and from the actor's values and attitudes,
among others. Chronologically, the definition of the situation immediately
precedes the individual's behavior which has responded, in turn, to the sub-
jective situation. A definition Is completed only after a period of conscious
reflection. Goal-directed behavior is delayed until the Individual has defined
-------
23
the situation. Action unfolds commensurate with the definition. Stebbins allows
for trial-and-error behavior during reflection. Goal-directed behavior, how-
ever, becomes possible only after meaning is established for the situation in
terms of the action orientation. As well, he notes that definitions vary in
amounts of reflection required; habitual and routine situations call for less
reflection than the unusual, the relatively new or crisis situations. In any
case, definition is a necessary condition to goal-directed behavior.
Empirical work supporting definitional processes has focused on the
effect of defined situations as modifier of attitudes and the procedures
actors employ in the definitional process. Gordon [15] considers the person's
definition of the situation a modifier of attitudes toward certain objects.
McHugh £?] , using a laboratory setting, was able to demonstrate that differ-
ent actors appear to follow similar procedures for constructing substantially
different definitions.
It is recognized that explication of the definitional process has not led to
widespread theoretical development and empirical testing. One is not likely
at this stage to find sociologists making precise predictions which are capable
of empirical verification or refutation. But the definitional formulation pro-
vides a view for examining phenomena capable of affecting large numbers of
people simultaneously. As well, variables may be selected on the basis of
definitional literature that are worthy of further study. It is in this vein that
the definition of air pollution situations will be examined.
Definition of Air Pollution Situations
It can hardly be refuted that some aspect of air pollution -- loss of visi-
bility, odors, perception of particulate matter, to name a few -- has not im-
pinged on the consciousness of most Americans. Even before it became
fashionable to express environmental concerns, air pollution was (and still
is) defined by some as evidence of employment and economic activity. To
others, air pollution represents an environmental insult of the worst sort
threatening the survival of man and all other creatures living on this planet.
-------
24
That multiple definitions are possible from the same (or nearly the same) ob-
jective pollution situation is supported by studies relating public concern,
attitudes, beliefs and knowledge about air pollution levels, effects and con-
trols. The work of Crowe [H], among others, is typical. Moreover,
multiple definitions at one point in time may represent a redefinition for
some --a change in the components of meaning. One effect of a redefinition
may well be a change in attitude, defined as a change in the positive or
negative evaluative predisposition of an individual toward some aspect of air
pollution. One way to examine the postulated association between definition of
the situation and change (or no change) in attitude is to study inputs from the
objective situation that may have a bearing on definition or redefinition.
One such input is to study the mass media themes concerning air pollution
which have been published in a community over a relatively long period of
time. Themes, or assertions about some aspect of air quality, function in
defining situations by providing descriptions of the present. Concerning themes,
McHugh [27] asserts that "actors assume before the fact that a pattern of
meaning will be discovered in the events they observe. They are future oriented,
in that they take it for granted they will be able to make something of what is
yet to occur. They are also past oriented, in that they take it for granted that
what has already occurred will inform the future. Each of these orientations
folds back as an actor's description of the present. " One ingredient in defin-
ing the present is assumed to be the input of mass media content, as described
by frequency and direction of themes from the objective situation. A rigorous
content analysis should provide a basis for estimating definitional (or re-
definitional) characteristics of air pollution situations for individuals in a
community. Then if it appears that an air pollution situation has been defined
or redefined on the basis of a mass media content analysis, one can test if
attitudes toward air quality remain the same or are changed according to
definitional predictions. If, for example, it appears that an air pollution
situation is not redefined within a specified time range, then attitudes toward
this phenomena should not change within the same time range.
The inferential leap from content analysis of mass media themes to
attitude change of individuals is great, however. What should be specified
-------
25
as well is how Inputs from the mass media are related to attitude level or
change. If it is demonstrated empirically that a reciprocal causal link or
links can be established between attitude level or change, mass media use
and other endogenous variables, then one is on firmer ground for relating
results of a mass media content analysis to an individual1 s change or lack
of change in attitude. The formulation of such a communication model must
be drawn from work outside the definitional literature, however, since the
necessary statements capable of empirical tests are not available. Research
concerning the communication of scientific information to the public is likely
to prove more fruitful and some of this work will be examined next.
An Air Quality Communication Model
In a secondary analysis of science knowledge and attitude data from 27
sources, Tichenor [ 50J hypothesized that the communication process in
which attitudes and mass media exposure play a role can be schematised as
follows:
Variables predisposing Exposure Effect
toward exposure variables variable
More
education
-»
More general science
knowledge and more
positive attitude
toward science
-
More exposure to
mass media and
interpersonal
sources
->
More science
knowledge
Figure 1. A communication model for accumulation of science knowledge.
He stipulates that his predisposition variable toward exposure (those
which occur prior to exposure) are represented by two variables. Level of
education is designated as an antecedent variable, primarily because it occurs
first in the time sequence. Formal education establishes an individual's level
of general science knowledge as well as his positive or negative evaluation, or
attitude, toward science. Exposure variables -- exposure to mass media and
interpersonal sources -- represent complementary channels through which
additional science knowledge flows. Acquisition of such additional knowledge
represents the effects variable.
-------
26
The sample size of his data base was large enough for him to partial out
or control for the effects of a number of demographic, attitudinal and exposure
variables in order to determine if postulated effects, such as a level of edu-
cation, serve as an independent predictor of science knowledge. His central
findings suggest that both the predisposition and exposure variables add to
the predictive power of accounting for variation in science knowledge. His
analysis, however, did not permit him to state precisely the direction of all
the arrows (J-. £. , the effects) in the above model. It is still not known if
attitudes toward science lead a person to acquire more science knowledge
through greater mass media exposure, or if these three variables represent
an interdependent system. It is conceivable, as Hovland [19] has pointed
out, that attitudes may predispose a person to acquire certain types of
information but such information may also trigger a change in attitude
which stimulates seeking of still more knowledge. In addition, it is also
possible to reconceptualize the scheme so that the exposure and effects
variables represent specific science items, such as mass media exposure
to information and more knowledge about air quality. In fact, this is the
nature of much of the data with which Tichenor was dealing. Predisposition
variables included measures of level of education and other demographic
characteristics, measures of general science knowledge and attitudes toward
science. Measures of exposure and effects variables, however, often dealt
with specific science fields, such as mental health, medicine, psychology,
fluorine and tooth decay, smoking-cancer link, etc. Moreover, the list of
variables in each class is not exhaustive and thus can be expanded to include
others which conceivably underlie increased mass media exposure and science
knowledge. With these considerations, the Tichenor model might be profitably
reformulated to describe more precisely the relationship between an individual's
attitudes toward air quality, his level of mass media exposure to air quality
information and his level of knowledge about air quality.
-------
27
More
; interpersonal .
discussion (X )
More
with a
(X2)
^
More positive
attitude toward
air quality (Yj
t
experience
ir pollution
*x
^
r
Gre
infl
More knowledge
about air ^
quality (Y3)
t
More science
education (X )
ater social More
uence (X ) education (X4)
^ I
More exposure to
^. mass media about
air qualify (Y_)
s
/ Greater use of
environment (X^)
Figure 2. A reformulated model for accumulation of air quality
information by the public.
The model shown in Figure 2 stipulates that three variables are mutually
interdependent and are denoted as "Y" variables in the figure. These variables
are: attitude toward air quality, mass media exposure to information about air
quality and the level of knowledge about air quality. The interest here is not
with a single independent-dependent relationship but rather with a system of
relationships in which the "Y" variables are imbedded and which leads to a
description of the interdependent relationships among these variables. With
such a model, then, the existence of mutual interdependence among attitude,
exposure and knowledge variables can be tested. Is, for example, knowledge
dependent on attitude and exposure? Does such knowledge, in turn, affect
attitude and exposure? The use of a set of simultaneous equations should
provide an answer to these types of questions.
The other variables, denoted as "X's" in the model, are "outside" the
system. They serve as important predictors and, as will be noted later
in the section on methods, represent important operational considerations
for analyzing the system empirically. Direction of arrows does not repre-
sent a causal relationship but only predictions that relationships exist.
-------
28
The model can be described mathematically as a system of simultaneous
equations:
Y3 = P31Y1 + P32Y2 + ^31X2 + ^32X5 + ^3X6 + 63 (3)
where
Y. is the observed value for the appropriate endogenous
variable in the system^
p.. and v.. are constants, or regression coefficients,
J J
X. is an independent predictor or exogenous variable
outside the system, and
e. is a stochastic disturbance term assumed to be inde-
pendently and normally distributed, with mean equal to
zero and a constant, finite variance o . .
This concludes the discussion of the theoretical model. A review of
literature is presented as Appendix A to support the postulated relationships
within the model.
Methods for testing the mass media content - attitude change hypothesis
are presented next, followed by the results and discussion. Then, method-
ological procedures for testing the communication model are presented,
followed by the results and discussion.
-------
29
METHODS -- CONTENT ANALYSIS AND ATTITUDE CHANGE
Methodological considerations necessary to test the mass media content-
attitude change hypothesis are divided into four phases:
1. The selection of locations and times of known air pollution situations
appropriate for hypothesis testing.
2. The development of operational measures of mass media content
and attitude change.
3. The gathering of sufficient data to provide a basis for an empirical
test of the hypothesis.
4. The specification of analytical procedures employed to test the
hypothesis.
Each phase will be described in its respective order.
Selection of Location and Time
Several conditions are necessary to test the mass media-attitude change
hypothesis. One is the selection of a community or communities where air
pollution effects are severe enough to impinge on the public's consciousness.
Another is a pollution situation which occurs with sufficient regularity that
one can design a panel study and plan personal interviews to capitalize on
this regularity. A third is a relatively high level of attention to air pollution
sources and effects by the local mass media, particularly newspapers.
It seems most likely that one will find these conditions met in a situation
where pollution externalities occur outside the polluting community and when
an enforcer and polluter agree that feasible emission controls are not available.
Social conflicts concerning emission restrictions are likely to occur, according
to the discussion in the introduction (page 19). This conflict is likely to be
reflected in relatively high levels of mass media attention and public
awareness.
-------
30
Several communities were considered for study shortly after the project
began in early 1969. The cities of Eugene and Salem, Oregon, were selected.
Because of its location at the base of the Willamette Valley, Eugene has ex-
perienced poor natural ventilation, whereas Salem, located in the mid-
Willamette Valley, has had good natural ventilation. Therefore, a contrast
in visibility would be expected to occur when there is air pollution in the Valley.
Eugene's population was estimated at 109,614; Salem's, 95,748, according
to the 1970 Census of Population [7].
One predictable source of air pollution was agricultural field burning.
Farmers burn approximately 250,000 acres of post-harvest residue from
grass seed crops in late summer or early fall in the Willamette Valley. Heat
produced through straw and stubble burning provides at least partial control
of a number of grass seed diseases and disorders. As well, other benefits
recognized include effective and inexpensive residue removal, destruction of
weeds and weed seeds and physiological stimulation of seed yields. No
economically feasible alternative to field burning has yet been developed.
The largest and most concentrated area of grass seed production lies in
the area between Albany and Eugene (see Figure 3). Emissions from burning
consist of smoke (carbon particles of various sizes), ash and certain gases.
Most individual carbon particles are minute and cannot be seen without visual
aid. Larger particles can be seen easily by persons with normal vision.
Large numbers of minute carbon particles are called "suspended particulates "
and are a major source of reduced visibility. Larger particles, which settle
more readily, are a source of soiling. Relatively large volumes of smoke
emissions from agricultural field burning were considered an important
stimulus for public concern, particularly in Eugene. For this study, a
measure of visibility was employed as the measure of air quality. For a full
presentation of the scientific basis for visibility, or visual range as it is
sometimes called, see Middleton [31], Robinson [37], for example, notes
that the deterioration of visibility is probably the first indication of air
-------
25
31
Figure 3. Willamette Valley Basin and vicinity.
-------
32
pollution about which a citizen becomes aware. Visibility was selected
because it is most likely to have impinged on the consciousness of individuals
in the study areas, compared to other forms of air pollution.
As well, any test of the mass media content-attitude change hypothesis
must take into account any past definitional situation. Ideally, one would
like to measure attitudes toward air quality prior to any definition (or
redefinition); wait, hopefully, for events to be defined or redefined, then re-
measure attitudes of the same individuals. This ideal situation is not likely
to occur in practice, for the researcher has no control over mass media
attention to air quality. An alternative procedure is that, given the inordinate
time involved in content analysis, one artibrarily selects a time difference
for measuring and remeasuring individuals when attitudes conceivably will
change on the basis of known changes in air quality. After all, Zajonc [54]
asserts that "mere repeated exposure" of the individual to a phenomenon is
sufficient to enhance his attitude towards it. Attitude change scores are
then related to changes in level of air quality. If the mass media content
analysis shows no definition or redefinition within the time period of
measures for attitude change, then the definitional hypothesis is not sup-
ported if an attitude change does, in fact, occur. If the content analysis
specifies definition prior to measures of attitude change, and no measurable
change occurs, then one would consider this data support for the definitional
hypothesis. But, this support is not as strong as it would be if attitude
measures were obtained prior to and after definition.
Such a design calls for measures of the same individuals during periods
of high and periods of low visibility. Two interviewing waves were planned.
The first, using Schedule A, contained measures of both endogenous and
exogenous variables of the model. The second, using Schedule B, contained
parallel forms of measures of attitude and knowledge variables. It was
important to guarantee that use of Schedules A or B was not correlated with
time of interview or visibility level. Otherwise, a confounding of effects
might occur. Thus, the sample was drawn so that a random half of
-------
33
respondents was interviewed from each community in the following manner:
May 1970 Aug. -Sept. 1970 January 1971
% respondents % respondents % respondents
Schedule A 50 50
Schedule B 50 50
It was anticipated that visibility levels would be high in May 1970 and
January 1971 but low in August-September 1970. It was also anticipated
that the pattern of high and low visibility would be more pronounced in Eugene
than in Salem, 1. e. , Eugene would have lower visibility levels than Salem in
August-September. Timing interviews in this fashion would balance any
effects between type of schedule (A or B) and time of interviewing. Those
who were interviewed with Schedule A in May 1970 were reinterviewed with
Schedule B in August-September 1970; those who were interviewed with
Schedule A in August-September 1970 were reinterviewed with Schedule B
in January 1971. Observers in both communities, trained by a representative
of the Oregon State University Department of Atmospheric Sciences, completed
visibility observations every two hours on the hour, beginning at 8 a. m. and
continuing until dusk. This meant that seven observations per day were made
in May and August-September, five per day in January. Observations were
completed each day interviewers were in the field. In this fashion, profes-
sionals in the air quality field could compare levels of visibility between
interviewing periods and between communities.
Five year visibility data from U. S. Weather Service stations at the Eugene
and Salem airports were gathered to estimate more precisely the timing for
interviews. This analysis disclosed that a four-week period from mid-May
to mid-June and a four-week period in January were consistent periods of
relatively high visibility. The analysis did not suggest a consistent
timing for periods of low visibility; however, estimates by professionals
in the O. S. U. Department of Atmospheric Sciences suggested that late
August to mid-September was likely to be a period of deteriorated air
quality.
-------
34
Development of Operational Measures
Mass media themes were defined in the introduction as "assertions
about some aspect of air quality". Themes served as recording units for
measuring the content of communication; in this instance, mass media
available in Eugene and Salem, Oregon, the communities selected for study.
The first step in theme development in this study was to factor analyze
intercorrelations among attitudinal items that were available from a pre-
test conducted among a random sample of individuals in Portland, Oregon.
(Details of the pretest will be presented in the section concerning the
methodology for testing the communication model. Suffice to say that pre-
test data were the earliest available attitudinal information that could be
developed in the study. ) The factor analysis disclosed six orthogonal factors
and the first list of potential themes was developed from these factors. Each
factor was characterized by a single statement. These statements, or tenta-
tive themes, could be written to express both a positive and a negative direction
toward some aspect of the referent, air pollution. A random sample of 52
weeks of newspaper articles, in which some aspect of the referrent "air
pollution" appeared at least once, was selected from an 11-year (1960-1970)
collection of such articles. These articles had been reproduced from micro-
film copies of the Eugene Register-Guard and the (Salem) Oregon Statesman.
Two trained coders, working independently, tested the six tentative themes
on a sub-sample of articles, with the article serving as the unit of analysis.
Discrepancies in coding were resolved by clarifying broad terms employed in
the themes, by rewriting ambiguous themes and by writing new themes.
Overlapping themes were omitted. The final list of 15 themes was then tested
on the remainder of the sample articles. The final version of 15 themes
was written as two lists to facilitate coding; one list, "for" clean air; the
other, its negation, "against" clean air. These themes are:
Theme "For" Clean Air (Code 1) Themes "Against" Clean Air (Code 2)
1. Air pollution is a health hazard. 1. Air pollution is not a health hazard.
-------
35
Theme "For" Clean Air (Code 1)
2. Air pollution is a menace to the
public welfare.
3. Air pollution is becoming a
serious problem and public
opinion must pressure govern-
ment into doing something to
control it.
4. Before the problem of air pollu-
tion becomes too serious, we can
find the resources and technology
to eliminate it by working together
now.
5. Air pollution is a problem caused
by industry, and it is industry's
responsibility to clean up the air.
6. Industry is willing to cooperate
in cleaning up its air pollution.
7. Polluters are willing to use alter-
native methods of waste disposal
even if they are convinced it is
neither economically feasible nor
technologically practical.
8. Air pollution control agencies need
more power if they are to control
the problem of air pollution.
9. If air pollution control agencies
are to be effective, they must be
organized on a regional basis.
10. Air pollution may be a problem,
and industry and economics are
not as important as clean air.
11. Industry is a major contributor
to the air pollution problem.
12. (Not useful. See Code 2)
13. Air pollution has always existed
and it may be a. problem, and
something must be done to
eliminate it.
Themes "Against" Clean Air (Code 2)
2. Air pollution is not a menace to
the public welfare.
3. Air pollution is becoming a serious
problem, but the use of public
opinion to pressure government is
not the way to control it.
4. (Not useful. See Code 1)
5. Air pollution is not a problem
caused by industry, and it is not
industry's responsibility to clean
up the air.
6. Industry is not willing to cooperate
in cleaning up its air pollution.
7. Polluters are willing to use
alternative methods of waste
disposal if they are convinced it
is economically feasible or
technologically practical.
8. Air pollution control agencies
already have enough power to
control air pollution.
9- Air pollution control agencies
need not be organized on a regional
basis to be effective.
10. Air pollution may be a problem,
but industry and economics are
more important than clean air.
11. Industry is only a small contribu-
tor to the air pollution problem.
12. It is difficult to place the blame for
air pollution since it is not con-
fined to any physical boundary.
13. Air pollution has always existed
and it may be a problem, but there
is nothing that can be done to
eliminate it.
-------
36
Themes "For" Clean Air (Code 1) Themes "Against" Clean Air (Code 2)
14. The available means of control- 14. The available means of controlling
ling industrial air pollution is industrial air pollution is ineffect-
effective or technologically ive or technologically impractical.
practical.
15. (Not useful. See Code 2) 15. Air pollution is not a problem.
The negation of three themes was not useful and were omitted. An
example is the theme "Air pollution is not a problem". Its negation, "air
pollution is a problem" is not useful, since all themes in this direction imply
the existence of a problem.
The measure for attitude change will be described in the section on
analytical procedures, presented below, and in the section concerning attitude
measurement which is presented in the methodology section for the communi-
cation model.
