Homosexuals and the Military:
An Analysis of the Spiral of Silence
William J. Gonzenbach, Ph.D.
Cynthia King Jablonski
Patrick M. Jablonski
University of Alabama
submitted to
Communication Theory and Methodology Division
Association for Education in Journalism and Mass Communication
Meeting in Atlanta, GA
August 1994
Abstract
Homosexuals and the Military:
An Analysis of the Spiral of Silence
This study examines Elisabeth Noelle-Neumann's spiral of silence theory about
the
controversial issue of whether homosexuals should be allowed to serve in
the United States
military. The study tests the theory's hypotheses in light of a content
analysis of the
media's treatment of the issue. Results provide qualified support
for the theory, and
suggestions are made for expansion of the model.
Homosexuals and the Military:
An Analysis of the Spiral of Silence
"For certainly, at the level of social life, what is called the adjustment of
man to his
environment takes place through the medium of fictions. By fictions
I do not mean lies.
I mean the representation of the environment which is in lesser or
greater degree made by
man himself."
Walter Lippmann
Public Opinion, 1922
Lippmann's representation of the environment subsumes two primary concepts, the
mass
media's representation of the environment and the individuals'
perceptions of the
environment. A primary question in public opinion formation is the
relationship of these
two concepts. Elisabeth Noelle-Neumann's spiral of silence theory
attempts to integrate
the concepts of individual perceptions of opinion about an issue and
the media's
representation of that issue into a theoretical explanation of the public
opinion for
mation process. This study tests the theory's hypotheses in relation
to the media's
representation of the controversial and morally-loaded issue of whether
homosexuals should
be allowed to serve in the United States military.
The Spiral of Silence
The spiral of silence theory is a broad-based examination of public opinion
encompassing
behavioral/attitudinal, communications and social theory (Donsbach
and Stevenson, 1984).
Noelle-Neumann develops a chain of hypotheses that chart the
formulation of public opinion
based upon the primary link in this chain, which is that people will
accommodate their
opinion to their perception of the majority opinion because they are
afraid to isolate
themselves from the majority (Noelle-Neumann, 1984, p. 39-40; Merten,
1985). This
conformity hypothesis drives Noelle-Neumann's definition of public
opinion, which she
defines as opinions on controversial issues that one can express in
public without
isolating oneself (Noelle-Neumann, 1984, pp. 62-63). Noelle-Neumann
posits two conditions
and four primary variables that must be examined in a test of the spiral of
silence:
the extent to which the issue is emotionally charged and morally
loaded; the intensity of
and slant in the way a topic is treated in trend-setting media;
majority opinion of the
issue; perception of the majority opinion; one's expectation about
success of one's
position in the future; and willingness to speak up publicly about one's
position on the
issue (1989, p. 20).
The Perception Hypothesis
The spiral of silence offers a fairly powerful-effects model of the media's
influence in
the formation of public opinion as suggested by the theory's first
hypothesis which is
that the media affect individuals' perceptions of the majority
opinion about an issue
(Media _ Perception of Majority Opinion). Noelle-Neumann hypothesizes
that individuals
make judgments about their perception of the majority opinion from
two sources, direct
observation and "the trend-setting media," and the theory in general
gives more specific
consideration to the role of the mass media, particularly
television, as an agent of
influence in the public opinion formation process than personal
observation of opinion
(see 1984, pp. 157-173). These two forms of observation may result
in a gap in perc
eption, which Noelle-Neumann terms the "dual climate of opinion" (1989).
Individuals who
have high exposure to "the media take a different view of the
climate of opinion and of
where the majority is to be found than persons who do not have much
exposure to the media
and essentially depend on their immediate observations " (1989, p.
21). Noelle-Neumann
hypothesizes that the mass media, particularly television, possess
three characteristics
that overcome the resistance of selective perception, and presumably
selective retention
as well: ubiquity, the over-powering presence of the media as a
source of information;
cumulation, the continuous repetition of messages in various media
and over time; and
consonance, the congruence of the journalists values and the content
of their messages
that is seen as a direct reflection of their values (Glynn and
McLeod, 1985). She argues
that the mass media play an important role through agenda-setting,
"according certain
persons and arguments special prestige and relaying the urgency or
chances of success for
the various opinions" (Salmon and Kline, 1985, p. 5).