Data for Hypothesis Testing
Community newspapers provided the most appropriate source for
measuring mass media content by counting frequency of themes. Microfilm
copies of newspapers were readily available to reproduce selected articles
for analysis. All newspaper articles -- news stories, features, editorials,
letters-to-the-editor, advertisements -- which contained at least one refer-
ence to some aspect of air quality were reproduced from an 11-year period
(1960-1970). The newspapers were the Eugene Register-Guard and the
(Salem) Oregon Statesman. Two coders, working independently, read each
article, noting the presence or absence of each theme. Each article was
also coded for type of article (news, editorial, etc. ); length (in column
inches); story placement (front page or other); source of pollution cited
(field burning, industrial, automobile, commercial, residential, meteor-
ological, forestry, natural, other); and scope of problem (local, regional,
state, out-of-state, national, world). As well, month, year and newspaper
-------
37
in which article appeared were coded. Discrepancies between coders were
resolved by a referee. At least one coding discrepancy occurred in approxi-
mately 25 percent of the articles.
A total of 3, 054 articles was coded -- 1, 963 from the Eugene Register-
Guard and 1, 091 from the Oregon Statesman. Approximately 16 percent of
the articles from each newspaper contained no themes. These articles
provided procedural information concerning some aspect of air pollution,
such as notices of hearings before regional authorities. As well, coders
agreed that 5 percent of the articles from each newspaper contained themes
which could not be coded with the available list of themes. The remaining
79 percent of articles contained at least one theme and these articles pro-
vided the data base for analysis.
Analysis of Data
Analysis of mass media content frequencies will be described first;
then measurement of attitude change will be presented. For mass media
content, analysis consisted of a visual examination of a frequency distri-
bution of total number of articles, of article type, of sources of air pollution
and of themes across time. Time, in months, was the abscissa and fre-
quency the ordinate. A three-month moving average was employed to
smooth out the frequency curve. Nine graphs were planned. First, the
frequency of total number of articles for the newspapers from each
community was plotted on one graph. Second, the frequency of article
type -- news story, feature, editorial, letter-to-the-editor -- was plotted,
one graph for each community. Third, the frequency of pollution sources
mentioned -- field burning, field burning and other sources, other sources
only, and no source -- was plotted, one graph for each community. Fourth,
the "for" clean air themes were plotted, one graph for each community.
Finally, the "against" clean air themes were plotted.
If air quality (or some aspect of air quality) is defined as an issue, one
-------
38
would expect to find some sustained attention to the topic over time. Prior
to being defined as an issue, one should find frequencies generally low with
no systematic pattern discernable. At some point in time, however, fre-
quencies should increase markedly and continue at relatively high rates with
easily explanable patterns occurring. For example, once field buring is de-
fined as an issue, one should expect increased attention to occur during the
burning season, compared to the non-burning season. Yet, frequencies
during post-definitional non-burning months should be higher than during
pre-definitional months. Themes associated with definition should aid in
conveying a description of the definition itself. By examining the frequency
of themes, one may well be able to construct a dialogue for describing the
mass media content at any specific period of time and, by comparison,
note differences among particular time periods. For example, one would
expect a pre-definitional description of field burning to differ markedly from
a post-definitional description.
As well, one can fix the date after which to expect definition. Field
burning began during World War II and reached a plateau of approximately
240, 000 acres by the late 1950's. Assuming that the pesticide controversy
of the early 1960's was the forerunner of the present environmental move-
ment, and that the publication of Silent Spring played a major role in defining
the pesticide issue, then one should expect environmental issues to develop
sometime after 1962, the date of publication. Public concern over air quality
is expected to be one of these issues. The intent of this analysis is to
An estimate of grass seed acres burned in the Willamette Valley can be
made from acres harvested, since the practice of burning had been widely
adopted by I960 for all grass seed crops (See Hardison [17]). Acres har-
vested showed no great fluctuations between 1959 and 1970, according to
Middlemiss and Coppedge [30] and other sources at Oregon State University.
The yearly total grass seed acres harvested were: 1959 - 241, 000; I960 -
235, 700; 1961 - 216, 500; 1962 - 222, 000; 1963 - 215, 300; 1964 - 227, 500;
1965 - 225, 600; 1966 - 228, 100; 1967 - 216, 000; 1968 213, 000; 1969 -
225, 300; 1970 - 242, 700. Some grain acreage was burned in the I960 decade
but reliable estimates are not available.
-------
39
determine the date of definition of air quality as an issue.
This concludes the analysis proposed for the mass media content fre-
quencies. Next, analysis of attitude change scores will be described.
A regression equation, employing dummy variables, was written in an
effort to adjust attitude change scores of individuals in each community for
a number of effects, including the effect of a change in air quality. The
equation was:
where
= f(X0' *!' X2'
Y. is final attitude score minus initial
attitude score of the individual,
X is 1 for mean effect,
X is 1 if high visibility occurred first,
0 if low visibility occurred first,
X is 1 if Form A used first,
0 if Form B used first, and
X is initial attitude score of the individual.
Significance of coefficients was tested by Student's t_. A significant t
value for the X coefficient would provide evidence that a change in air
quality was associated with a change in attitude.
-------
40
RESULTS AND DISCUSSION --
CONTENT ANALYSIS AND ATTITUDE CHANGE
Results in this section are divided into four phases:
1. Results of the content analysis.
2. Analysis of visibility observations.
3. Analysis of attitude change scores.
4. Discussion of results.
Each section will be discussed in its respective order.
Results of the Content Analysis
Frequency distributions of types of content are presented in Figures 4-22.
Time, in months, is the abscissa and frequency the ordinate for all Figures.
The frequency curve is a three-month moving average. Figure 4 is the-fre-
quency curve for the total number of articles for Eugene and Salem, respect-
ively. Figures 7 and 8 are frequency curves of the sources of air pollution
mentioned for Eugene and Salem, respectively. Figures 9-20 are frequency
curves for the "for clean air" themes for Eugene and Salem, respectively.
There are two themes per chart, each theme number corresponding to those
listed on Page 34. Curves representing the 12 themes were too concentrated
to present on one chart. Two themes were selected which would appear to
vary sufficiently in each graph so one can easily follow each curve visually.
Figures 21 and 22 are frequency curves for the "against clean air" themes
for Eugene and Salem, respectively. Figures 9-22 are presented as
Appendix C.
It had been noted in the section on Methods -Content Analysis (Page 29)
that analysis would consist of a visual inspection of the different frequency
distributions to determine the approximate date when air quality was defined
-------
60 T
50 "
40
o
a
3 30
u*
-------
35
30 -
25 "
20
15
10
5
mi ijitu n it IHJI
1960 1961 1962 1963 1964 1965 1966 1967 1968
Figure 5. Frequency distribution for type of article, Eugene, 1960-1970.
1969
1970
-------
35
30
25 .
20 '
15
10
5
1960 1961 1962 1963 1964 1965 1966 1967 1968 1969
Figure 6. Frequency distribution for type of article, Salem, 1960-1970.
1970
OJ
-------
35 .
30 .
25 . .
20
Field Burning &
Non Field
Burning
0 MIII i>miiyiiiTiiiMnpnii n ii ii [i n ii rYMTKji IMKTTI nvpi iiYlvi n i| IXTTIIII n [i n n u n n )i irmm i i[li mi n ni| n n n mi t( i
I960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970
Figure 7. Frequency distribution for type of air pollution cited, Eugene, 1960-1970. 4>.
e-
-------
40 T
35 4-
30 -f
25 4-
o
2J 20
15 -I-
10 -I-
5 +
Non Field Burning :
Field Burning &
Non Field
Burning
1960
1961 1962 ' 1963 1964 1965 1966 1967 1968 1969
Figure 8. Frequency distribution for type of air pollution cited, Salem, 1960-1970.
1970
-------
46
as an Issue. It was asserted that one should look for some sustained attention
to the topic over an extended period of time. Prior to being defined as an issue,
frequencies would be generally low with no systematic pattern discernible.
At some point in time, however, frequencies should increase markedly and
continue at relatively high rates with explainable patterns occurring.
Applying this very general guide to the data, it appears that the time
of definition for Eugene occurred in 1966 or, possibly, early 1967. Fre-
quencies never did reach the low levels prior to these dates -- even in early
1968 when the news was devoted to the presidential primaries. The pattern
of "for clean air" themes was mixed with some continuing in use from 1966,
others occurring primarily during the three-month field burning season of
mid-July to mid-October of each year. The only exception to the definitional
pattern was the frequency of "against clean air" themes. Frequency of these
themes disappeared in the Eugene press in early 1968. The sharp rise of
total mass media attention during the 1969 burning season underscores the
occurrence of an episode in Eugene. On August 12, severe visibility re-
strictions prompted a delegation of Eugene residents to call on the state
governor who returned with them to Eugene. The governor ordered a seven-
day emergency ban on all field burning and called the Environmental Quality
Commission into emergency session. Following the ban, a new quota system
was established with farmers drawing lots to obtain permits to burn a share
of the limited daily acreage allowed. The quota system was in effect for
the 1970 and 1971 burning seasons.
The time of definition in Salem is less clear-cut. It is likely to have
occurred in the late summer of 1969, perhaps as an effect of the episode in"
Eugene. Actions of the state governor, who resides in Salem, the state
capital, may have defined the issue not only for Salem but for the state as
well. There was an increase in frequency in 1967 but this increase dropped
to near zero in early 1968. Sustained use, however, appears after the 1969
burning season; thus, the issue was likely defined by late 1969-
-------
47
A study of the frequency of themes, types of articles and types of pollution
sources cited suggests that the air quality issue in both communities was well
defined prior to our interviewing of respondents in 1970. On the basis of the
definitional hypothesis, one would not expect attitudes to change during the
three-month interval between interviews. A null hypothesis is postulated.
Analysis of Visibility Observations
Visibility observations were grouped by distance and by source of re-
striction -- natural conditions or air pollution. The results are summarized
in Table 5 (for Eugene) and Table 6 (for Salem). These results were analyzed
by a faculty member of the Department of Atmospheric Science?, and he
concluded
a. For Eugene, visual restrictions due to air pollution
were greatest in August-September and least in
May and January. Visual restrictions due to natural
conditions were greatest in January.
b. For Salem, visual restrictions due to air pollution
were greatest in August-September and least in
May and January. Visual restrictions from natural
sources were greatest in January.
c. Visibility restrictions due to air pollution in Eugene
were greater than in Salem for any distance in August -
September.
Timing of interviews occurred during anticipated levels of visibility
for each community as measured by visibility observations. May and January
interviews were during periods of relatively high visibility; the August-
September interview was during a period of relatively low visibility. As well,
visibility in Eugene was less than visibility in Salem for the August-September
interviews.
17
Personal communication from Dr. E. W. Peterson, Assistant Professor
of Atmospheric Sciences, Oregon State University, April 5, 1971.
-------
48
Table 5. Summary of Visibility Data at Four Distances for Eugene, Oregon
*
Miles Visibility
O
<2 P
N
O
<3 P
N
O
<4 P
N
O
<8 P
N
***
(N)
May 1970
(7/13)**
. 99
. 01
0
. 99
. 01
0
. 97
. 03
0
. 77
. 23
0
(91)
Aug. - Sept.
(7/17)
. 95
. 05
0
. 92
. 08
0
. 76
. 24
0
. 34
.66
0
(119)
Jan 1971
(5/10)
.86
.02
. 12
. 64
. 10
.26
.48
. 16
. 36
.42
. 14
. 44
(50)
*
O - Percent of observation marker(s) visible.
P - Percent of observation marker(s) not visible due to air pollution.
N - Percent of observation marker(s) not visible due to natural
conditions (fog, rain, snow, etc. ).
sjcafc
Number of observations per day divided by the number of days.
***
Number of observation-days.
-------
49
Table 6. Summary of Visibility Data at Five Distances for Salem, Oregon
J{
Miles Visibility
0
<2 P
N
O
<3 P
N
O
<4 P
N
O
< 5 P
N
O
<8 P
N
***
(N)
May 1970
(7/14)**
.98
. 01
.01
. 98
. 01
. 01
. 93
. 05
. 02
. 98
. 01
. 01
. 92
.06
.02
(98)
Aug. -Sept.
(7/20)
. 97
. 02
.01
. 96
. 02
. 02
. 90
. 09
. 03
. 94
. 04
. 02
. 66
. 29
. 05
(132)
Jan. 1971
(5/10}
. 82
0
. 18
. 56
0
. 44
. 40
0
. 60
. 42
0
. 58
. 28
0
. 72
(50)
*
O - Percent of observation marker(s) visible.
P - Percent of observation marker(s) not visible due to air pollution.
N - Percent of observation marker(s) not visible due to natural
conditions (fog, rain, snow, etc. ).
***
Number of observations per day divided by the number of days.
!e
Number of observation-days.
-------
50
Analysis of Attitude Change Scores
A regression equation (Equation (4), Page 39) was written to adjust
attitude change scores for regression effects and to determine if a change in
air quality was associated with a change in attitude. The coefficients obtained
for Eugene and Salem were:
Y^ = 15. 8529 + 0. 3670 X, + 0. 3829 X. - 0. 2655 X0 (17)
Eugene (P. 490) 1 (0.507) 2 (-6.100) 3
Y = 2. 3071 + 0.0295 X + 2. 2883 X - 0. 3771 X (18)
Salem .(0.037) (2.957) (-9.107)
Numbers in parentheses under the coefficients are Student's Rvalues.
For Eugene, only the coefficient associated with regression on initial attitude
scores, X was significant. For Salem, the coefficients associated with
form effects, X?, and regression, X , were significant. In neither com-
munity was the coefficient associated with level of visibility, X , significant.
Thus, one can conclude that a change in air quality was not associated with
a change in attitude (Y.) in either community. A high proportion of the
variance in attitude-change scores can be accounted for by regression:
effects; i. e. , errors of measurement associated with scores of the dependent
variable and, for Salem, by differences in form effects as well.
Discussion of Results
It was postulated, on the basis of a content analysis of the mass media,
that one would not find a change in attitudes occurring during the 1970 inter-
viewing period for respondents in both communities. The data presented in
Equations (17) and (18) support this hypothesis. It appears that the mass media
had defined the situation well before respondents were interviewed. If one
had wished to detect an attitude change toward air quality in Eugene, he should
have completed his initial interview prior to 1966; in Salem, prior to 1969,
according to the content analysis.
-------
51
An examination of the frequency curve in Figure 4 (for Eugene only) and
Figure 7 (for field burning only) reveals an interesting pattern associated with
the field burning months of each year (mid-July to mid-October) beginning with
1965. Mass media attention appears to increase sharply during these months
each year. Speculating a bit on this repeated pattern of attention leads one to
ask if the media wereresponding to greater visibility losses each year or if
some other phenomenon was operating. One wonders why, after a relatively
long history of field burning and, ostensibly, reduced visibility in the Fugene
area, the mass media began showing an interest in 1965 and in subsequent
years. The question may have been partially answered in the discussion
concerning the environmental movement and definitional process presented
in the Introduction.
Yet, one must ask further if increased attention to field burning in Fugene
in the 1965 to 1969 years altered the field burning habits of farmers sufficiently
to contribute to the episode of 1969.
To answer this latter question, one must assume as a premise the stipu-
lation; that grass seed farmers are entrepreneurs; j^ ^. , they are risk-taking
decision makers who are motivated by profits. As entrepreneurs, they have
established a set of priorities for farming practices, including field burning,
to achieve their objective. The time of field burning, then, is normally
dictated by profits.
Now, suppose farmers perceive hostility toward one of their essential
production practices, field burning, which threatens their ability to obtain
profits. Their possible behavior alternatives may be ranked on the basis of
guarantees that field burning will be completed each year. Since each farmer
is an entrepreneur, it is likely that field burning will tend to be completed as
soon as possible after harvest -- in some cases fields just harvested will be
burned before other fields'are harvested. Suppose, further, that rain or other
See Footnote 1, Page 38, for an estimate of grass seed acres burned yearly
between 1959 and 1970. The data indicate that the greatest acreage burned
was in 1959 and in 1970; the smallest acreage burned was in 1967 and 1968.
-------
52
conditions prt-vrnl harvest or field burning when some farmers would normally
harvest or burn. Potential acres for burning accumulate and are burned im-
mediately when conditions permit. Visibility restrictions then are likely to
be more pronounced in Eugene and an increase in attention and concern by
community residents follows.
It may well be that a sequence of events, such as those described above,
contributed to the episode of 1969- If this conjecture hats merit, it means
that the process culminating in the episode was set in motion in 1965 when
attention to field burning effects first appeared in any substance in Eugene.
As well, the continued arousal of attention and concern may have contributed
to a worsening of air quality until some form of workable controls was es-
tablished.
The process of visualizing the opinions of another group with some
accuracy is termed 'co-orientation1. In this situation two or more groups
may independently perceive a phenomenon and simultaneously comprehend
their joint perceptions of the phenomenon. For a more complete description
of the co-orientation concept and its role in consensus, see Scheff [39]. The
concept may well be useful in explaining or accounting for episodes in other
air pollution situations as well.
One would need to gather at least two diverse types of data to evaluate
the efficacy of the co-orientation notion for explaining this particular episode.
One type of data is an evaluation of visibility levels for Eugene prior to and
including the establishment of field burning as a practice, and prior to and
including the establishment of community attention and concern over the
effects of field burning. Moreover, one would suspect that co-orientation is
operating if there is a lack of correspondence between levels of visibility
and mass media attention prior to 1965 and a correspondence between these
variables from 1965 to 1969. The second type of data would be an in-depth
interview of field-burning farmers to determine if their perceptions con-
cerning reactions of Eugene residents altered their decisions about time of
burning.
-------
53
METHODS -- AIR QUALITY COMMUNICATION MODEL
A theoretical model has been developed which postulates that a person's
attitude, mass media exposure and knowledge about air quality represents an
interdependent system and that this system can be explained by a set of
predictor variables.
Methodological procedures required to test the model empirically were
divided into three phases:
1. The development of operational measures of the variables in the
theoretical model. Here, summated response scores to items asked of the
general public in a standardized questionnaire were used as measures in
all cases.
2. The gathering of sufficient data to provide the basis for an empirical
test of the model. A personal interview survey of a random sample of
residents living in Eugene and Salem, Oregon, was completed.
3. The specification and estimation of the parameters of a set of
equations by which the model was tested. A set of simultaneous equations
has been specified and a test of significance of variables by two-stage least
squares was planned.
Each phase will be described in its respective order.
Development and Evaluation of Operational Measures
Appropriate questions (items) were developed to represent operational
measures of the variables in the model. Specifically, these variables were
an individual's:
-------
54
a. attitude toward air quality (attitude)
b. level of exposure to messages in the mass media about
air quality (mass media exposure)
c. level of air quality knowledge (knowledge)
d. level of exposure to interpersonal information sources
(interpersonal exposure)
e. level of experience with air pollution (experience)
f. level of social influence (influence)
g. level of formal education (demography)
h. level of formal science education (demography)
i. level of use of the environment (use)
(Note: the terms in parentheses will be used to designate each
variable respectively)
Measurement of demographic, exposure and influence variables was straight-
forward and required only pretesting and evaluation. A first draft of the
questions was written at the outset and incorporated in the pretest question-
naire.
Measures of the remaining variables were less straightforward. They
required additional information before the items could be specified. First,
measures of attitude and knowledge were considered; then measures of ex-
perience and use were considered.
Attitude and Knowledge. A content analysis of the local mass media
and magazines published in recent years was employed. For the mass media,
microfilm tapes of the daily newspapers in the two target cities were read
independently by teams of two trained individuals each. All news stories,
editorials, letters to the editor, etc. , concerning air pollution were noted.
Newspapers were the Eugene Register-Guard and the (Salem) Oregon Statesman.
In addition, the Reader1 s Guide was surveyed for titles of magazine articles
concerning air pollution. Time interval in all cases was the past six years --
January 1, 1964, through December 31, 1969. Appropriate material from
these sources was reproduced by xeroxing. This resulted in 1, 417 articles
-------
55
from the Register-Guard, 687 articles from the Statesman and 163 articles
from popular magazines.
Trained coders performed a content analysis of the xeroxed articles.
The statement served as the unit of analysis. A statement is defined as an
expression which denotes a complete idea and which has a referrent -- some
aspect of air pollution, in this case. Statements were recorded on cards
and coders independently classified them into either attitude or knowledge
categories and into qualitative similar sub-classes within each category.
A total of approximately 3, 200 attitude statements and 5, 350 knowledge
statements was available for consideration as items.
Empirical support of the model requires knowledge of content or inform-
ation presented in the mass media. The importance of this requirement can
be noted quickly. It is possible that one will not find a relationship be-
tween mass media exposure and knowledge when an empirical test is made
of the model. One explanation might be that items used to test such knowledge
were not related to information available in the mass media. By confining
one's attention to areas in which knowledge test items are constructed from
statements in the mass media, one can rule out this alternative explanation.
Attitude statements were divided into mutually exclusive sub-categories.
Items that were judged to be unambiguous and representative of various
aspects of the respective attitude categories and that fit the "agree-disagree"
type of response alternative were selected. The final list of potential attitude
items was composed of 50 statements. These items represented the measure
of the attitude variable in the pretest.
Knowledge statements were selected that were judged to be unambiguous
and representative of various aspects of the respective knowledge categories.
The final list of potential test items was composed of 223 statements. An
accuracy rating of these statements was performed by a panel of experts in
the air quality professions from the National Air Pollution Control Administration,
-------
56
Raleigh and Durham, North Carolina; and from professionals in Oregon and
California. A total of 65 raters was asked to cooperate; 57 completed and
returned their ratings.
Critical ratios were calculated for each item. A significant CR (at the
. 05 level) was established .as a necessary condition for acceptance of an item
in the knowledge test battery. Representativeness of each knowledge category
and ease or appropriateness in an interview setting were other conditions
necessary for acceptance. The final set of knowledge items pretested was
composed of 40 items --20 true and 20 false.
Experience and Use. Intensive interviews of a sample of the public
(40 in Eugene, 10 in Salem) were examined for statements concerning the
public's experience with air pollution and use of the environment. Probes
were directed toward these two areas. These interviews yielded several
statements about past living and visiting experiences in areas where the
respondents believed air quality problems existed and about use of the out-
doors for work, leisure and living. Items were developed from these re-
sponses for pretesting as measures of experience and use.
Fvaluation of Pretest Measures. A pretest interview schedule was
developed from items representing measures of the several variables. The
schedule was administered by personal interview to a random sample of 100
adults in the Portland (Oregon) Metropolitan Area in February 1970.
Responses were coded and scores key-punched for analysis. Analysis
consisted, first, of selection of items from among the 50 attitude state-
ments. Then estimates of reliability and the construction of parallel or
-------
57
comparable halves (forms) were completed for the items measuring attitude
and knowledge.
Fifty attitudinal items were subjected to an item analysis. Twenty-
five respondents with the highest total scores and 25 with the lowest total
scores were selected. These two groups served as criterion groups for
evaluation of individual items. The following ratio was calculated for
each item:
X, - X
1 = ~ (5)
"h "l
where
X is the mean score on a given statement for
the high group,
X is the mean score on the same statement for
the low group,
2
S is the variance of distribution of responses
of the high group to the statement,
S is the variance of distribution of responses
of the low group to the statement, and
is the number of responde
low groups, respectively.
n , n is the number of respondents in the high and
Parallel or comparable halves (forms) were used to avoid effects of memory
transfer when individuals were remeasured in the full study. The procedure
for item selection suggested by Horst [18] was followed. Form reliabilities
reported in Table 1 may be inflated since they were computed from the same
data used to construct the forms.
-------
58
The value of t is a measure of the extent to which an item differentiates
between high and low groups. If a t_ equaled 1. 75 or greater, it was considered
evidence that the average response of the high and low groups to a statement
differed significantly. This analysis resulted in a deletion of 10 items with
low t_ values.
Estimated Spearman-Brown split-half (odd-even) reliability coefficients
for the several measures are presented in Table 1.
Table 1. Estimated Spearman-Brown Split-Half Reliability Coefficients
for Pretest Measures of Variables in the Theoretical Model
No. of Reliability
Variable Items Coefficient
a. Attitude toward air quality
(Form A) 20 . 83
b. Attitude toward air quality
c.
d.
e.
f.
g-
h.
i.
j-
k.
(Form B)
Level of exposure to mass media
Level of air quality knowledge
(Form A)
Level of air quality knowledge
(Form B)
Level of exposure to interpersonal
information sources
Level of experience with air
pollution
Level of social influence
Level of education
Level of formal science education
Level of use of the environment
20
3
20
20
2
3
5
1
19
16
, 78
. 44
. 70
. 80
.76
. 30
. 75
. 92
. 77
Reliability estimates for measures of mass media exposure and experi-
ence with air pollution are low and suggest a need for item revision. An
examination of items measuring both variables disclosed evidence of ambigu-
ous wording which would likely produce responses based on different frames
-------
59
of reference. New items were written in an attempt to overcome this difficulty,
resulting in an increased number of items for measuring both variables.
An item analysis also was completed in an attempt to increase knowledge
test reliability. Guliksen [16], for example, has shown that the test mean,
standard deviation and reliability may be estimated from item parameters of
difficulty and reliability index. An attempt was made to increase test relia-
bility lay/making the average item variance smaller or the average item re-
liability index larger. This practice did not improve test reliability appreci-
ably. Details are available in a separate publication (see Mason [36] ).
Execution of the Sample Survey
Pretest questions surviving tests of reliability and item analysis consti-
tute the basis for accepting items for use in the sample survey. Summed
response scores of these items represented measures of variables in the
model. The new interview schedule constructed for the sample survey was
pretested itself to insure that its make-up maintained rapport between inter-
viewer and respondent. The final interview schedule employed in the sample
survey is presented as Appendix B.
Parallel or comparable forms had been developed for attitude and know-
ledge measures. As well, response alternatives were reversed for the third
endogenous variable, mass media exposure and items reordered randomly in
the battery. This produced a permutation of forms, or batteries, as measures
of the three endogenous variables. These permutations were employed in
order to randomize effects between forms and time of interview. The number
3
of permutations was eight (2 - 8). Since both Eugene and Salem were divided
into eight census tracts, one combination of forms was assigned at random to
each census tract for the first, or initial, interview.
Sampling of Respondents. A two-stage area sample was employed for
random selection of respondents from each community. Sample size was
-------
60
planned to achieve results having a specified degree of precision, viz. ,
n =
where
n is the sample size,
k is the constant associated with the
probability that the sample «alue will
have a relative error no greater than
±D>
V is the coefficient of variation, and
D is the relative error allowed.
The value of k was arbitrarily set at 3; V, calculated from pretest;.
data, averaged 0. 30 or less for variables in the model and D was arbitrarily
established at 0. 06. Solving for n in Equation (6) resulted in a sample size
of 250 for each community. A design effect of 1. 20 was assumed, raising
the sample size to 300. A response rate of 0. 85 for the first interview and
of 0. 77 to 0. 80 for the second was also assumed. Application of these rates
increased the sample size so it ranged from 444 to 464 for each community.
The latest metropolitan area maps were purchased from the office of the
local planning commission for each community. Each map was divided into
eight areas, following census boundary lines. It had been arbitrarily deter-
mined to select four households at random from each block in an effort to
reduce effects of clustering or intraclass correlation. Random selection of
14 blocks from each census area would yield 448 households per city for
interviewing -- within the range calculated in Equation (6).
Both metropolitan areas received an on-site inspection. Non-residential
and new residential blocks were noted as were blocks that were above- or
below-average in number of dwelling units (DU's). Non-residential blocks
were eliminated prior to numbering and new residential blocks added.
Above-average blocks were divided and below-average blocks were combined,
all in an effort to adjust block size to a calculated average number of dwelling
-------
61
units. Blocks were numbered in serpentine fashion, and 14 blocks were
selected at random from each census area, seven blocks for each pair of
interview "waves".
Sampled blocks received an on-site visit to 1) ascertain the number of
valid (L e. , occupied) dwelling units, 2) select at random a point to begin
selection of DU's, and 3) select, by use of a table of random permutations,
four dwelling units. Dwelling units within apartments or other contrived
blocks were selected randomly as well. Addresses and distinguishing features
at each DU were noted for preparation of interviewer listing sheets. One
adult respondent was selected at random wibhin each household by the inter-
viewer, using a "face sheet" or "cover sheet" commonly employed for such
purposes. Proper use of the "face sheet" guaranteed random selection of a
respondent within each household.
Completion of Interviews. The first interviewing wave began May 15,
1970, and terminated June 2, 1970. This was considered the first "high"
visibility condition. A total of 224 household addresses was specified in
each city for each interview. Oversampling was planned in order to secure,
if possible, more than 150 completed interviews. Panel attrition may have
been high for the subsequent re-interviewing with Schedule B. Interviews
for the "low" visibility condition began August 26, 1970, and were completed
September 16, 1,970. The third (and final) interview wave began January 5,
1971 and terminated January 16, 1971. This was considered the second
"high" visibility condition.
Response rates for Schedule A were calculated by dividing the number of
successful completions by 224. These are reported in Table 2. Response
rates for Schedule B were the number of usable schedules divided by the
number of usable Schedule A's.
See Page 32 for the interview design employed to test effects of changes
in air quality.
-------
62
Table 2. Response Rates for Interviewing
Interview
Schedule A
Schedule B
May
1970
77
Salem
September
1970
87
83
January
1971
%
87
May
1970
87
Eugene
September
1970
78
87
January
1971
%
87
The percent not-at-home and refusals are presented in Table 3.
Table 3. Percent Not-at-Home and Refusals
Salem Eugene
May September January May September January
Interview 1970 1970 1971 1970 1970 1971
Schedule A
Schedule B
14(9)* 7(6)
9(8)
6(7) 11(11)
8(5) 7(6)
9(4)
*
Percentages in parentheses refer to refusals.
The not-at-home rate for Salem in May and the refusal rate for Eugene
in September, both with Schedule A, appear high. An examination of block
clusters assigned to interviewers did not reveal a pattern of not-at-home1 s
for S;alem. They appeared to be distributed uniformly across all blocks.
The distribution of refusals in Eugene, however, was concentrated in blocks
surrounding the University of Oregon and interviewers noted the cause of re-
fusal in some instances as a saturation of previous interviewing by University
students.
Any effect of Schedule A non-response may be partially reduced by the
definition of the sampling universe. It was defined as residents who live- in
their respective dwelling units for six consecutive months or more after first
-------
63
contact. A set of items was developed and used for predicting early in
Schedule A if a respondent was likely to move within the next six months. If
it appeared that he was likely to move, the interview was terminated.
Since these "likely-to-move" respondents were not defined in the
sampling universe, other dwelling units in the blocks where these respond-
ents lived were sampled at random in an effort to build up the level of
completed interviews. It turned out that "likely-to-move" households were
correlated within their respective blocks. It was not always possible to
complete four interviews in these blocks. No more than four "likely-to-
move" respondents in such blocks were included among the "successful
completions" for the calculation of Schedule A response rates.
Assuming that measures to determine length of residence were valid, it
is likely that some not-at-home1 s or refusals would include some people not
defined in the population. If so, any non-response bias would likely be re-
duced commensurately. But, without confirming information, it is safer to
state that some non-response bias is likely to exist for the Schedule A inter-
views. The direction and extent of this bias are not known. The Schedule B
response rate averaged 86 percent and effects of non-response bias from
reinterviewing sources are expected to be less than that for Schedule A.
Completion of Coding, Key-Punching, and Verifi.cation. The response
rates represented 673 usable Schedule A's and 579 usable Schedule B's.
Efforts employed to transfer the information in these questionnaires to data
processing cards .were, first, to code the questionnaires according to a pre-
determined code key. Next, two coders independently checked each question-
naire for accuracy. Third, data were key-punched and verified. Fourth, a
print-out was obtained for all data and was proofread and rechecked with the
questionnaires. Fifth, a computer column displacement program was applied
and discrepancies sight-checked. Finally, a second computer print-out was
proofread and rechecked with the questionnaires.
-------
64
Analysis of Data
Analysis consisted of, first, an estimate of reliability for measures
representing variables of the model, and, second, a test of significance
for variables in the model by two-stage least squares.
Estimates of Reliability. Spearman-Brown split-half analyses were
completed for scores representing variables in the model, except for the
measure of attitude change. In this case, Formula (7) was applied:
r + r - 2r
- xx yy xy m
r j j ~ T ~T I''
dd 2 - 2r
xy
where
r is the estimated reliability of the attitude
dd i
change scores,
r is the estimated reliability of the Form A
scores,
r is the estimated reliability of the Form B
yy j
scores, and
r is the Form A, Form B correlation coefficient.
xy
Estimates of reliability for variables in the model are given in Table 4.
Table 4. Estimates of Reliability for Variables in the Model
Variable
Attitude change (20 items vs. 20 items)
Attitude level (40 items)
Mass media exposure (newspaper,
magazine, T"V)
Knowledge (40 items)
Interpersonal exposure (neighbors, friends)
Experience with air pollution (5 items)
Social influence (5 items)
Level of science education (19 items)
Use of the environment (16 items)
Reliability
Eugene
. 45
. 90
. 71
. 81
- 79
. 33
. 54
. 91
. 73
Coefficient
Salem
. 52
. 91
. 65
. 75
. 73
. 19
. 58
. 94
. 71
-------
65
Items employed for measurement of the "experience with air pollution"
variable did not intercorrelate particularly high. Thus, the reliability coef-
ficient was low. It was considered too low for the variable to function in the
analysis of the model. Therefore, this variable was dropped from further
consideration. Reliability estimates of other variables were considered
sufficiently high for their inclusion in the subsequent analysis.
Two-Stage Least Squares. Equations (1), (2), and (3) have been written
(Page 28) to describe the theoretical model. As a number of scholars have
demonstrated (e-_g. , Johnston [22] , Valvanis [ 53] ), least squares applied
to each of the above equations will lead to estimates of the parameters which
are biased and inconsistent. In addition, such estimators are less efficient
than those obtained by alternative procedures.
For example, consider two- stage least squares:
where
TT.. = f(P..; \ ..), and
ij iJ ij
v. = g(e.; (3. .; v ..).
i i ij ij
Since the e.'s are normally distributed with mean equal to zero, finite
1 2
and constant variance (r . , and the v.'s are linear forms of e 's, then the
iL i
v.'s themselves are normally distributed, with mean equal to zero and finite
1 2
and constant variance
-------
66
Having estimated the functions in Equations (8), (9) and (10), one
proceeded to obtain estimated values for the endogenous variables, Y,, Y_,
A A A l *
and Y , and these will be denoted as Y , Y , Y . The next step was to fit
the function:
Yl= PllPl2*HXl + "' + *12X2 + el (H)
Y2 = P21Y1 + B22Y
by ordinary regression techniques.
The resulting estimators, the (3 ' s and the -y 's are biased; but are con-
sistent and more efficient than straightforward least squares estimators.
Having completed this routine, and having acquired estimators of the
several (3's and v 's in the system, and knowing the values of X ' X,, it is
possible to solve the original system (Equations (1), (2), and (3)), for the
values Y,, Y_, and Y_. To distinguish these estimates from those
J. Ct o
obtained from the reduced form (Equations (8), (9), and (10)), they will be
. /\ A
ft A A
indicated as Y^ Y_, and Y , and referred to as structural equations:
Therefore, parameters of the structural equations were estimated by the
method of two-stage least squares. Significance of constants in these equations
was tested by Student's ^, recognizing, however, that some degree of bias
is involved.
By comparing scores of individuals in the two communities, the null
hypothesis concerning the difference of each endogenous (Y.) variable between
communities was tested.
-------
67
RESULTS AND DISCUSSION--A TEST
OF THE AIR QUALITY COMMUNICATION MODEL
Results in this section are divided into four phases:
1. Results of accuracy of air pollution perceptions.
2. Empirical test of the communication model.
3. Empirical test of the revised communication model.
4. Discussion of results.
Results of Accuracy of Air Pollution Perceptions
A condition stressed throughout this study has been the selection of an
air pollution situation which has "impinged on the consciousness of people
in the community selected for study". An operational corollary to this
condition is the contention that people are able to compare visibility levels
from one time period to another and that the substance of this comparison
can be ascertained by survey methods. While visibility differences may well
exist between specified time periods, as measured by trained observers,
there is no assurance that accurate perceptions will be distributed uniformly
among people in the community. It may well be that the communication
process concerning air quality for those who hold accurate perceptions will
differ from those who hold inaccurate perceptions or no perceptions at all.
Before testing the communication model it may be profitable to examine the
nature of perceptions concerning visibility to determine if analysis of the
model should be adjusted to include effects of perceptual differences.
An opportunity existed to examine such differences in this study. Per-
ceptual accuracy was determined by scoring responses to one question asked
at the end of the second (Schedule B) interview. In September, for example,
this question was:
"Within the last three weeks, would you say that the air pollution
in Salem (Eugene) is better or worse than when we talked to you
earlier this year in May?"
Responses to this question were compared to the analysis of visibility obser-
vations. A "worse" response in the September interview and a "better"
-------
68
response in the January interview, for example, were considered accurate.
Other response codes employed were "no difference" and "don't know" but
these were not used to determine perceptual accuracy or inaccuracy. These
responses are summarized in Table 7.
Table 7. Summary of Perceptions to Visibility
Perception of
Visibility
*
Accurate
Inaccurate
No difference
Don't know
TOTAL
(N)
May-September 1970
Eugene
52
31
14
_ 3
100
(152)
X2 = 15. 58, 3 df;
Salem
53
15
29
3
100
(139)
p< . 01 X2 =
Sept. 1970 - Jan. 1971
Eugene
72
6
18
__ 4
1001
(142)
13. 76, 3 df; p<
Salem
51
12
31
6
100
(146)
. 01
"Worse" response in May-September interview, better response in
September-January interview.
"Better" response in May-September interview, "worse " response in
September-January interview.
Accuracy of perceptions concerning visibility varied by community (location
effect) and by time of interview (timing effect). A statistical analysis of the
data in Table 7 suggests a highly significant location effect, if one sums the
Chi squares and degrees of freedom presented in the table. Switching the
independent variables, _i. _e. , controlling for location effects and testing for
timing effects, resulted in a non-significant Chi square for Salem respondents
2
but a highly significant Chi square (X = 28. 90, 3 df; p<. 001) for Eugene
respondents. Testing for a possible location x timing interaction effect in-
volved the routine suggested by Snedecor [43] . This resulted in a pooled
Chi square of 11. 67, with 3 degrees of freedom, significant at the . 01 level.
Thus, much of the variation in perceptual accuracy can be accounted for by
three effects --a location effect, a timing effect (for Eugene only) and an
interaction effect. Location effects had been anticipated, given the history
-------
69
of air pollution in the Eugene area. Timing and interaction effects, however,
wero a surprise. There seemed to be little theoretical basis for predicting
that Eugene should experience a higher level of inaccurate responses in
September than in January. Responses to an open-ended item asking re-
spondents to describe how or in what way they believed air pollution was
better were studied for those interviewed with Schedule B in September.
A sample of verbatim responses associated with accurate and inaccurate
perceptions is presented as Appendix D.