The Conformity Hypothesis
Noelle-Neumann's definition of public opinion and the spiral of silence are
built upon the
fear of isolation hypothesis, which is a version of the conformity hypothesis
that
specifies the causal relationship of the next link in the theory's
chain that perception
of the majority opinion leads to one's opinion -- Perception of
Majority Opinion _ Opinion
(see Noelle-Neumann, 1974, 1979, 1984, 1989; also see Scherer, 1991).
Noelle-Neumann
hypothesizes that to avoid social isolation, the individual scans the
environment to
determine what opinions are held by others (what climate of opinion
exists) and to
determine the dominant opinion of an issue through quasi-statistical
perceptions. As the
individual scans the environment, the individual also perceives
which opinions are gaining
strength and which are declining (Merten, 1984, 31-32; Noelle-Neumann 1974, p.
45).
While Noelle-Neumann argues that fear of isolation drives the theory's
conformity
hypothesis, other researchers suggest alternatives to fear of isolation
as the source of
conformity. Some argue that positive social attraction may have
more influence on o
pinions and speaking out than fear of isolation (Glynn and McLeod, 1985;
Salmon and Kline,
1985), and others suggest that informational forces may also drive conformity
(Price and
Allen, 1990). Noelle-Neumann counters that positive motives play a
role in conformist
behavior, but "positive sanctions would not suffice to make most
members of a community
... strive for conformity (Noelle-Neumann, 1985, p. 70).
Gonzenbach's (1992) experimental
analysis of the Iran-Contra affair offered qualified support for the conformity
hypothesis (Under the threat of impending public scrutiny, Perception of
Majority Opinion
_ Opinion), however the study could not test the alternative of the conformity
hypothesis, which is the projection hypothesis -- Opinion _ Perception of
Majority
Opinion (see Scherer,1991). Scherer (1991) argues that the projection
hypothesis is a
better theoretical explanation of the relationship between opinion
and perception of the
majority opinion than is the conformity hypothesis. His panel study
addressed the order
of causality between the two competing hypotheses and indicated that
one's opinion led to
one's perception of the predominant opinion. The arguments against
the conformity
hypothesis and for the projection hypothesis are generally based on
research about the
inability of individuals to accurately gauge the opinions of others
(Salmon and
Kline,1984; Glynn and McLeod, 1984). Noelle-Neumann counters that the
quasi-statistical
sense has been interpreted as the ability to correctly assess actual
majority and minority
opinion, however "no such ability exists ... there is an ability to reliably
perceive an
increase or decrease in attitudes, which coincides with actual
developments in attitudes"
(1989, p. 21).
The Assessment of Future Trend Hypothesis
Noelle-Neumann's theory places a heavy emphasis on individuals' expectations
about the
future success of competing sides of an issue and the relationship of
the media's coverage
of an issue to these expectations. This expectation about future success of an
issue is
indirectly related to the media's presentation of the issue via the
perception hypothesis
and becomes a primary variable in the chain of relationships in the
theory that lead to
one's willingness to publicly speak out about the issue. Taylor's
study (1982) of the
spiral of silence hypothesized that one's opinion and one's
assessment of the majority
opinion are both assumed to influence one's judgment of the future
course of public
opinion. Taylor only specifically tested the hypothesis that "those who
favor the
majority position tend to believe that the future trend is toward greater
support for
their position (Opinion and Assessment of the Future are positively
related)" (p. 320).
He concluded that this hypothesis was "not evenly supported" for the
three issues he
examined (p. 328). However, as noted by Taylor, Noelle-Neumann implies
that the influence
of one's opinion on one's perception of the future course of opinion is
stronger for
those who believe they are in the majority (p. 320), therefore
suggesting that perception
of the majority opinion, while controlling for opinion, predicts
assessment of future
trend of the issue (Controlling for Opinion, Perception of Majority
Opinion _ Assessment
of Future Trend).
The Speaking Out Hypothesis
The final hypotheses of the theory examines the relationships among perception
of opinion
distribution and trend and willingness to publicly express one's
opinion and is a central
component in Noelle-Neumann's definition of public opinion --
opinions on controversial
issues that one can express in public without isolating oneself.
Noelle-Neumann
hypothesizes:
Willingness to expose one's views publicly varies according to the individual's
assessment
of the frequency distribution(perception of majority opinion) and the trend of
opinions
in his social environment(assessment of future trend). It is
greater if he believes his
own view is (perception of majority opinion), and will be
(assessment of future trend),
the dominating one or (though not dominating now) is becoming
more widespread (1974, p.