It became clear that one reason for a "better" response was an evaluation
of attempts by the Oregon Department of Environmental Quality to manage
smoke emissions during the 1970 field burning season. Daily acreage burning
quotas were established with burning permitted when winds would likely blow
smoke away from the Eugene area. The Eugene Register-Guard carried
several articles on this management proposal prior to the first interview in
May. It is conceivable, therefore, that facts concerning the forthcoming
management of field burning became incorporated as part of the definition of
the situation for many respondents. If such were the case, some respondents
may well have ignored the May-September time frame posed by our perception
question. Instead, they imposed their own defined time frame which included
a comparison of 1970 management efforts with the pollution episode of the
previous year. If this were the case, a "better" response would lead to
spuriously inaccurate responses in the September Schedule B interview and
to spuriously accurate responses in the January interview. In the first instance,
respondents would be wrong for the wrong reasons; in the latter instance,
right for the wrong reasons.
All September and January Schedule B open-ended responses were
examined to test the definitional hypothesis. Responses were studied for
statements concerning the management (or lack of management) of field
burning emissions. If the definitional hypothesis is to be supported, one
would expect fewer statements among Salem respondents, compared to
Eugene respondents. Smoke emissions from field burning were less an issue
in Salem than in Eugene. Next, a 'better" response would be associated with a
-------
70
positive evaluative statement, je._g. , that field burning was being effectively
controlled, compared to the previous year. As well, a "worse" response
would be associated with a negative evaluative statement, _e. jj- , that field
burning was not being controlled. No other response combination should be
present; _L^. , one should not find a "worse" response associated with a
positive evaluative statement nor a "better" response associated with a
negative evaluative statement. The data are presented in Table 8.
Table 8. Relationship Between Perceived Management of Field Burning
Smoke Emissions and Accuracy of Visibility Perceptions
August-September 1970 January 1971
5|C #5{C SJCSfS 5JC
Level of Accurate Inaccurate Accurate Inaccurate
Management Eugene Salem Eugene Salem Eugene Salem Eugene Salem
Field burning
controlled
Field burning
not controlled
Total number
of respondents
0
10
(79)
0
5
(73)
40
0
(47)
24
0
(21)
27
0
(103)
7
0
(75)
0
22
(9)
0
6
(18)
0£
"Worse" response.
>'f>k
"Better*1 response.
The data support the definitional hypothesis. In all cases more Eugene
respondents cited evaluative statements than Salem respondents. Moreover,
it does appear that timing and interaction effects were produced, in part,
by the application of a pre-defined time span in response to our air quality
perception question. Thus, any attempt to measure the accuracy of perceptions
in Eugene is likely to be confounded by past definitions of the situation. Such
confounding raises questions concerning the validity of the method employed for
determining perceptual accuracy in Eugene. This confounding is not apparent
for the measurement of perceptual accuracy in Salem. Effects of perceptual
differences will be considered prior to testing the communication model in
Salem.
-------
71
An additional source of potential error also was examined. Known
visibility levels were employed for scoring accuracy of responses concerning
perceptions of air quality. It may well be that visibility was not employed by
some respondents in their evaluation of air quality. Odors and health effects
are some of the possible alternatives, for example. Open-ended responses
were coded for source and effect of air pollution mentioned. Analysis dis-
closed that visibility or visibility-related phenomena were cited by 92 percent
of the Eugene respondents and 91 percent of the Salem respondents who prof-
fered perceptual judgments. It appears that visibility was, indeed, an aspect
of air quality that impinged on the consciousness of most of these respondents.
Empirical Test of the Communication Model
Preliminary analysis of the model involved the plotting of scatter dia-
grams between each exogenous and endogenous variable prior to analysis by
two-stage least squares. The purpose of such diagrams was to ascertain
visually the order of polynomial fitting the data. It was determined from
this inspection that the relationship between all variables was linear and
linear terms were employed for all reduced-form equations.
Measurable differences in perceptual accuracy were found among Salem
respondents. It may well be that the communication process for those who
hold accurate perceptions will differ from those who hold inaccurate or who
hold "no difference "perceptions. A test of the appropriate residual sums of
squares was made to determine if data for the three perception subgroups
in Salem could be pooled and if the data for both Eugene and Salem could
be pooled as well. The procedure Ostle [36] describes was followed.
Pooling of data for perception subgroups in Salem was tested first. Then,
pooling of data for Eugene and Salem groups was tested.
When several sets of sample data are available, one is interested in
determining if a single regression line can be used for all the data. The
'I am indebted to Dr. Norbert Hartmann, Department of Statistics at Oregon
State University, for suggesting this procedure and for completing the
calculations.
-------
72
first step is to calculate the total residual sum of squares for each reduced
form equation (attitude, mass media use, knowledge) in the model for Salem.
The second step is to calculate the residual sum of squares of each perception
subgroup for each reduced form equation of the Salem data and to sum these
residual sums of squares. The third step is to subtract the residual sum of
squares calculated in the second step from the total residual sums of squares
calculated in the first step. One now has obtained two residual sums of
squares -- one from Step two and one from Step three. Each of these residual
sums of squares is divided by its respective degrees of freedom. Finally,
an F-ratio is calculated for each equation with n] and n degrees of freedom.
If this F-ratio is statistically significant, the regression line for that
equation differs among the different subgroups. If the F-ratio is not sig-
nificant, regression lines for each subgroup are considered the same, and
the data can be pooled.
The F-ratios for the different Salem perception subgroups and for the
Eugene and Salem groups are presented in Table 9-
Table 9. Analysis of Variance, Degrees of Freedom and Significance Levels
for Residual Sums of Squares of Equations of the Communication
Model
Equations
Combining over levels of
perception for Salem:
Attitude (Y.)
Mass media exposure (Y_)
Knowledge (Y,)
Combining over communities:
Attitude (Y.)
Mass media exposure (Y,)
Knowledge (Y )
F-Ratio
1. 98
1. 24
<1.00
<1. 00
1. 34
3. 34
Degrees of
Freedom
14,
14,
14,
7,
7,
7,
264
264
264
565
565
565
Significance
Level
NS
NS
NS
NS
NS
p< .01
-------
73
The data show that one can pool the data for perception subgroups in Salem,
but cannot pool the data for Eugene and Salem. The F-ratio associated with
the knowledge equation is statistically significant.
Multiple correlations and zero-order correlation Coefficients between
each exogenous and endogenous variable are presented in Table 10.
Table 10. Multiple Correlation Coefficients and Zero-Order Correlations
Between Each Exogenous and Endogenous Variable*
Mass Media
Attitude Use Knowledge
Exogenous Variable (Y.) (YJ (YJ
L C* .3
Eugene
Interpersonal discussion (X ) . 447 . 520 . 497
Social Influence (X3) . 382 . 379 . 425
Education (X) .351 .190 .436
Science education (X ) . 344 . 238 . 470
Use of environment (XJ .272 .262 .362
b
Multiple correlation coefficient , 591 . 534 . 622
Salem
Interpersonal discussion (XJ . 416 . 510 . 400
Social Influence (X ) . 308 . 362 . 342
Education (X4) .281 .244 .421
Science education (Xg) . 253 . 277 . 419
Use of environment (X,) .274 .210 .373
b
Multiple correlation coefficient . 459 . 524 . 538
*p<.05 = . 113; p< . 01 = . 148.
All the correlations in Table 10 are significant and the difficulty this
poses will become apparent when the model, analyzed by two-stage least
squares, is presented. The structural form of the model is shown in
Equations (19) to (24). Numbers in parentheses under the coefficients are
-------
74
Student's t_ values: (*) denotes . 05
-------
75
Equation (21) of the Eugene model (set of equations). An alternative pro-
cedure was employed to reformulate the model. To begin with, chances are
good that low observed t_ values were produced, in part, by the effects of multi-
collinearity, JL. e_. , at least one variable in an equation was adding little to
the explained variance that was not already accounted for by other variables.
One can check for this possibility by examining the ratio of the main diagonal
element of the inverse matrix corresponding to the ivariable and the re-
ciprocal of the corrected total sum of squares of the i variable. The ideal,
but uncommon, situation is for this ratio to equal unity; however, unity can
occur only if all other explanatory variables in the model were completely
uncorrelated with the i-
sociological variables.
uncorrelated with the i variable. This rarely occurs in practice with
It can be shown that the relationship between the i ratio and the
variance of the i variable accounted for by other explanatory variables in
the model corresponds to the function:
D. = -
1 1 - R2
ij
where
D. is the ratio of the main diagonal element
for the it*1 variable, and
R.. is the variance of the i*" variable accounted
^ for by all other explanatory variables.
Thus, when D. is large, the variance of the i variable is already
largely accounted for by other explanatory variables in the model. Variables
with relatively high D. values will tend to have large variances of their esti-
mated coefficients; thus, very low precision of estimation will result -- even
to the point that one is not sure of the sign of the coefficients. The values
2
of Student's t, D., and R.. obtained for variables in the model are pre-
- i ij
sented in Table 11. Variables with high D. values were usually associated
with low Student's t_ values. Moreover, the occurrence of a large
number of relatively high ( i. e. , greater than 10) D. values was considered
'I am indebted to Dr. William G. Brown, Professor of Agricultural Economics
at Oregon State University, for suggesting this procedure.
-------
76
Table 11. The Values of Student's Jt_, D. and R.. Associated with
Variables in Each Model * IJ
Equation Variable
(19) Xx
Y^
Y3
(20) X3
X4
A
Yl
(21) X5
X6
A
Yl
Y2
(22) Xj
Y2
Y3
(23.) X3
X4
A
(24) X5
X6
*1
*2
Student's t_
-0. 049
0. 346
3. 350
-0. 384
-3. 886
7. 379
1. 833
1. 105
2. 523
-0. 331
1. 577
-0. 916
2. 149
0. 980
-1. 363
6. 216
1. 296
1. 133
2. 268
0. 034
D.
L
53. 659
78. 538
5. 803
2. 289 "
1. 935
3. 588
3. 153
3. 945
15. 585
85. 523
35. 700
58. 880
12. 370
1. 829
1. 600
2. 360
7. 346
2. 183
13. 928
105. 429
R2
. 981
. 987
. 828
. 563
. 483
. 721
. 683
. 747
. 936
. 988
. 972
. 983
. 919
. 453
. 375
. 576
. 864
. 542
. 928
. 990
unacceptable. Exogenous variables associated with high D. values were noted,
and intercorrelations among exogenous variables examined. The X exogenous
variable, interpersonal discussion, had the highest D. value of all exogenous
variables; as well, this variable was the only one correlating significantly
with all other exogenous variables. And, as noted in Table 10, this variable
correlated highest with all endogenous variables, except for knowledge (Y,)
in Salem. It quickly became apparent that this exogenous variable was
-------
77
improperly specified and was likely the source of difficulties encountered in
the test of the model. The interpersonal discussion variable should have
been endogenous.
A new model was formulated. Interpersonal discussion was considered
a new endogenous variable. Social influence was specified as exogenous.
Level of formal education was deleted (it correlated . 80 or higher with level
of science education) and a new exogenous variable, perceived seriousness
of air pollution in a respondent's community of residence, was added.
This variable was postulated to be positively related to attitude towards air
quality. Thus, the new model contained four exogenous and four endogenous
variables.
Conclusions based on significance tests of the reformulated model,
however, must be viewed with caution. The equations were developed after
data from the previous model had been examined. Such selection tends to
capitalize on relationships which might be high because of sampling fluctu-
ations. This research must be considered exploratory, due to the tentative
nature of the theoretical model. Previous work cited to support the model
usually employed comparisons between two variables considered one at a
time; no previous work reported has attempted to estimate the more general
relationships specified among all the variables simultaneously. When con-
fronted with this type of situation, it is often necessary to experiment with
different hypothesized systems. Consequently, one must be particularly
cautious in interpreting tests of significance for the reason cited earlier.
Data from the previous model will be employed to test the reformulated
model. This test is less meaningful than if new data had been acquired.
Empirical Test of the Reformulated Communication Model
Pooling of data for perception subgroups in Salem and of data for
Eugene and Salem groups was completed in a manner similar to that for the
original model. The results are presented in Table 12.
-------
78
Table 12. Analysis of Variance, Degrees of Freedom and Significance
Levels for Residual Sums of Squares of Equations for the
Revised Communication Model
Model
F-Ratio
Degrees of
Freedom
Significance
Level
Combining over levels of
perception for Salem:
Attitude (Y )
Mass media exposure (Y_)
Li
Knowledge (Y )
Interpersonal discussion (Y )
3.
1.
1.
1.
18
10
43
15
10,
10,
10,
10,
270
270
270
270
p<. 01
NS
NS
NS
The data show that one cannot pool the data among the Salem subgroups.
The F-ratio associated with the attitude equation is statistically significant.
Further tests among the different residual sums of squares indicated that
one could pool the inaccurate and "no difference" perception subgroups.
Note that it would be inappropriate to test for combining over communities,
since the equation for attitudes (Y.) in Salem was significant.
Multiple correlations and zero-order correlations between each exogenous
and endogenous variable of the reformulated model are presented in Table 13.
It was noted in Table 10 that all correlations were statistically significant;
'however, some correlations in Table 13 are not significant. As well, several
of the intercorrelations among the exogenous variables of the reformulated
model are not significant. Thus, the magnitude of the difficulty encountered
in testing the original model is not likely to be repeated in the test of the re-
formulated model.
-------
Table 13. Multiple Correlation Coefficients and Zero-Order Correlations
Between Each Exogenous and Endogenous Variable
Mass Media
Endogenous Variable
Social influence (X )
Science education (X )
Use of environment (X,)
o
Seriousness (X )
Attitude
'V
Eugene*
. 382
. 344
. 212
. 361
Use
. 379
. 238
. 262
. 124
Knowledge
.425
. 470
. 362
. 207
Peer
Discussion
!no<.differenc.e"perceptions)
Social influence (X3) . 375 . 432 . 431 . 564
Science education (X) .272 .295 .464 .250
Use of environment (XJ .409 .312 .434 .375
b
Seriousness (X ) . 349 . 310 . 209 . 281
Multiple correlation coe
coefficient .540 .410 .588 .617
*#
Salem (accurate perceptions)
Social influence (X )
Science education (X )
Use of environment (X,)
o
Seriousness (X )
Multiple correlation
coefficient
p<. . 05 = 113; p< . 01 = . 148
**p < . 05 = . 159; p<. 01 = . 208
258
235
154
309
384
. 292
. 254
. 123
. 221
. 359
. 260
. 373
. 304
. 258
. 450
. 509
. 236
. 279
. 187
. 530
-------
80
The structural form of the revised model is presented as Equations (26) -
(37). Equations (26) - (29) are for Eugene; Equations (30) - (32) for Salem
respondents who held either inaccurate or "no difference" perceptions;
*
Equations (34) - (37) are for Salem respondents who held accurate .perceptions.
Again, numbers in parentheses under the coefficients are Student's Rvalues:
(*) denotes . 05
Interpersonal
discussion
Use of
environment
(X6)
Seriousness of
air pollution
Mass media
use (Y2)
Science
education
(X5)
Figure 23. Diagram of the communication model for Eugene.
-------
81
In Equation (26) Air Duality Knowledge (Y ) and Perceived Seriousness
5 A
of Air Pollution Ln Eugene (X_) are significantly related to Attitude Level (Y )
A
while Interpersonal Discussion (Y ) is not. In Equation (27) Air Quality
A A
Knowledge (Y,) and Interpersonal Discussion (Y ) are significantly related
A A
to Mass Media Use (Y_). In Equation (28) Mass Media Use (Y,), Use of
£* L*
the Environment (X,) and Science Education (X ) are significantly related to
A
Air Quality Knowledge (Y ). In Equation (29) Attitude Level ($) and
Social Influence (X0) are significantly related to Interpresonal Discussion
(Y4) while Air Quality Knowledge (Y ) is not.
The next set of equations represents the communication model for Salem
respondents who held inaccurate or "no difference" air quality perceptions:
Y = 75. 5231 -1- 2. 6610 X + 1. 4089 Y + 2. 3386 Y (30)
(2.006)** (1.888)** (1.400)*
^ A
Y2 = -3. 8072 + 0. 3042 X + 0. 1262 Y (31)
(1. 691)** (3. 573)***
^ A
Y = 5. 0937 + 0. 1344.X + 0. 1475 X, + 0. 9432 Y- (32)
(2.770)*** (2.043)** (2.787)***
Y = -0. 1090 + 0. 8132 X + 0. 2381 Y - 1. 2257 Y (33)
(2.487)*** (1.861)** -1.232)
Social
influence
\
Mass media
use (Y2)
Social
(X3)
/
Knowledge
f3)
Science
(X5)
Figure 24. Diagram of the communication model for Salem respondents with inaccurate or 'no
difference "air quality perceptions.
-------
82
A
In Equation (30) Air Quality Knowledge (Y.J, Interpersonal Discussion
A *
(Y ) and Perceived Seriousness of Air Pollution in Salem (X ) are significantly
A ' A
related to Attitude Level (Y ). In Equation (31) Attitude Level (Y.) and
/\
Social influence (X ) are significantly related to Mass Media Use ^Y~).
A
In Equation (32) Mass Media Use (Y_), Science Education (X_) and Use of-
& 5
the Environment (X,) are all significantly related to Air Quality Knowledge
(Y3). In Equation (33) Attitude Level (Y ) and Social Influence (X ) are
significantly related to Interpersonal Discussion (Y ) while Mass Media
Use (Y_) is not.
The final set of equations represents the communication model for
Salem respondents who held accurate air quality perceptions:
w
YI = 90. <
V -
Y3=-5.I
V ' =
Social
i '
Interpersonal
785 + 2. 5098 X + 1. 7735 $
(1. 787)** (2. 606)***
* A
5974 + 0. 2464 Y + 0. 3846 Y
(1. 768)** (1. 447)* 4
787 + 0. 0955 X + 0. 1034 X + 0. 19
(1. 751)** (1. 670)** (2. 51
A
i990 + 0. 5212 X + 0. 1271 Y
(5.004)*** (1.480)*
Attitude ^
.
^\ /
^ Knowledge
(34)
(35)
40 Y" (36)
6)***
(37)
Seriousness
-------
83
In Equation (34) Air Quality Knowledge (Y ) and Perceived Seriousness
of Air Pollution in Salem (X ) are significantly related to Attitude Level (Y ).
In Equation (35) Air Quality Knowledge (Y_) and Interpersonal Discussion (Y )
«J *
are significantly related to Mass Media Use (Y2). In Equation (36) Attitude
Level (Y ), Science Education (X ) and Use of the Environment (X,) are sig-
nificantly related to Air Quality Knowledge (Y ). In Equation (37) Air
Quality Knowledge (Y_) and Social Influence (X ) are significantly re-
lated to Interpersonal Discussion (Y ).
Discussion of Results
Results concerning statistically significant relationships among
endogenous variables will be discussed for each model. Then comparisons
will be made among the different models.
For Eugene respondents, the relationship between mass media use
and air quality knowledge is interdependent, L e. , one serves as the basis
for the other and vice versa. High knowledge levels are associated with
positive attitudes toward air quality which, in turn, are related positively
to interpersonal discussion. High levels of interpersonal discussion are
positively associated to high levels of mass media use.
For Salem respondents who held inaccurate or "no difference" air
quality perceptions, positive attitudes toward air quality were positively
associated with high levels of mass media use. Mass media use was posi-
tively related to air quality knowledge which, in turn, was related positively
to attitudes. An interdependent relationship was found between attitudes and
interpersonal discussion.
For Salem respondents with accurate air quality perceptions, an inter-
dependent relationship was found between attitudes and air quality knowledge.
High knowledge scores were positively related to mass media use and to
interpersonal discussion. Interpersonal discussion was positively related
to mass media use.
-------
84
Moreover, the postulated relationship between mass media use and
attitudes, as stipulated by the definitional hypothesis, is not supported
by the data. It may be more correct to state that mass media use is posi-
tively related to knowledge and knowledge is positively related to attitudes.