45; italics added).
One's willingness to speak out changes the climate of opinion so that the
dominant opinion
will become stronger. Conversely, if individuals think they hold the less
dominant
opinion, they will become more silent so that this opinion is less
frequently heard, and
therefore the less dominant opinion is diminished. This describes
the spiraling process
used in the theory's name (Merten, 1985). Taylor (1982) and Merten
(1985) also suggest
that opinion predicts willingness to express one's opinion.
Noelle-Neumann, however,
tends to focus more on perception of majority opinion and assessment
of future trend of
opinion as predictors of willingness to express one's opinion and
does not specifically
specify this hypothesis. Taylor found strong confirmation for this
hypothesis, even when
the relationship was controlled for by perception of majority
opinion and assessment of
future trend (1982, p. 332). Therefore, the speaking out hypothesis
can be modeled as
Opinion or Perception of Majority Opinion or Assessment of Future
Trend _ Speaking Out.
The Media's Treatment of the Issue
Content Analysis
Noelle-Neumann (1989) argues that a content analysis must be included in any
analysis of
the theory, because, except for rare exceptions, the predominance of
certain points of
view in the media always results in a greater willingness to speak up
by supporters of
those views. This study's content analysis examined all stories
about homosexuals in the
military presented by the early evening news of the three U.S.
television networks (ABC,
NBC and CBS), the New York Times, and the television news show
Nightline from September 1,
1992 to February 28, 1993. While these sources do not exhaust the avenues of
media
coverage, they were chosen because they are typically representative of
national news
coverage. The time frame of the content analyses is based upon
findings from
agenda-setting research. Winter and Eyal (1981) suggest that the "optimal
effect span" is
between 4 and 6 weeks: Stone and McCombs suggest that it takes 2 to 6 months
for changes
in the media agenda to be fully translated into the public agenda; and,
Shoemaker et al.
(1989) suggest maximum effects at 1 and 4 months.
The content analysis compromises 286 New York Times' articles and 77 television
news
stories for a total of 363 stories about homosexuals in the military.
---Table 1 about here---
The analysis presents nine mutually exclusive categories for the articles. The
categories
are defined as Workability, stories about the likelihood and advantages and
disadvantages
of lifting the ban on homosexuals in the military, as well as reports on
governmental
office candidate's election year stands and Presidential leanings on
the issue;[1] Lawsuits
and Laws, stories about state regulations and gay rights' laws and
lawsuits; Public
Opinion; stories focusing on public opinion including poll results and
comments on public
and group sentiment; International, stories on foreign country bans
and laws; Effects on
Other Issues, stories regarding the effects of maintaining or
eliminating the ban on other
issues and events; Government, stories about the interplay between the
military,
Pentagon, Joint Chiefs, Religious Right and Les Aspin; Violence and Gay
Rights Activities,
stories about violence against gays and activists attempts to stop the ban;
Discrimination, stories that discuss parallels with racial integration in the
military and
historical accounts of discrimination in the armed forces; and Miscellaneous,
stories
that would not fit in any of the specified category, most of which
only indirectly
concerned the issue of homosexuals in the military.
The Media's "Picture" of the Issue
The content analysis reveals that the media appeared to focus more on
governmental
interplay surrounding the ban rather than the direct emotional or social
ramifications of
the issue for the public. Over one-third (41.14%) of the stories
fell into one of two
categories, Workability or Government. The stories in these
categories dealt almost
exclusively with the posturing and interplay among various governmental
organizations.
Most of the coverage focused on conflict between President Clinton
and members of Congress
or Clinton and the military.
---- Table 2 about here ----
These findings should not be interpreted to mean that these battles did not
concern the
social and emotional ramifications of the issues. To the contrary,
most of these
exchanges dealt solely with such issues (e.g.. the morale of the troops,
gay rights, e
tc.). What is does indicate, however, is that such issues were framed
in terms of the
governmental camps that took sides on these issues. Relatively fewer
articles discussed
public and non-governmental group opinions (26). It is difficult to
predict what effect
this trend may have on public perceptions.