This type of relationship was found for all models.
It may be profitable to compare the different models, particularly since
the analysis of mass media content disclosed such wide differences in the
time of definition of air pollution as an issue in both communities. As well,
the role of mass media use appears to vary according to community and to
accuracy of perception. For example, an interdependent relationship was
found in Eugene between mass media use and air quality knowledge. A
dependent relationship was found between mass media use and knowledge
among Salem respondents holding inaccurate air pollution perceptions.
And, a dependent relationship was found between knowledge and mass media
use among Salem respondents with accurate perceptions.
What may well be going on -- as one views the different models along
a postulated time continuum --is the evolution of a single communication
process concerning this issue.
Initially, the issue is poorly perceived or defined with the communication
process functioning much like that of Figure 24. The direction of the arrows
for the endogenous variables suggest selectivity in information seeking.
Those with more positive attitudes toward the topic interact or discuss it
more with peers. As well, this interaction forms and reinforces their weak
but positive attitudes toward air quality. Those who hold positive attitudes
are greater users of the mass media and it is argued that this use is likely
to be biased or selective in favor of these weakly-held attitudes. People
tend to seek out information which supports their positive attitudes and to
avoid or misperceive messages which threated or fail to support them.
Greater mass media use, however, leads to higher levels of knowledge which
tends to form and reinforce positive attitudes.
-------
85
Next, after the passing of time, the accumulation of information from the
mass media has led to a higher level of knowledge. The communication
process now may resemble that of Figure 25. Mass media use is no longer
selective; _i. e_. , it is not a function of attitude level. Rather, attitudes are
well established and are now based primarily on level of accumulated know-
ledge. High knowledge levels now produce greater information seeking in
both the mass media and wi th other people. It is argued that selectivity
associated with the process in Figure 24 is likely to have been abandoned.
Now, information seeking is unbiased, with knowledgeable people having
formed well-differentiated and elaborate* cognitive structures concerning
2/
air pollution. Positive attitudes associated with these elaborate cognitive
structures are relatively impervious to change -- unless there is a significant
change in the cognitive structure itself. Incompatible information is not
avoided or misperceived, for example, but is likely to be sought out and
discounted through counter-argument. The mass media is used by both those
high in knowledge and those who are involved in interpersonal discussion.
It may well be that there is insufficient information available in the mass media
channels to meet the information needs of more knowledgeable people.
Otherwise, the direction of the arrow would likely go from mass media use
to knowledge.
Finally, the process may evolve to a system much like that of Figure 23.
The interdependence between mass media use and knowledge suggests a suf-
ficiency of information about air quality in the mass media channels and that
the more knowledgeable people are actively seeking out this information.
Student's t values were calculated for attitude and knowledge means between
Salem accurate and inaccurate or "no difference" . subgroups. Means of both
groups for both variables were statistically significant in the predicted
direction (accurate means were higher) at the . 05 level or less. More-
over, it is not likely that these differences were produced as artifacts of
differences in exogenous variables. Student's t_ values were also calculated
to test the significance of difference between means of each exogenous vari-
able for both groups. No difference was statistically significant.
2/
The term 'cognitive structure' refers to the way an individual organizes his
physical and social worlds, including all his facts, concepts, beliefs,
expectations and the pattern of their interactions about the topic.
-------
86
Moreover, higher knowledge levels support the more positive attitudes toward
air quality which lead to greater interpersonal discussion. Those who discuss
air pollution with others are also greater users of the mass media. Again,
it is argued that information seeking at this stage of the process is also un-
biased. Information inconsistent with or in refutation of existing knowledge,
for example, is likely to be verified with some authority. Inaccurate or in-
consistent public statements, for example, are likely to be challenged. And,
one is likely to find that the air pollution knowledge levels of citizens in a
community which has an attentive local mass media is remarkably sophisticated.
One can conclude that more than one communication model is function-
ing in a community at one point in time, depending on perceptions people
have about the environment. As well, whether or not different models
are functioning may well depend on how the local mass media, particularly
newspapers, define the situation. Mass media use appears to play a sig-
nificant role in the accumulation of knowledge which contributes to the
establishment of attitudes. The decisions of mass media gatekeepers appear
to be most relevant in determining the extent of knowledge the citizenry of a
community may have about the local aspect of a complex issue. Word-of-
mouth channels can function up to a point, but their ability to communicate
valid information of a technical nature is limited, compared to newspapers,
for example. The outcome of efforts to obtain clean air' for a community
may well rest, in part, with the decisions of these local mass media gate-
keepers in the selection of information they employ to define the issue.
It is doubtful if one will find this particular communication process at
work in all other types of air pollution situations. Consider, for example,
the situation where the community which receives both the economic benifits
and the externalities of a polluter and where a polluter and enforcer disagree
over the feasibility of controls. One is not likely to find a positive linear
relationship between attitudes and knowledge among community residents
over this issue. Instead, a curvilinear relationship is more likely -- those
who hold extreme positive or negative attitudes will have relatively high
-------
87
knowledge scores. Attitudes of those with low knowledge scores will likely
be neutral, L £. , tend to have "no opinions". In this instance the role of the
local mass media -- particularly newspapers -- may well have a bearing on
any empirical relationship between attitude and knowledge. For if the local
media "manage" the issue by not reporting it (or not reporting it fully), then
one will need to consider the effects of possible alternative mass media
sources, such as state newspapers, in the model. Otherwise, an empirical
test of the model may disclose a null relationship between mass media use
and knowledge. One decision required in developing such a model would be
to build in the effects of the "localness" of the media itself.
The role of exogenous variables should be considered as well. The role
of social influence and interpersonal discussion concerning air pollution can
be more precisely defined. It was recognized in the review of literature that
social influentials expose themselves not only to pertinent forms of the mass
media, compared to those less influential, but are involved more in inter-
personal discussion as well. Interpersonal discussion, however, plays a role
in attitude formation at the initial stage of the communication process. The
importance of interpersonal discussion may be attentuated once knowledge levels
and attitudes are established. To be sure, social influence is related to inter-
personal discussion at the latter stages of the communication process. But
interpersonal discussion appears to lead only to greater mass media use.
This use is related to higher knowledge levels. A direct relationship between
interpersonal discussion and either knowledge or attitudes is not supported
by the data. Thus, any information effort to improve attitudes concerning air
pollution should take into account the significant role of these influentials at
the initial stage of the effort.
The positive relationship between science education and knowledge, for
example, suggests that knowledge levels among the better educated are likely
to be higher than among the less educated. And, over time, the knowledge gap
between the highly- and poorly-educated segments of a community will likely
become larger, not smaller. Such a gap stems not only from higher initial
-------
88
knowledge levels but also from the use of the mass media as a learned skill
for obtaining information. Tichenor jet aL , [5l] have pointed out the statistical
regularity associated with an increasing knowledge gap for different educational
levels and the implications this has for the "mass" impact of the media; viz. ,
that of reinforcing or increasing existing inequities in public knowledge.
Thus, any program to employ the mass media in reaching large numbers of
citizens with information concerning air pollution, for example, is likely to
reach the already well-informed. This does not mean that the well-informed
will not profit from any new information effort. It suggests that these individuals
will be the primary beneficiaries of such an effort.
Use of the environment for recreation purposes, a significant predictor
of air quality knowledge at all stages of the communication process, also plays
a role a-s a source of attitude change as well. One question likely to arise
when using two-stage least squares is to compare the relative effects of all
the exogenous variables on each endogenous variable. For example, if a 1%
change occurred in each exogenous variable, how much of a percent change
would occur in each endogenous variable? Using level of attitudes as the
endogenous variable, an answer to this question showed that social influence
would produce the greatest change for Eugene and for Salem respondents who
held accurate air pollution perceptions. In these instances a 1% change in
social influence would change attitudes toward air quality by 10% for Eugene
respondents and by 7% for Salem "accurate" respondents. Effects of other
exogenous variables were relatively small. However, a change in use of the
environment for recreation was by far the greatest producer of attitude
change among Salem respondents who held inaccurate or "no difference" air
pollution perceptions. A 1% change in use of the environment would produce
a 16% change in attitudes. The implication is clear. One way to sensitize
respondents who hold inaccurate or "no difference" perceptions is to persuade
them to use the environment more for recreation. Individuals who make greater
use of the environment for recreation appear to hold more positive attitudes
'I am indebted to Dr. Albert N. Halter, Professor of Agricultural Economics
at Oregon State University, for suggesting this procecure and for completing
the calculations.
-------
89
toward air quality. In turn, they are able to more accurately compare visi-
bility differences and may see the relatively clean air of their recreation area
as an "escape" from the visibility restrictions they perceive at home.
There is little one can do to change the social influence of an individual.
Thus, producing changes in attitude by changing the level of this exogenous
variable would not be particularly effective for Eugene and Salem "accurate"
respondents. But, for Salem "inaccurate" respondents, efforts to change
attitudes toward air quality are likely to produce positive results. One has to
find some way to persuade these people to spend more of their recreation
hours out-of-doors away from their home and, perhaps, to spend more of their
time on recreation activities.
Finally, the belief that air pollution is serious -- based on the individual's
experience living or working in the community -- is a significant exogenous
effect that appears to have a bearing on attitudes at all stages of the communi-
cation process. Beliefs based on experience may well provide a basis for
the acceptance of one of several definitions of air pollution situations with
which an individual may be confronted. It may be redundant to assert that
the first step in improving air quality is for a community to recognize that it
has a serious problem. It is, nonetheless, a truism to state that efforts
to improve air quality are not likely to succeed without this recognition.
A pre-condition for the establishment of a successful air pollution control
program is that the citizens of an airshed recognize the scope and magnitude
of the problem. This pre-condition is satisfied when they perceive odors and
loss of visibility, for example, for what they are and in terms which they
believe are important to them.
-------
90
LITERATURE CITED
1. Gordon Allport and Leo E. Postman, "The Basic Psychology of Rumor, "
in Readings in Social Psychology, Eleanor E. Maccoby, T. M.
Newcomb, and E. L. Hartley, eds. , New York: Henry Holt,
pp. 54-65, 1958.
2. E. Jackson Baur, "Opinion Change in a Public Controversy, " Public
Opinion Quarterly, Vol. 26, pp. 212-226, 1962.
3. Kenneth Boulding, "The Economics of the Coming Spaceship Earth"
in Henry Jarrett, ed. , Environmental Quality in a Growing
Economy, Baltimore: The Johns Hopkins Press, 1966.
4. James E. Brinton and L. N. McKown, "Effects of Newspaper Reading
on Knowledge and Attitude, " Journalism Quarterly, Vol. 38,
pp. 187-196, 1961.
5. Charles F. Cannell and James C. MacDonald, "The Impact of Health
News on Attitudes and Behavior, " Journalism Quarterly, Vol. 33,
pp. 315-323, 1956.
6. Rachel Carson, Silent Spring, New York: Houghton Mifflin, 1962.
7. Census of Population, Advance Report, PC(VI) - 39, Oregon, U. S.
Department of Commerce, Bureau of Census, December 1970.
8. James S. Coleman, Community Conflict, Glencoe, 111. The Free
Press, 1957.
9. Barry Commoner, Science and Survival, New York: Viking, 1967.
10. Lewis Coser, The Functions of Social Conflict, New York: The
Free Press, 1956.
11. M. Jay Crowe, "Toward a 'Definitional Model1 of Public Perceptions
of Air Pollution, " J. of the Air Pollution Control Association,
Vol. 18, pp. 154-157, 1968.
12. Robe.rt C. Davis, The Public Impact of Science in the Mass Media,
Ann Arbor: Survey Research Center, University of Michigan,
1958.
13. Paul B. Downing, "Solving the Air Pollution Problem: A Social
Scientist's Perspective, " Natural Resources Journal, Vol. II,
pp. 693-713, 1971.
-------
91
14. Jacob J. Feldman, The Dissemination of Health Information, Chicago:
Aldine, 1966.
1 5. Raymond L. Gordon, "Interaction Between Attitude and the Definition
of the Situation in the Expression of Opinion, " American Sociological
Review, Vol. 17, pp. 50-58, 1952.
16. Harold Guliksen, Theory of Mental Tests, New York: Wiley, 1950.
17. J. R. Hardison, "Status of Field Burning and Alternative Methods for
Disease Control in Perennial Grasses, " in R. M. Alexander, ed. ,
Agricultural Field Burning in the Willamette Valley, Corvallis,
Oregon: Air Resources Center, Oregon State University, pp. 18-24,
1969.
18. Paul Horst, Psychological Measurement and Prediction, Belmont,
Calif. : Wadsworth, 1966.
19. Carl I. Hovland, "Reconciling Conflicting Results from Experimental
and Survey Studies of Attitude Change, " The American Psychologist,
Vol. 14, pp. 3-14, 1959.
20. Hans Huth, Nature and the American: Three Centuries of Changing
Attitudes, Berkeley, Calif. : University of California Press,,
19,57.
21. Herbert H. Hyman and Paul B. Sheatsley, "Some Reasons Why Infor-
mation Campaigns Fail, " Public Opinion Quarterly, Vol. 11,
pp. 413-423, 1947.
22. J. Johnston, Econometric Methods, New York: McGraw-Hill, 1963.
23. V. O. Key Jr. , Public Opinion and American Democracy, New York:
A, A. Knopf, 1961.
24. Joseph T. Klapper, The Effects of Mass Communication, Glencoe,
111. : The Free Press, 1961.
25. Jerome M. Levine and Gardner Murphy, "The Learning and Forgetting
of Controversial Material, " J. of Abnormal and Social Psychology,
Vol. 38, pp. 507-517, 1943.
26. Robert G. Mason, Selection of Items for an Air Pollution Knowledge
Test, Corvallis, Oregon: Special Report 324, Agricultural
Experiment Station, Oregon State University, 1971.
27. Peter McHugh, Defining the Situation, New York: Bobbs-Merrill, 1968.
-------
92
28. Jack M. McLeod and James W. Swinehart, Satellites, Science and the
Public: A Report of a National Survey on the Public Impact of Early
Satellite Launchings, Ann Arbor: Survey Research Center,
University of Michigan, 1959.
29. Nahum Z. Medalia, Community Perception of Air Quality: An Opinion
Survey of Clarkston, Washington, Cincinnati: Environmental
Health Service, U.S. Department of Health, Education, and
Welfare, 1965.
30. Willis E. Middlemiss and Robert O. Coppedge, Oregon's Grass and
Legume Seed Industry in Economic Perspective, Corvallis, Oregon:
Special Report 284, Cooperative Extension Service, Oregon State
University, 1970.
31. W. E. K. Middleton, Vision Through the Atmosphere, Toronto:
University of Toronto Press, 1952.
32. Roderick Nash, Wilderness and the American Mind. New Haven:
Yale University Press, 1967.
33. Jum Nunnally, Popular Conceptions of Mental Health, New York:
Henry Holt, 1961.
34. , "Experimental Studies of Communicative Effectiveness, "
in Wilbur Schramm, ed. , Studies of Innovation and Communication
to the Public, Stanford: Institute for Communication Research,
Stanford University, pp. 175-202, 1961.
35. Clarice N. Ol/sin, et. aL , "The Community Editor's Power and the
Reporting of Conflict, " Journalism Quarterly, Vol. 45, pp. 243-
252, 1968.
36. Bernard Ostle, Statistics in Research, Ames, Iowa: The Iowa State
University Press, 1963.
37. Elmer Robinson, "Effects of the Physical Properties of the Atmosphere, "
in A. C. Stern, ed. , Air Pollution, New York: Academic Press,
Vol. I, pp. 349-400, 1968.
38. Merrill E. Samuelson, et.al., "Education, Available Time and Mass
Media Use, " Journalism Quarterly, Vol. 40, pp. 491-496, 1963.
39. Thomas J. Scheff, "Toward a Sociological Model of Consensus, "
American Sociological Review, Vol. 32, pp. 32-46, 1967.
-------
93
40. Wilbur Schramm, "Science News and the Public Mind, " in Wilbur Schramm,
ed. , Studies of Innovation and of Communication to the Public,
Stanford: Institute for Communication Research, Stanford
University, pp. 261-286, 1962.
41. Jean J. Schueneman, "Air Pollution Control Administration, " in A. C.
Stern, ed. , Air Pollution, New York: Academic Press, Vol. Ill,
pp. 719-795, 1968.
42. Muzafer Sherif and Carolyn W. Sherif, An Outline of Social Psychology.
New York: Hapers, 1956.
43. George W. Snedecor, Statistical Methods, Ames, Iowa: The Iowa
State College Press, 4th edition, 1946
44. Shirley A. Star and Helen MacGill Hughes, "Report on an Educational
Campaign: The Cincinnati Plan for the United Nations, " American
J. of Sociology, Vol. 55, pp. 389-400, 1950.
45. Robert A. Stebbins, "A Theory of the Definition of the Situation, " The
Canadian Review of Sociology and Anthropology, Vol. 4, pp. 148-
164, 1967.
46. Arthur C. Stern, "Air Pollution Standards, " in A. C. Stern, ed. ,
Air Pollution, New York: Academic Press, Vol. Ill, pp. 601-720,
1968.
47. The Quality of Outdoor Recreation: As Evidenced by User Satisfaction,
Report to the Outdoor Recreation Resources Review Commission,
Study Report 5, Washington: U. S. Government Printing Office, 1965.
48. The Roper Report, New York: The Roper Organization, Inc. , March
1972.
49. W. I. Thomas, The Child in America, New York: A. A. Knopf, 1928.
50. Phillip Tichenor, Communication and Knowledge of Science in the Adult
Population of the U. S. , unpublished Ph. D. Dissertation, Stanford
University, 1965.
51. P. J. Tichenor, ei_ &L , "Mass Media Flow and Differential Growth in
Knowledge," Public Opinion Quarterly, Vol. 34, pp. 159-170,
1970.
52. P. J. Tichenor, e±aL, "Environment and Public Opinion," The J. of
Environmental Education, Vol. 2, pp. 38-42, 1971.
-------
94
53. Stefan Valvanis, Econometrics, New York: McGraw-Hill, 1959-
54. R. B. Zajonc, "Attitudinal Effects of Mere Exposure, " J. of Personality
and Social Psychology Monograph Supplement, Vol. 9, Part 2, 1968.
-------
APPENDICES
-------
APPENDIX A
Review of Literature
-------
95
APPENDIX A
Review of Literature
Few propositions are as widely accepted as the one -which claims a re-
lationship between the attitudes people have toward some object and the
knowledge they possess about it. Such a relationship is assumed in any
attempt to create, reinforce or convert attitudes by the dissemination of
factual information. It is in this context that a review of literature will
provide much in the way of a foundation to support the theoretical model.
The review will consider first the postulated relationships of variables in
the interdependent system, then the postulated relationships of the predictor
variables.
Effects of the Interdependent System
Attitudes. The role of attitudes and attitude change in predicting
levels or changes in exposure and knowledge has been studied in several
areas. Tichenor [50 ] , in his re-analysis of the effect of a "threat of science"
attitude scales on science knowledge, found that people with lower science
knowledge scores perceived science as more "threatening". Nunnally
[33, 34 J , utilizing laboratory experiments, discovered that it is relatively
easy to convey new information about mental illness. Any seemingly authori-
tative and comprehensible information will be accepted by the public.
Attitudes are more difficult to change, but they can be changed. Public
anxiety over mental illness concepts is one of the facets affecting attitude
change toward this topic. If anxiety-producing features of the message are
not diminished, adverse attitude toward the topic develops and a premature
closure of information that may be relevant follows. Acquiring information,
regardless of its validity, often tends to reduce fear that may be built up
about mental illness. Thus, the mere act of acquiring information leads to
"constructive" attitude change. Conversely, when information is invalidated --
through counterargument by scientific authority, for example -- anxiety about
the topic tends to increase. He concludes that it may be unwise to invalidate
-------
96
existing public misinformation unless a valid and understandable body of
knowledge can be given in return.
Mass Media Exposure. Schramm [40] considers use of the mass media
to be an important predictor of scientific information; after school years,
most of the increment in science knowledge comes from the mass media.
Amount of exposure to the mass media is often used to predict science
knowledge, with or without controlling for predisposition effects, such as
level of education. Brinton and McKown [4] found that newspaper sub-
scribers differed significantly from non-subscribers in knowledge concern-
ing fluoridation of a public water supply, at a time and in an area marked
by a dispute over this issue. They concluded that a high degree of personal
involvement in an issue and a perceived need for information are not necessary
for a person to expose himself to relevant messages in the mass media.
McLeod and Swinehart [28] have found that media exposure was re-
lated to level of science information. Their 1958 (post-Sputnik) study
showed that mass media exposure was closely linked to attribution of a
scientific purpose to satellites, even with the effect of education controlled.
Since the two variables (education and exposure) were independent pre-
dictors, this provided evidence that mass media exposure may be viewed
as a distinct effect and not merely an artifact of education.
Davis [ 12] was able to describe the amount and kind of media used
after controlling for level of science knowledge. He found that highly
informed people were heavy users of the mass media, used more forms
of the print media (newspapers, books, magazines, etc. ) and used them
more intensely than those less informed. Attention to the electronic media
was less clear-cut. Better informed persons were more likely than the
less informed to be in the radio and television audience, but they spent
less total time with these mediums. The most extensive attention to radio
and television was among the less informed. In a comprehensive survey
on the dissemination of health information, Feldman [ 14] has noted that
-------
97
better informed in dividual a were also newspaper and magazine readers, and
constituted a read/ audience for the health educator.
Science knowledge can affect mass media exposure and the question of
whether knowledge precedes or follows exposure -- or both -- is often
unresolved. Studies of selective processes have frequently suggested that
persons with higher interest and prior knowledge are more likely to expose
themselves to sources of information to gain even more knowledge. For
example, Myman and Sheatsley [21] report that persons showing high
interest in the atomic bomb were more than twice as likely as uninterested
persons to report having been exposed to information about an early atomic
energy report. Selective processes also work to "screen out" threatening
messages from the individual's perceived environment. Illustrative of
this process are the findings of Cannell and MacDonald [ 5] . They dis-
covered that readership of articles on health -- particularly those dealing
with the possible relationship between smoking and cancer -- were con-
sistently read by 60 percent of the non-smoking males but by only 32 per-
cent of the smoking males to whom the information presented was more of
a threat. Among women, for whom the threat was considerably less, they
found that the respective readership percentages were 69 and 59.
Knowledge. A widely recognized tenet of social psychology is that
people tend to select and retain information which is consistent with their
prior attitudes. Messages which conflict with established attitudes are often
misperceived, distorted or ignored in order to "fit" with the pre-existing
attitudes. Illustrative of this principle is the Allport and Postman [1] study
on rumor transmission. Persons who were shown a picture of a white man
and a Negro arguing, with the white man holding a knife, later reported that
the Negro was wielding the knife, consistent with their prior attitude toward
Negroes. Levine and Murphy [ 25] presented pro- and anti-Communist
statements to students. Afterwards, students tended to recall only informa-
tion consistent with their prior attitudes: Pro-Communists were more likely
to remember pro-Communist information, and vice-versa. Star and
-------
98
Hughes [44], noting a failure of a massive information campaign in Cincinnati
to increase level of information concerning the United Nations, concluded that
any increase in exposure was determined largely by peoples' previous
orientation. Interest and prior knowledge determined who would attend the
information campaign and those who attended were relatively unaffected by
the campaign itself. Thus, what one knows about a subject often can serve
as a predictor of his attitude about the subject. Tichenor [50] has supported
this in two respects. He found that general science knowledge was related
to three attitudinal items that were combined into an index of "threat of
science". It was noted earlier that those who had lower science knowledge
scores perceived science as more "threatening", i. e. , fear that science may
mean control by a few over many, concern for science-caused rapid changes
and the belief that science tends to break down moral ideas. In addition, he
reported that persons with lower science knowledge scores tend to hold
negative attitudes toward science. This pattern held at all levels of education
and was not associated with level of exposure to science information in the mass
media.
Proposed changes in the environment, however, have formed attitudes
among individuals who perceived that they might affect the outcome of these
changes. In his study of a local controversy concerning alternative methods
of flood control and-water conservation, Baur [ 2] found that attitudes were
formed, in part, through subtle changes in kinds of knowledge about the
issue.
Effect of Predictors
Level of Exposure to Interpersonal Information Sources. Exposure to
interpersonal discussion, mass media use and level of knowledge may com-
plement each other in forming or reinforcing attitudes toward air quality.
Information reaching individuals via the mass media is considered sus-
ceptible to the mediation of other people, as Klapper [24] has noted.
Moreover, Baur [ 2 ], in his study of attitude formation and change in a
-------
99
public controversy, concluded that opinions emerged and were changed pri-
marily through interpersonal communication, although changes were initi-
ated by formal organizations and projected by the mass media. Level of
interpersonal discussion is considered a predictor of attitude.
Experience with Air Pollution. Experience refers to the level of parti-
cipation or living in an area in which a known deterioration in air quality
has occurred. It is considered a basis for both attitude formation toward,
and knowledge about, air quality. Medalia [29] , for example, has reported
that length of residence in Clarkston, Washington, (an area subjected to
continuous pollution from a kraft pulp mill) was strongly related to concern
with air pollution. Approximately two-thirds of the persons interviewed
who had lived in Clarkston in 1950 or before were highly disturbed by air
pollution as a problem, compared to a little over a third of those who had
moved to the city in 1951 or later.
Level of Social Influence.. People differ in their level of social
influence, and a well-established proposition concerning the behavior
of influentials is their level of exposure to the mass media, particularly
the printed word. Those high in influence tend to expose themselves to
pertinent forms of the mass media compared to those less influential,
according to Klapper [24] . Level of social influence is considered
a predictor for level of mass media exposure.
Level of Education. Satisfactory completion of successive years of
education provides a person with greater knowledge and skills. The knowledge
acquired, of whatever quality, can only expand a person's total world of
actuality and reality. Each increment of schooling adds to his competence
through the acquisition of knowledge and Incidental skills. Further, advance-
ment into college or university training Is taken to reflect the potential of
better performance, compared to others who are screened from higher
education. The reading skills learned have Important consequences In later
life for Increased exposure to the mass media, particularly the printed word.
This fact has been well documented by many communication studies (e._g. ,
-------
100
Samuelson, £t aL [38]), In turn, it is not at all surprising to find, as
Schramm [40] has reported and as Tichenor [50 ] has confirmed, that an
individual's level of education is the chief predictor of his level of exposure
to science information in the mass media and his amount of science knowledge.
Level of Science Education. Level of formal education may not be the
whole story. It is conceivable that high school or college graduates -- with
no formal science education -- may not acquire the vocabulary and content
necessary to evaluate and comprehend science to the same extent as those
who have completed some formal science courses. In his secondary
analysis, Tichenor [50 ] reported that the amount of science education in
high school or college also may be an important predictor of science know-
ledge. These attributes were strikingly low for high school and college-
educated individuals with no formal science course work,
Level of the Use of the Environment. People differ in the percent of
time they spend outdoors and individuals who are outdoors much of their
daylight hours may have a greater opportunity to differentiate and evaluate
the air quality of where they live, compared to those who are outdoors little,
if any. This may be particularly true in the realm of outdoor recreation use.
Air quality plays a role in the preference people have for outdoor recreation,
according to a report of the Outdoor Recreation Resources Review Commission
[47] . Deterioration in air quality was one of the reasons cited for recreation-
user dissatisfaction. Thus, it appears that the quality of the environment is
an important aspect for sustained outdoor recreation use, and the level of
outdoor use may be a predictor of knowledge toward air quality.
-------
APPENDIX B
Interview Schedules A and B
-------
APPENDIX B. Interview Schedule A
101
Budget Bureau Number 85-S70010
Approval Expires March 31. 1971
Schedule A
OREGON STATE UNIVERSITY
Final
Hello, I'm working on a survey for Oregon State University and I would like to ask you a
lea interesting questions, If* you don't mind.
1 -
2 _
3 -
4 -
4
3
2
1
3
2
1
Years
X DK, NA
Own
Rent
Other
DK, NA
Years
Yes
No
DK, NA
First, how many years have you, yourself, lived in the
Sale^n (Eugene) Metropolitan Area?
May I ask whether you own or rent the place in which you are
now living?
Approximately how long have you lived in this particular
dwelling unit?
Some people plan to move within six months to a year, while
others plan to stay put in the same home or apartment. How
about you do you have any plans to move to another place
within, say, the next six months?
NOTE TO INTERVIEWER: If respondent answers "Yes" to Q 4 or_ Rents and has lived in current
DU less than one year, terminate the interview.
(ASK OF EVERYONE)
5 -
6 -
7 -
8 -
2
1
3
2
1
4
3
2
1
0
4
3
2
1
0
Yes - Asked
No or DK
Convince
Active part
Listen or DK
A lot
Quite a bit
Little
None
DK, NA
A lot
Quite a bit
A little
None
DK, NA
Looking back over the past few months, can you remember any
time someone outside your immediate family has asked your
opinion about state or community affairs?
When a discussion is on about community affairs, do you
usually listen, take an active part, or try to convince
others especially?
Some people earn all or part of their living by working out-
doors, while others' jobs keep them inside. How about your
family? About how much time, if any, did the chief bread-
winner spend working outdoors for pay last year a lot,
quite a bit, a little, or none at all?
Some people spend quite a bit of time out-of-doors while
others have other pastimes. Not counting work for pay, how
much time, if any, did you, yourself, spend out-of-doors last
year none at all, a little, quite a bit, or a lot?
-------
102
Now, here are some specific out-of-door activities. After I read each one, will you
please tell me how much time, if any, you spent participating in that activity during
the past twelve months a lot, quite a bit, a little, or none at all?
9 -
10 -
11 -
12 -
13 -
14 -
15 -
16 -
17 -
18 -
19 -
20 -
21 -
22 -
23 -
24 -
25 -
Lot
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
Quite
bit
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
Little
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
None
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
DK, NA
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Activity
Bicycling
Attending outdoor cultural events
Gardening
Picnicking
Camping
Pleasure driving, sightseeing
Boating or water skiing
Horseback riding
Hunting
Attending outdoor sporting events
Playing golf
Ocean beach activities
Walking for pleasure, hiking,
study, rockhounding, etc.
Swimming
Fishing
nature
Playing outdoor games and sports
Other
(Specify)
26 - Suppose one of your friends came to you and asked you how he could find out what is
known about air pollution. Where would you advise him to go for that information?
(PROBE for both someone to talk to and something to read)
Anything else?
-------
103
27 - Do you feel you personally know more or less about air pollution than most people?
(PROBE!)
Anything else?
28 - Where do you feel that you, yourself, have gotten the most valuable information
about air pollution? (PROBE!)
Anything else?
29 -
30 -
31 -
32 -
33 -
1
2
3
4
5
1
2
3
4
5
Oregon Statesman (R-G)
Capital Journal
Other
None
DK, NA
Press (Reminder)
Other
Other
None
DK, NA
Number
Number
What daily newspaper, or newspapers, if
into your home yesterday?
What weekly newspaper, or newspapers, if
into your home last week?
Not counting car radios , how many radios
do you have in your home?
any , came
any , came
, if any,
And, how many TV sets, if any, do you have in your
home?
Which magazines or publications, if any,
your home this past month? Any others?
came into
0 None
-------
104
34 - Listed on this card are several sources that many people use to obtain information.
From which one of these sources do you believe you get most of your information
about air pollution? (HAND RESPONDENT CARD A) Which is second most in importance?
Which is j-east in importance?
Most Next Least
1 Newspaper
2 Magazine
3 Radio
4 Television
5 Other
6 DK, NA
1
2
3
4
5
6
Newspaper
Magazine
Radio
Television
Other
DK, NA
1
2
3
4
5
6
Newspaper
Magazine
Radio
Television
Other
DK, NA
I have a list of statements here concerning different sources of information about
air pollution. As I read each one, will you please tell me what you know or how
you feel about them?
Quite, Little, Nothing,
)uite Not too Never DK
35 - 4 3 2 1 0 How much have you talked to someone at
Willamette University (Univ. of Oregon) or
Oregon State about air pollution not at
all, a little, quite a bit, or a lot?
36 - 4 3 2 1 0 How much have you viewed on TV about air
pollution a lot, quite a bit, a little,
or nothing?
37 - 4 3 2 1 0 How often have you talked to someone at
the state or local Air Pollution Control
Authority about air pollution never,
not too often, quite often, or very often?
38 - 4 3 2 1 0 How much have you read in the newspaper
about air pollution nothing, a little,
quite a bit, or a lot?
39 - 4 3 2 1 0 How often have you discussed air pollution
with your neighbors very often, quite
often, not too often, or never?
40 - 4 3 2 1 0 How often have you listened to something
about air pollution over the radio
never, a little, quite a bit, or a lot?
41 - 4 3 2 1 0 How much have you talked to your friends
about air pollution a lot, quite a bit,
a little, or never?
42 - 4 3 2 1 0 How much have you read in a magazine about
air pollution nothing, a little, quite
a bit, or a lot?
-------
(ASK OF EVERYONE)
43 - 2 Yes (Where?
3 Yes (Where? .
4 Yes (Where?
1 No (Skip to Q 46)
0 DK, NA (Skip to Q 46)
105
Within the past five years, have you lived in
any state or states other than Oregon where
you felt there r-ras an air oollution problem
while you were living there?
(IF YES IN Q 43, ASK FOR EACH AREA)
44 - How serious did you consider their air pollution problem to be not too serious,
somewhat serious, or very serious?
Area(s)
4
3
2
1
Very serious
Somewhat
Not too
DK, NA
4
3
2
1
Very serious
Somewhat
Not too
DK, NA
4
3
2
1
Very serious
Somewhat
Not too
DK, NA
45 - What type or types of air pollution was there in this (these) area(s) where you
used to live just describe it in your own words? (PROBE!)
(INT: Write in area(s) with response opposite)
Area; Response
Area:
Response
Area:
Response
-------
106
(ASK OF EVERYONE)
46 -
2 Yes (Where?
3 Yes (Where?
4 Yes (Where?
1 No (Skip to Q 49)
0 DK, NA (Skip to Q 49)
) Within the past five years, have you visited
) any state or states other than Oregon where
) you felt there was an air pollutiori problem
while you were visitine there?
(IF YES IN Q 46, ASK FOR EACH AREA)
47 - How serious did you consider their air pollution problem to be very serious,
somewhat serious, or not too serious?
Area(s)
4
3
2
1
Very serious
Somewhat
Not too
DK, NA
4
3
2
1
Very serious
Somewhat
Not too
DK, NA
4
3
2
1
Very serious
Somewhat
Not too
DK, NA
48 - What type or types of air pollution was there in this (these) area(s) where you
visited just describe it in your own words? (PROBE!)
(INT: Write in area(s) with response opposite)
Area; Response
Area:
Response
Area:
Response
-------
107
(ASK OF EVERYONE)
49 - 2 Yes (Where?
3 Yes (Where?
4 Yes (Where?
1 No (Skip to Q 52)
0 DK, NA (Skip to Q 52)
) Within the past five years, have you lived in
) any community or communities in Oregon other
~) than the one you live in now where you felt
there was an air pollution problem while you
were living there?
(IF YES IN Q 49, ASK FOR EACH AREA)
50 - How serious did you consider their air pollution problem to be very serious,
somewhat serious, or not too serious?
Area(s)
4
3
2
1
Very serious
Somewhat
Not too
DK, NA
4
3
2
1
Very serious
Somewhat
Not too
DK, NA
4
3
2
1
Very serious
Somewhat
Not too
DK, NA
51 - What type or types of air pollution was there in this (these) area(s) where you
lived just describe it in your own words? (PROBE!)
(INT: Write in area(s) with response opposite)
Area: Response
Area:
Response
Area:
Response
-------
108
(ASK OF EVERYONE)
52 - 2 Yes (Where?
3 Yes (Where?
4 Yes (Where?
1 No (Skip to Q 55)
0 DK, NA (Skip to Q 55)
Within the past five years, have you visited
any part or parts of Oregon where you felt
there was an air pollution problem while you
were visiting there?
(IF YES IN Q 52, ASK FOR EACH AREA)
53 - How serious did you consider their air pollution problem to be not too serious,
somewhat serious, or very serious?
Area(s)
4
3
2
1
Very serious
Somewhat
Not too
DK, NA
4
3
3
1
Very serious
Somewhat
Not too
DK, NA
4
3
2
1
Very serious
Somewhat
Not too
DK, NA
54 - What type or types of air pollution was there in this (these) area(s) where you
visited just describe it in your own words? (PROBE!)
(INT: Write in area(s) with response opposite)
Area: Response
Area:
Response
Area:
Response
-------
109
(ASK OF EVERYONE)
55 - 2 Yes
1 No (Skip to Q 58)
0 DK, NA (Skip to Q 58)
Now, thinking of the area within a mile of
where you now live in Oregon, do you feel
that there is an air pollution problem in
this area?
(IF YES IN Q 55)
56 - 4 Very serious
3 Somewhat
2 Not too
1 DK, NA
How serious do you consider this air pollu-
tion problem within a mile of where you now
live very serious, somewhat serious, or
not too serious?
57 - What type, or types, of air pollution problems do you have in this area where you
now live just describe them in your own words? (PROBE!)
(ASK OF EVERYONE)
Here are some statements that have been made recently about air pollution. As I
read each one, would you please tell me how strongly you agree or disagree with
it? (HAND RESPONDENT LIST OF Q's 58-77)
58 -
59 -
60 -
61 _
60
cu a
0) O
cu
cu
J-l
bO
g
o P<
53 O
0)
0)
t-l
60
rt
CO
co s*,
CU rH
)-i M
00 C
Cfl O
CO M
H 4J
n w
Statement
We don't have to worry about air pollution
here. The problem will solve itself.
People must be ready to pay increased
taxes, utility rates, and higher prices
on consumer goods if we are to have
clean air again.
As a good place to live, the Willamette
Valley is deteriorating from air pollution
year by year.
More public support is needed for air
pollution control programs.
-------
110
0) >
C a)
-------
111
75 -
76 -
77 -
60
<0 fi
0) O
QJ
01
60
No
opinion
Disagree
Disagree
strongly
0
0
0
Statement
Air pollution produced by Kraft paper
mills stinks.
I don't feel that air pollution has any
effect on the economy of this community.
While sources of air pollution may be
environmental villains, they are also
social and economic necessities.
Most industries won't do anything about con-
trolling their air pollution unless they are
forced to.
78 - 1 Farm group
2 Lodge or fraternal
order
3 Church social or
service group
4 Labor union
5 Civic or service club
6 Political group
7 PTA
8 Occupational or
businessmen's group
9 Social club
0 Conservation, environmental
or clean-air group
+ Other
(Specify)
- None (Skip to Q 85)
Are you, yourself, presently a member of any
of the following kinds of organizations?
(HAND RESPONDENT CARD C)
79 - What are the names
of the groups to
which you, yourself,
belong? (NOTE: List
specific names of
groups circled in
Q 78)
80 -
About how many times a
year does this group meet?
(Exc. Sun. at Church)
81 -
About how many times
did you attend last
year?
-------
112
82 - Have you held an office
in any of the groups to
which you belong?
2 Yes (What group?)
1 No, DK, NA
83 - How long ago did you hold office in
the group?
84 -
84a -
1 Yes (Ask Q 84a)
2 No (Skip to Q 85)
3 DK (Skip to Q 85)
Do you feel that membership in any of the groups to
which you belong helps you obtain valuable information
about air pollution?
(IF YES) Which group or groups do you feel provide
you with valuable information about air
pollution?
84b -
0 DK
(IF MORE THAN ONE GROUP MENTIONED IN Q 84a) Which one
group do you feel you received the most valuable air
pollution information from?
(ASK OF EVERYONE)
I have here another list of statements that have been made lately about air pollu-
tion. As I read each one, would you please tell me quickly if you think they are
true or false? (HAND RESPONDENT LIST OF Q's 85-]04)
True False NA, DK Statement
85-1 0 - Persons who suffer from respiratory ailments, such
as asthma, emphysema and bronchitis, are particu-
larly affected by air pollution.
86-0 1 ~ Jet airplane engines might destroy enough ozone
to permit a harmful increase in ultraviolet light
reaching the Earth.
87-0 1 - A physician can measure the amount of air pollution
by examining heart and lung functions of some of
his patients.
88-1 0 - Areas of Oregon not included in the regional air
pollution control authorities will be under the
jurisdiction of the Oregon Environmental Quality
Commission.
89-0 1 - Education and quiet persuasion have usually proved
to be better tools than rigid abatement proceedings
in waging the war against air pollution.
-------
113
True False NA, DK Statement
90-0 1 - Forest slash burning does not produce the same
kind of smoke that agricultural field burning
produces.
91-1 0 - Sunlight causes hydrocarbons to combine with
oxides of nitrogen to produce photochemical smog.
92-1 0 - Air pollution affects visibility.
93-1 0 ~ Carbon monoxide is one type of an air pollutant
emitted by automobiles.
94-1 0 - The U. S. has both the technology and the wealth
to drastically reduce air pollution.
95-0 1 ~ About half the states have air pollution control
laws.
96-1 0 - Carbon monoxide reduces the oxygen-carrying capa-
city of the blood.
97-0 1 - Air pollution in the Willamette Valley of Oregon
is as bad as the air pollution in Los Angeles.
98-1 0 - Foreign cars sold in the U. S. will have to carry
new anti-pollution equipment required by Federal
standards.
99 - 0 1 - While Kraft mill air pollutants may have an
unpleasant odor, they are not harmful to human
health.
100-0 1 - The atmosphere cannot be cleansed once it is
polluted.
101-0 1 - There is no way of knowing the amount of pollutants
emitted into the air from automobile exhausts.
102-1 0 - Dust storms and sand storms, forest fires, earth-
quakes, landslides, and windborne pollens are
ways nature pollutes the air.
103-Q 1 - Smoke in the air does not aggravate asthma or
upper respiratory trouble.
104 - 1 0 - Fine particles of lead are emitted as air pollu-
tion by automobiles.
105- 2 Yes - asked Looking back over the past few months, can you
1 No or DK remember any time someone outside your immediate
family has asked your opinion about air pollution?
-------
114
106 -
107 -
108 -
3
2
1
2
1
4
3
2
1
Convince
Active part
Listen or DK
Yes
No or DK
College - complete
College - partial
High school
Grade or no schooling
When a discussion is on about air pollution, do
you usually listen, take an active part, or try
to convince others especially?
Do you happen to know any political office holder
in the state personally?
Would you mind telling me the last grade you
completed in school?
109 - (IF RESPONDENT WENT TO HIGH SCHOOL OR BEYOND) Would you tell me what science or
mathematics courses, if any, you took in high school (and college). As I read each
one will you tell me whether or not you took them?
Didn't
H.S. College take Mathematics
1 20 Algebra
_1 2 0 Geometry
1 20 Trigonometry
1 20 Calculus
1 20 Statistics
1 20 Other
(Specify)
Science
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
0
0
0
0
0
0
0
0
0
0
General science
Biology
Chemistry
Physics
Geology
Botany
Astronomy
Zoology
Microbiology
Other
(Specify)
Behavioral Science
1 20 Economics
J. 2 0 Psychology
1 20 Sociology
1 2 0 Anthropology
1 20 Political science
1 20 Other
(Specify)
-------
115
110-
(Type of industry)
(Specific job)
What type of work does the chief
breadwinner in the family do?
111-
1
2
3
4
5
20-29
30-39
40-49
50-59
60 & above
May I ask your approximate age?
BY OBSERVATION
112 -
1
2
Male
Female
113 - 1 Less than 30 minutes
2 30 minutes - 1 hour
3 1 hour - 1^ hours
4 More than 1^ hours
Length of interview
114 - 3 Positive
2 Neutral
1 Negative
Rating on cooperation
X I hereby certify this interview was actually taken at the following address and repre-
sents a true and accurate account of the interview.
(Address)
(City)
(Interviewer's Signature)
Phone Number (For Name
verification only)
FOR OFFICE USE ONLY: Interview verified by
Date
-------
116
APPENDIX B. Interview Schedule B Budget Bureau Number 85-S70010
Approval Expires March 31, 1971
, . , n OREGON STATE UNIVERSITY
Schedule B Final b
Hello, I'm , and I'm working on a research project for Oregon State University.
May I ask if you are . Thank you. As you may remember, you were kind
enough to grant us an interview a short while ago on some interesting topics, one of
Which was air pollution. We'd like to ask you to grant us just a few more minutes of your
time to get the benefit of some more of your reactions. As you can see, (show quex) this
is much shorter than the first one and will only take a few minutes of your time. Your
interview is so important to us since you are a vital part of our cross-section, and you
will be doing the University a great favor.
The results are tabulated for the area as a whole, not for any one person. Any-
thing you say will be held in the strictest confidence.
(ASK OF EVERYONE)
Here are some statements that have been made recently about air pollution. As I
read each one, would you please tell me how strongly you agree or disagree with
it? (HAND RESPONDENT LIST OF Q's 1 - 20)
x, 0) 0) >s
iH C
-------
117
>. >
rH C QJ 0) i-l
Of O H M 60 <
<0 ti -i -rt 03 MM >
-------
118
I have a list of statements here concerning different sources of information about
air pollution. As I read each one, will you please tell me what you know or how
you feel about them?
21 - 4
22 - 4
23 - 4
24 - 4
Little,
Not too
Nothing,
Never
DK
0
How much have you read in a magazine about
air pollution a lot, quite a bit, a
little, or nothing?
How much have you viewed on TV about air
pollution nothing, a little, quite a
bit, or a lot?
How often have you discussed air pollution
with your neighbors never, not too
often, quite often, or very often?
How often have you talked to someone at
the state or local Air Pollution Control
Authority about air pollution very
often, quite often, not too often, or
never?
25 -
26 -
27 -
28 -
How often have you listened to something
about air pollution over the radio
a lot, quite a bit, a little, or never?
How much have you talked to your friends
about air pollution never, a little,
quite a bit, or a lot?
How much have you read in the newspaper
about air pollution a lot, quite a bit,
a little, or nothing?
How much have you talked to someone at
Willamette University (Univ. of Oregon) or
Oregon State about air pollution a lot,
quite a bit, a little, or not at all?
29 - Listed on this card are several sources that many people use to obtain information.
From which one of these sources do you believe you get most of your information
about air pollution? (HAND RESPONDENT CARD A) Which is
Which is least in importance?
1
2
3
4
5
6
Most
Newspaper
Magazine
Radio
Television
Other
DK, NA
1
2
3
4
5
6
Next
Newspaper
Magazine
Radio
Television
Other
DK, NA
second most
1
2
3
4
5
6
in importance?
Least
Newspaper
Magazine
Radio
Television
Other
DK, NA
-------
119
(ASK OF EVERYONE)
I have here another list of statements that have been made lately about air pollu-
tion. As I read each one, would you please tell me quickly if you think they are
true or false? (HAND RESPONDENT LIST OF Q's 30 - 49)
True False NA, DK Statement
30-1 0 - Agricultural field burning during the summer is a
source of air pollution in the upper Willamette
Valley of Oregon.
31-0 1 ~ Scientific evidence is inconclusive that sulphur
dioxide is a hazard to human health.
32-1 0 ~ Air pollution emissions from industrial processes
have damaged livestock, crops, fruit trees, and
flowers.
33-1 0 - Air pollutants fade fabrics.
34-0 1 - The best known air pollution control devices still
don't work effectively.
35-1 0 - Some air pollution control devices earn profits for
industry from the pollutants recovered from the air.
36-0 1 - California is the only state trying to reduce auto-
mobile pollution through the installation of tail-
pipe afterburners.
37-0 1 ~ The great majority of Americans has never seen smog.
38-1 0 ~ Many companies have waited until local authorities
have forced them to control their air pollution
emissions.
39-0 1 ~ Sulfur oxides are known to cause cancer, though it
takes higher concentrations than are usually found
in the air.
40-1 0 ~ Air pollution is irritating mentally as well as
physically.
41-0 1 ~ There is no scientific evidence that air pollution
leads to emphysema, chronic bronchitis, or lung
cancer.
42-0 1 ~ There is as yet no definite medical evidence about
the relationship between air pollution and human
health.
43-1 0 ~ There is an economic issue involved with air pollution
control.
-------
120
True False NA. DK Statement
44 - 0 J. - Rain has no effect on temperature inversions.
45-1 0 - Troublesome gaseous contaminants may pass through
air conditioners unhindered.
46 -
47 -
48 -
49 -
50 -
01-
Ol-
lO-
10-
3 Better
2 Worse
1 No diff. (Skip to
Q 53)
0 DK, NA "
Most devices for reducing auto exhaust fumes have
been designed to fit on a car's tailpipe.
Sulphur dioxide cannot be controlled through devices
on stacks.
Polluted air and dust can be carried for long dis-
tances by the wind before reaching the ground.
Pollution exists in two basic forms particles and
gases.
Within the last three weeks, would you say that the
air pollution in Salem (Eugene) is better or worse
than when we talked to you earlier this year in May?
51 - 3 Lot
2 Somewhat
1 Little
0 DK, NA (Skip to
Q 53)
How much (better) (worse) would you say the air
pollution is in Salem (Eugene)a lot, somewhat,
or a little?
52 - How or in what way(s) would you say that the air pollution in Salem (Eugene)
is(better) (worse) than when we talked to you last May? (PROBE!)
Anything else?
-------
121
.BY OBSERVATION
53 -
1 Male
2 Female
54 - 1 Less than 30 minutes
2 30 minutes - 1 hour
3 1 hour - 1% hours
4 More than 1^ hours
Length of interview
55 -
3 Positive
2 Neutral
1 Negative
Rating on cooperation
X I hereby certify this interview was actually taken at the following address and
represents a true and accurate account of the interview.
(Address)
(City)
(Interviewer's Signature)
Phone Number (For verification only)
FOR OFFICE USE ONLY: Interview verified by
Date
-------
APPENDIX C
Frequency Curves of Clean Air Themes
(Figures 9-22)
-------
15 -
14 .-
13 ..
12
11
10
9
8 ..
7
6 ..
5
4 -.
Theme 1. Air pollution is a health hazard.
Theme 10. Air pollution may be a problem, and industry
and economics are not as important as clean air.
Theme 1
ii if ii mi i ii ii ii n ii
i ii ii ii ii i mm ii mum
I960 1961 1962 1963 1964 1965 1966 1967 1968 1969
Figure 9. Frequency distribution of "for clean air" themes 1 and 10, Eugene, 1960-1970.
1970
ho
-------
o
I
cr
-------
o
c
s
tr
15 j
14
13
12 , .
11
10
9
8 ..
7 ..
6 ..
5
2 ..
1
0
Theme 2. Air pollution is a menace to the public welfare.
Theme 5. Air pollution is a problem caused by industry, and
it is industry's responsibility to clean up the air
im 1111111 fiiiTTiimH
MM inn n\n 111 i7i iVij IN i n 1111 i₯|irn-HH
1
I960 1961 1962 1963 1964 1965 1966 1967 1968 1969
Figure 11. Frequency distribution of "for clean air" themes 2 and 5, Eugene, 1960-1970.
1970
-------
15 T
14
13
12
11 , .
10 .
9
Theme 2. Air pollution is a menace to the public welfare.
Theme 5. Air pollution is a problem caused by industry, and
it is industry's responsibility to clean up the air.
mil mi ii mi ii ii mi i im HI
l ii n iiini it MII >nn»n n
I960
1961 1962 1963 1964 1965 1966 1967 1968 1969
Figure 12. Frequency distribution of "for clean air" themes 2 and 5, Salem, 1960-1970.
1970
-------
O
01
3
cr
01
15 j
14 . .
13 . ,
12 ..
11
10 ' '
9
Theme 3. Air pollution is becoming a serious problem and
public opinion must pressure government into
doing something to control it.
Theme 8. Air pollution control agencies need more power if
they are to control the problem of air pollution.
1960 1961 1962 - 1963 1964 1965 1966 1967 1968 1969
Figure 13. Frequency distribution of "for clean air" themes 3 and 8, Eugene, 1960-1970.
1970
-------
15 -
14 .
13
- 12 ..
11
10
9 , .
r
OJ
cr1
01
7 ' '
6
5 ..
Theme 3. Air pollution is becoming a serious problem and
public opinion must pressure government into
doing something to control it.
Theme 8. Air pollution control agencies need more power if
they are to control the problem of air pollution.
1960
1961 1962 1963 1964 1965 1966 1967 1968 1969 1970
Figure 14. Frequency distribution of "for clean air" themes 3 and 8, Salem, 1960-1970.
-------
o
0)
cr
En
15
14-
13' '
12
11
10 "
Theme 4. Before the problem of air pollution becomes too
serious, we can find the resources and technology
to eliminate it by working together now.
Theme 6. Industry is willing to cooperate in cleaning up
its air pollution.
1960
1961
1962
1963
1964
II II IMj M II If I Mil |HI MfMMI jl II II M lfH( II II II II IH| II II
1965
1966
1967
1968
1969
1970
Figure 15. Frequency distribution of "for clean air" themes 4 and 6, Eugene, 1960-1970.
00
-------
Theme 4. Before the problem of air pollution becomes too
serious, we can find the resources and technology
to eliminate it by working together now.
Theme 6. Industry is willing to cooperate in cleaning up
its air pollution.
1 '
n ini iYi[ IYI mi n ii 11 n im n n p n i
mm
I960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
Figure 16. Frequency distribution of "for clean air" themes 4 and 6, Salem, 1960-1970.
-------
15 ,
14
13
12 "
11
10 "
9 .-
51 8 +
QJ
O" -74.
o) 7 T
M
Theme 9. If air pollution control agencies are to be effective,
they must he organized on a regional hasis.
Theme 14. The available means of controlling industrial air
pollution is effective or technologically practical.
1960
1961 ' 1962 1963 ' 1964 1965 1966 1967 ' 1968 1969 1970
Figure 17. Frequency distribution of "for clean air" themes 9 and 14, Eugene, 1960-1970.
-------
15 ,
14
13
12
11 ' '
10
B- s
d
Hi
3
cr 7
QJ '
3 '
Theme 9. If air pollution control agencies are to be effective,
they must be organized on a regional basis.
Theme 14. The available means of controlling industrial air
pollution is effective or technologically practical.
[ i m i-m 1111 m 111 n m 111 n 1111 /TTf i
m mm mm 111n ii n mi 11 IHI iiMINI
I960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
Figure 18. Frequency distribution of "for clean air" themes 9 and 14, Salem, 1960-1970.
-------
15"
14. ,
13'
12 , .
11
10
9 "
8
7
6 , .
5 ..
4
3 . .
Theme 11. Industry is a major contributor to the air
pollution problem.
Theme 13. Air pollution has always existed and it may
be a problem, and something must be done to
eliminate it.
Theme 13
IIH ii fin it iiit ii ii ii ii ii m mi ii ii mi
I960
1961
1962
1963 1964 1965 1966 1967 196* 1969
Figure 19. Frequency distribution of "for clean air" themes 11 and 13, Eugene, 1960-1970.
1970
u>
NJ
-------
15
14
13
12
11
10
Theme 11. Industry is a major contributor to the air pollution
problem.
Theme 13. Air pollution has always existed and it may be a
problem, and something must be done to eliminate it.
o
c
-------
o
I
cr
-------
17 "
16
15 "
14 "
13 "
12
11
10
9 "
8
7 "
6 "
5 '
4
3
Theme 6. Industry is not willing to cooperate in
cleaning up its air pollution.
Theme 8. Air pollution control agencies already have
enough power to control air pollution.
Theme 10. Air pollution may be a problem, but industry
and economics are more important than clean air.
Theme 14. The available means of controlling industrial
air pollution is ineffective or technologically
impractical.
I960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
Figure 22. "Frequency distribution of ""against:' "clean air themes, Salem, 1960-1970.
LO
01
-------
APPENDIX D
Verbatim Responses of Individuals
Holding Accurate or Inaccurate
Visibility Perceptions
-------
136
APPENDIX D
Verbatim Responses of Individuals Holding Accurate
or Inaccurate Visibility Perceptions
How or in what way(s) would you say that the air pollution in Salem (Eugene)
is (better) (worse) than when we talked to you last May (September)?
(Accurate responses)
I think field burning is the greatest cause (of air pollution). It's dusty and
vegetation is dying and it's the asthmatic ones that suffer terribly.
(F, Over 60, Widow of Auto Sales Owner, Salem, Sept. )
The grass burning makes your eyes burn. A couple of times it was so thick
it was like fog. One evening about 6 o'clock we were driving down the free-
way and everyone had to have their lights on. Sometimes it chokes you up
and makes you cough. Trucks and buses really put the smoke out, too.
(F, 40-49, Wife of Dispatcher for Can Manufacturer, Salem, Sept. )
I notice that the field burning causes real bad pollution. I don't know how
harmful the field burning smoke really is compared to factories. I don't
think it is as harmful but it certainly affects the visibility and you can smell
it in the air. (F, 20-29, Wife of Correctional Officer, Salem, Sept. )
Smoke--field burning. Whenever the south winds blow you get smoke and
gases from the paper mill and it irritates your nose. (M, 50-59, Power
Saw Operator, Salem, Sept. )
It(air pollution) is due to field burning. It hasn't affected me personally
but it obstructs my view. (M, 30-39, Minister, Salem, Sept. )
The smog and smoke due to field burning seems to be down over town more.
Visibility is limited and the smell is terrible. (F, 40-49, Wife of Office
Machine Technician, Salem, Sept. )
It's (air pollution) from the field burning. With these new burning laws,
we get the smoke from all directions. (M, 50-59, Floor Coverer,
Salem, Sept. )
It seems like the smoke is even coming from the East now. I lived in
Salem all my life and I remember it as clean without all these odors, etc.
I think the taxpayers pay enough. The people causing the pollution should
clean it up and I know they can. I don't really know why it's worse but the
mill out toward Albany is the worst and the one downtown blackens my house
so there is a lot of pollution to do that. Seems like field burning is better
but sometimes that blows in too. (F, Over 60, Teacher, Salem, Sept. )
-------
137
How or in what way(s) would you say that the air pollution in Salem (.Eugene)
is (better) (worse) than when we talked to you last May (September)?
(Accurate responses continued)
(You are not) able to see as far. Field burning is causing it in our area,
also automobiles, I guess. (F, 40-49, Wife of Wholesale Grocery Depart-
ment Manager, Salem, Sept. )
All I can do is look out of the window and I can't see the things I used to be
able to. I know it is worse. What's causing it I don't know unless it's
industries and automobiles. (F, 30-39, Wife of Sales Personnel
Administrator, Salem, Sept. )
For one thing the odor and the visibility are worse. These are the main
things. You can't even see the Coburg Hills anymore, and you've got a raw
throat. (F, 40-49, Wife of Construction Equipment Parts Manager, Eugene,
Sept. )
Field burning. It's completely overcast on a beautiful day, and it can affect
your health. It (air pollution) makes the day miserable when it should be
nice. There is really not a hell of a lot you can say about it, other than
it is miserable. (M, 20-29, Son of Television Electric Firm Owner,
Eugene, Sept. )
There is a little more smoke now that we didn't have in May. The smoke
bothers me. It congests my head. Smoke is caused by field burning from
up North. I notice it on my clothes. It looks like little stringy black
particles. It is especially bad on the days I hang out the laundry. (F,
Over 60, Wife of Retired Shoe Salesman, Eugene, Sept. )
There are lots of particles because of field burning down the valley. It is
settling where we have our recreation and in our pool and patio. Also the
smoke has been far worse (than in May). They've even had to stop burning
to look down at the forests to see if there are any forest fires. The field
burning is contaminating the forests in the Cascades -- just filthy.
(M, 50- 59, Self employed, Heating and Air Conditioning, Eugene, Sept. )
There is more soot around (now). Visibility is affected and sometimes
the air doesn't smell very good. (F, Over 60, Business College Teacher,
Eugene, Sept. )
We get smoke up here on the field burning. I have sinus trouble at that
time. (F, 30-39, Wife of Machine Shop Foreman, Eugene, Sept.)
(It is worse from) the field burning. The wind changes and we get the
smoke that was supposed to blow somewhere else. (F, 30-39, Wife of
Salesman for Publishing Firm, Eugene, Sept. )
-------
138
How or in what way(s) would you say that the air pollution in Salem (Eugene)
is (better) (worse) than when we talked to you last May (September)?
(Accurate responses continued)
The haze is over Eugene constantly, plus the visibility (is worse), and the
smell and my allergy is acting up and even the shots don't help, so I know
it's worse. (F, 40-49, Bookkeeper for Industrial Supply Co. , Eugene, Sept.)
It is overcast all of the time. You can't hang clothes out to dry because
little particles get all over them. (F, 30-39, Wife of Labeling Machine
Crew Operator for Cannery, Eugene, Sept. )
Some of the mills have started up again and there is an awful lot of field
burning. During the summer months the extra traffic from vacationers
makes it worse. (F, 40-49, Wife of Carpenter-Electrician, Eugene, Sept. )
There is smoke in the afternoons, the sun is hazy and you can't see the
hills. The outside of the house is filthy. (F, 40-49, Wife of Freight
Service Truck Driver, Eugene, Sept. )
It's been smokier than last month because of the field burning. I don't
suppose the winds have been just right for burning. (F, 40-49, Wife of
Insurance Agent, Eugene, Sept. )
(It's worse) because of the field burning. Cne thousand percent better
than last year but a lot worse than last May. (F, Over 60, Widow of
Construction Engineer, Eugene, Sept. )
You don't have the field burning now. The rain has a tendency to filter
the air. (M, 40-49, Insurance Field Representative, Salem, Jan. )
In September they were still doing some field burning. This is a big
cause of air pollution. The moisture in the air makes the air clearer.
Weather conditions are a big factor in air pollution. (M, Over 60,
Machinery Maintenance Worker, Salem, Jan. )
(There is) no burning in the fields now. I think people are getting more
conscious of pollution and becoming more careful about trash burning.
(F, 50-59, Wife of Body and Fender Shop Owner, Salem, Jan. )
It's raining now, besides the farmers aren't burning and the pollen count
is sure down. It's (air pollution) always better in the winter. (F, 40-49,
Wife of Civil Engineering Professor, Salem, Jan. )
It seems like when we have rain it (air pollution) always clears up and
last September they were burning and that made it bad then, poor farmers.
(F, 40-49, Wife of School Custodian, Salem, Jan.)
-------
139
How or in what way(s) would you say that the air pollution in Salem(Eugene)
is (better)(worse) than when we talked to you last May(September) ?
(Accurate responses continued)
The rain and winds clean the air a lot and in September we had a lot of the
burning and the air was more still. (F, 30-39, Wife of Finance Co.
District Manager, Salem, Jan. )
It (the air) is clear now aside from the fog. In the summer the visibility
is much worse. The field burning in summer and the sawdust up in the
Santiam area cause this. (M, 30-39, Physical Therapist, Salem, Jan. )
(Air pollution) hasn't burned my eyes this winter. When they were
burning the fields in September I could really feel it. We also had some
really bad days then. (F, 40-49, Wife of Meat Cutter, Salem, Jan. )
The wind and rain have diluted the pollution and increased the good fresh
air from the ocean. There are not as many forest fires, dust storms, etc.
in the winter, and not as much open burning either. (M, 40-49, Retail
Clothing Salesman, Salem, Jan. )
There are not as many particles in the air now. The air has been washed
and there is no field burning. (M, 50-59, Owner Retail Music Store,
Salem, Jan. )
(The air pollution is better) because the farmers quit burning. Now I can
see the mountains. There is more wind now and the rain has cleaned the
air. (M. 40-49, President of Underground Utility, Salem, Jan. )
There isn't as much smoke in the air now. There is less field burning and
I can see farther now than last summer. There is no doubt that there are
fewer cars now. The recent rains have no doubt cleaned the air some.
Some mills have closed down here and there, and .the gravel roads are not
so dusty. (M, 50-59, Laundry Washerman, Salem, Jan. )
I think it is simply the time of year. There is no field burning now and
no slash burning. The wind in winter brings in pollution from the Albany
mill when it is from the South and the mill here in town when the wind is
from the North, however. (F, 40-49, Wife of Service Station Owner,
Salem, Jan. )
I can see the mountains better from our house now. The atmosphere seems
better -- less smog around. (M, 20-29, Highway Department, Traffic
Analyst, Salem, Jan. )
I have not noticed any smells. There hasn't been any field burning now so
there is no smoke. There is not as much dust and I don't have to dust the
furniture as often. (F, 20-29, Wife of Bank Loan Officer, Fugene, Jan. )
-------
140
How or in what way(s) would you say that the air pollution in Salem(Eugene)
is (better) (worse) than when we talked to you last May (September) ?
(Accurate responses continued)
The air is cleaner now, there is better visibility and the air is easier to
breathe. The mill North of us got rid of their wigwam burner. Field
burning controls have improved the problem. (M, Over 60, Shipping and
Receiving Clerk, Fugene, Jan. )
They are not burning the fields, no forest fires, and there are not as many
cars and trucks on the road this time of year. (F, 40-49, Wife of Appliance
Parts Salesman, Fugene, Jan. )
Some of the industries are working on controlling their pollution now.
Field burning has stopped. The city of Fugene is passing an excessive
exhaust law and this will help. People are more informed but the only time
they get upset is during field burning. People seem to have accepted Weyer-
haeuser and its smell as part of the Fugene atmosphere. The non-leaded
gas, if everyone uses it, will help. The major companies are making great
efforts. The use of wigwam burners has been done away with and' this
will help the particulate pollution (problem). (F, 50-59, Wife of Railroad
Supervisor, Fugene, Jan. )
You don't have the black soot in the air now. The rains have definitely
cleared the air. It's as close to normal as can be. . I'm very upset by
how pollution bothers people with allergies and the threat (air pollution) is
to our wildlife. I really feel it's (air pollution) a lot better. It's not
perfect, but a lot better than in September. (F, 30-39, Wife of Dentist,
Fugene, Jan. )
They are not burning the fields now and the rain tends to clear the air.
The clouds may tend to hide the pollutants so they appear to be better. The
prevailing winds and rains clear the air. I'm sure we have more cars than
six months ago but the weather seems to give us cleaner air now. I don't
want them (undefined) to think they have done a lot to clean our air. They
haven't, the weather has cleaned it. (M, 30-39, Plumber-Steamfitter,
Fugene, Jan. )
It's clearer the air is clearer. You can see farther. (F, 40-49, Wife of
City Maintenance Worker, Fugene, Jan. )
It's quite a bit better because of the atmosphere as well as the controls.
The farmers aren't burning at all and there isn't any other burning. The
big waste burners are no longer used and I know Weyerhaeuser is trying
to reduce its pollution. (F, 50-59, District Court Register Clerk, Fugene,
Jan. )
-------
141
How or in what way(s) would you say that the air pollution in Salem (Eugene)
is (better) (worse) than when we talked to you last May (September)?
(Accurate responses continued)
I haven't noticed it (air pollution) burning my eyes like it did in September.
It might be just the different time of year. I don't notice it as bad as when
I first moved out here 12 years ago. (M, 50-59, Electronics, Eugene, Jan.)
Through visibility is the only way I have of judging it (air pollution). That
August and September period was bad every day. Now it might be bad only
every three days. It's nowhere near as evident. (M, 30-39, Wholesale
Drug Salesman, Eugene, Jan. )
It's (the air) clearer, not quite all the smoke and soot on the windows,
doorways and cars, etc. , now. They aren't field burning. (F, Over 60,
Wife of Physical Plant Maintenance Worker, Eugene, Jan. )
It (air pollution) has just cleared up. At that time (September) it was
smoky. I think this rain and cold weather has a lot to do with clearing the
air. The air is clearer and if you are out at night most of the stars are
visible. In September, only a few were visible once in a while. I just
think the air will eventually clean itself if we stop putting more stuff in it.
It's up to industry to take their money to research to eliminate the waste
products as much as possible. I notice the farmers are hollering for the
government to solve their pollution problem, but the industries are already
spending millions to work on theirs. I believe that the farming industry is
big enought to support research of their own. Let the farmers shell out
some of their profits. (M, 50-59, Filter Plant Operator, Eugene, Jan. )
-------
142
How or in what way(s) would you say that the air pollution in Salem (Eugene).
Is (better) (worse) than when we talked to you last May (September)?
(Inaccurate responses)
It isn't as smoky as it was then. I think maybe the heat and dryness cured
the fields somewhat. We had it bad for a few days when they were burning
but I think the cool evenings will help more. (F, 50-59, Drug Store Sales-
clerk, Salem, Sept. )
They just don't happen to be burning a field now as they were when you
talked to me before. It hasn't bothered me in the last couple of weeks.
(F, 20-29, Wife of Lumber Products Production Manager, Salem, Sept. )
Field burning is not nearly as smoky or on as many days as last year. I
do not smell the paper mills anymore. (M, 30-39, Psychiatric Aide,
Salem, Sept. )
Seems like the atmosphere is cleaner. It is not as smoggy or smoky. I
have no idea why this would be unless they are working to control it. We
haven't been able to burn all summer. (F, Over 60, Widow of Railroad
Yard Electrician, Salem, Sept. )
We have had more wind the last few days. It just isn't as smoky as it was.
A lot of smoke in the air would be better if they would allow them to burn
on a hot day instead they wait till a cloudy day and it all stays close to the
ground. (M, 40-49, State Fire Department Statistician, Salem, Sept. )
The atmosphere conditions in summer help carry it off so conditions are
better, but the situation itself is no better. (M, 20-29, Piano Sales Credit
Manager, Salem, Sept. )
They had controlled field burning this year. I think it helped a lot. No days
when the sun was obliterated. I know what we are doing at the plant. (F,
40-49, Wife of Accountant, Salem, Sept. )
Visibility--! listened to the pollution indexes from Portland, but I don't
know about here. I just grew up with it so I didn't realize it was pollution.
(F, 30-39, Secretary, Salem, Sept. )
The field burning has eased off. We have had a little rain. I lived here
all my life and it has always been smoky and it isn't the farmers. They
are getting blamed for big business who pollute the air 365 days a year
and the seed men burn their fields off in an hour. (M, Over 60, Newspaper
Circulation, Salem, Sept. )
The field burning isn't as bad as usual so it's been better since May.
(M, 40-49, Construction Contractor, Eugene, Sept. )
-------
143
How or in what way(s) would you say that the air pollution in Salem (Eugene)
is (better) (worse) than when we talked to you last May (September)?
(Inaccurate responses continued)
There's less smoke in the air. The reason it's been a good year for burning
wet grass makes more smoke and they are trying to control it. (M, Over 60,
Retired Farmer, Eugene, Sept. )
We have not had as much smell from Weyerhaeuser and Georgia-Pacific
isn't smoking as bad as they used to. They have done a good job with field
burning with a few exceptions when the wind turned. (M, 30-39, Mechanical
Engineer, Eugene, Sept. )
There aren't as many burned particles in the air. There is not as much
smoke. (The air) doesn't smell of smoke either. (F, 20-29, Wife of
Retail Sales Department Manager, Eugene, Sept. )
One's health is better. Individuals in the community have cared and done
more. Officers of the law call attention to the burning of trash. Police
have controlled dogs more in neighbors' yards. Mills have done a lot to
control and do away with stacks. Motor exhausts have been neglected.
Eugene has done just fine. I think something should be done about the
pollution in the rivers and the mill workers who are breathing in the dust.
(F, Over 60, Art Teacher, Eugene, Sept.)
Before we had lots of smoky days and couldn't see anything. Not so much
smoke now. Industry has tried better to do their best. Citizens are not
allowed to do outside burning and that helps. People are burned up about
that. (M, Over 60, Retired Meat Cutter, Eugene, Sept )
It's smoky for short periods of time, (but) it clears up early in the evening.
They have done a better job this year. (F, 20-29, Wife of Tree Planter,
Eugene, Sept. )
They have tried to control the burning so that it didn't go down the city.
I know: they had burning regulations. The state had it controlled. They
did it when the wind was just right. (F, Over 60, Wife of Physical Education
Equipment Supervisor, Eugene, Sept. )
There's been only two nights that we've come home and not been able to
see Spencer's Butte. Pollution in general is better. Not the field burning,
however. Those farmers make it cool, smoky and miserable for us.
I don't think we need to kill the agricultural owners living. (F, 50-59,
Wife of Architect, Eugene, Sept. )
There is industrial pollution over the East of us from the paper mills. It
is better because they are not burning fields but I imagine they'll burn again
soon as it stops raining. The Chairman of the Department of Environmental
Quality is a field burning farmer and I think that is absurd. (M, 20-29,
Attorney, Eugene, Sept. )
-------
144
How or in what way(s) would you say that the air pollution in Salem (Eugene) is
(better) (worse) than when we talked to you last May (September)?
(Inaccurate responses continued)
It's better because of the ban on field burning, but as far as every day
pollution goes, it is about the same. I don't recall getting odor from
Weyerhaeuser like I used to. (F, 30-39, Wife of Fire Department Dis-
patcher, Fugene, Sept. )
We haven't had quite as much smoke. Mondays have been bad for about
three weeks. Our pollution is only field burning. (F, 30-39, Wife of
Flectrician, Fugene, Sept. )
It is not quite so dense. I can see the sun once in awhile, at least. (F,
40-49, Housekeeper, Fugene, Sept. )
I don't see smoke in the air or valley now. I can't hardly tell till winter
comes. Winter is when you can tell it on the windows. It is greasy and
collects like fog, only it's blue stuff. (M, 40-49, Plywood Putty Patcher,
Fugene, Sept. )
They are burning only when the wind is blowing northeast. I think there
is less smoke. They have taken down several wigwams lately and the
foundry has been moved from our neighborhood since spring. (F, 40-49,
Wife of Mail Carrier, Fugene, Sept. )
It is better because the one sawdust mill is gone and there aren't as many
smoke burners now. When the wind is right it is a help because of the
fields (burning). One day, though, about everybody choked when the wind
turned. (F, Over 60, Wife of Foundry Worker, Fugene, Sept. )
In the winter it (weather) holds it (air pollution) down closer to the earth.
It gives me sinus drainage and plugs my nose up. Cars and industry
seem to be the things that cause it. I mean the paper mill especially.
(F, 20-29, Wife of Auto Mechanic, Salem, Jan. )
We get an awful lot of odor out here this winter. Paper or pulp mill smoke
(causes this). The summer field burning is bad. I suppose cars and trucks
also contribute. (F, 30-39, Wife of New Car Salesman, Salem, Jan. )
It seems that in comparison with previous winters, there is a lot more smog.
Very poor visibility. There is just a greater accumulation of air pollutants.
It must be industry because it could not be field burning now. (F, 40-49,
Wife of Community College Instructor, Salem, Jan. )
When the weather is calm, smog drifts down and settles and gets real bad.
Boise Cascade and the pulp mill at Albany (cause pollution). This combined
with carbon monoxide from cars and you really have something. (M, Over 60,
Baker, Salem, Jan. )
-------
145
How or in what way(s) would you say that the air pollution in Salem (Eugene)
is (better) (worse) than when we talked to you last May (September)?
(Inaccurate responses continued)
Weather conditions is the greatest reason for it. People are aware of the
problem but still haven't stopped burning and don't abide by the request.
It would mainly (be caused by) the burning and (people) will burn at any
time regardless of the request not tq. (F, 30-39, Wife of Mailman, Salem,
Jan. )
The smell as far as the mills go has gotten worse here. The smog makes
visability a little worse. There seems to be more and more of a build up
of air pollutants in the air all the time. (F, 20-29, Wife of Design
Draftsman, Salem, Jan. )
It (air pollution) has been noticed. Perhaps it is because I've been more
aware of it since talking to the first lady (interviewer). Before I'd have
thought it was fog. (F, 50-59, Fertilizer Sales Warehouse Foreman,
Salem, Jan. )
Nothing basically has been done (about air pollution). New industries
(are here) and the added population is causing more pollution from autos.
Certain climatic conditions (make it worse). (M, 40-49, School Principal,
Salem, Jan. )
It (air pollution) hurts your nose and makes your eyes run. A month or two
ago this may have been from more traffic. The new Fred Meyer store
creates more traffic, which means more car (pollution). Some of the cars
really throw off oil fumes. Lots of trucks taking orders to and from the
store, (Ff Over 60, Wife of State Forestry Radio Fngineer, Salem, Jan. )
There are still no controls and the production of pollution is growing.
There are more autos, more population and more industry and other
sources of pollution. The controls have not been put on population and
industry. (M, 50-59, VP of Engineering and Marketing Firm, Eugene, Jan.)
The air is so dense now. We didn't have a week of sun last year. It's true
that it was bad last September. I guess it was worse then. It's just that the
smoke is close to the ground now. When you're in Cottage Grove it's
always sunny and you look over here and there is always a cloud of smoke
hanging over. (F, 50-59, Wife of Locomotive Engineer, Eugene, Jan.)
Increased industrial, increased residential, and increased automotive
pollution (makes it worse now). (M, 20-29, Business Equipment Sales
Representative, Eugene, Jan. )
The last couple of weeks have been pretty ripe around here. The sewer
plant is on one side and Weyerhaeuser is on the other. (M, 50-59, Cemetery
Maintenance Worker, Eugene, Jan. )
-------
146
How or in what way(s) would you say that the air pollution in Salem (Eugene)
is (better) (worse) than when we talked to you last May (September).?
(Inaccurate responses continued)
When you live away from industrial parts, you don't notice it as much, but
it came out in the paper that snow recently was caused by air pollution
particles hitting the cold air which caused snow. They called it snow
pollution. (F, 50-59, Wife of Bakery Salesman, Eugene, Jan.)
Last week we had air pollution frozen in the air. Seems like with our fog
in our valley it (fog) holds it (pollution) in for a week at a time. The worst
thing is the irritation to your eyes. A woman had written in the paper
saying they cleaned up wigwam burners, but have to clear sawdust away
in diesel trucks which are polluting the air. In Garfield, creosol dumped
out of a pipe into the gutter and killed plants, and that eventually gets
into water. (F, 40-49, Wife of Athletic Department Worker, Eugene, Jan. )
------- |