While a break down of story type offers important insight into the media's
picture of the
issue, story valence is, perhaps, a more meaningful distinction for
understanding the
impact of media coverage on public perceptions. Story valence refers
to the "slant" of
the article (for or against allowing homosexuals in the military).
Predictions made
according to the spiral of silence theory are based more on this
valence, than story
category.
Regarding the homosexuals in the military issue, few, if any stories directly
advocated a
specific position. Many stories, however, focused heavily on individuals or
events
supporting one side of the issue. Stories were placed into one of
three categories, For,
stories that focused mostly on support for homosexuals in the
military and opposition to
the ban; and Against, stories that focused on opposition to
homosexuals in the military
and support for the band. Few articles presented a truly neutral
or moderate picture of
the issue, although many articles did cover extreme positions on
both sides of the issue.
The third category, therefore, was labeled Mixed, stories that focused about
evenly on
both sides of the issue. Story valence was broken down by month.
---Insert Figure 1 about here---
Apart from an unusually large number of positive stories in November, the issue
does not
really appear to emerge in the media until January, 1993. Press
attention to the issue
increases in November due to a variety of factors, including
Clinton's stated intention to
end the ban, and the legal proceedings surrounding the reinstatement of Keith
Meinhold,
an admitted homosexual, to the Navy. Press attention fell in
December, but then reemerged
in January, following an off-handed remark by Clinton who stated that ending
the ban
would be an early priority of his administration. In November,
January and February, a
plurality of articles for each month were positive (supporting gays
in the military)
reflecting comments made by President Clinton and his staff. Fewer
articles focused
primarily on support for the ban. This trend suggests that,
regardless of the exact time
frame for media impact, the results should lead to increased perceptions of
support for
gays in the military and opposition to a ban.
Hypotheses
Based upon the theoretical considerations presented about the theory and content
analysis
of the issue, we specifically hypothesize the following:
Hypothesis 1: Perception Hypothesis
Media Exposure is associated with Perception of Majority Opinion.
Hypothesis 2: Fear of Isolation Hypothesis
Perception of Majority Opinion is associated with Opinion.
Hypothesis 3: Assessment of Future Trend Hypothesis
Perception of Majority Opinion, while controlling for Opinion, is associated
with
Assessment of Future Trend.
Hypotheses 4ABC: Speaking Out Hypothesis
4A. Opinion is associated with Speaking Out.
4B. Perception of Majority Opinion is associated with Speaking Out.
4C. Assessment of Future Trend is associated with Speaking Out.
Method
Data
The study is based on data from a national survey conducted by a research
institute in the
South. The survey was conducted over 8 nights from 5 p.m. to 11 p.m. (CST)
from March 7
through March 11 and from March 14 through March 16.. The findings
are from telephone
interviews, based on random-digit dialing techniques, with 480 adults
in the continental
United States, 18 years and older. The survey used three
call-backs, when needed, for
each number. For results based on samples of this size, one can say
with 95 percent
confidence that the error attributable to sampling and other random
effects could be 4.5
percentage points in either direction.
Treatment of variables
The dependent variables for each hypothesis were assessed on multiple choice
scales.
Don't knows were excluded from all analyses. Media was created from
the combination of
two items; days of the week watch national news, days of the week
read newspaper. The
variable was then dichotomized into low media exposure (0-7) and high
exposure (8-14).[2]
The remaining variables were assessed on five point scales (1-5),
including perception of
Majority (large majority supports...large majority opposes), Opinion
(strongly support
ban...strongly oppose ban), Future trend (position rapidly
gaining...position rapidly
losing), and Speak Out (never speak out...definitely speak out). Each
scale included a
neutral mid-point.[3]
Statistical Analysis: General Linear Models
The General Linear Models procedure in SAS 6.06 was used. An analysis of
variance (ANOVA)
was performed for each hypothesis. The use of this analysis is advantageous
because it
does not assume a strictly linear trend. Since media exposure is
expected to have very
different effects depending upon one's opinion (consonant or
dissonant with the majority),
this procedures allows for an examination of the differences among the various
groups.
Follow-up analyses to compare mean differences were performed using
Tukey's LSD technique.
Results
Hypothesis 1 (Media => Perception of Majority Opinion) is supported, F(df 1,435)
is
4.26.(p <0.0395); those with high media exposure were more likely to
believe that the
majority disagreed with them (x=3.44) than those with low media
exposure (x=3.05).
Hypothesis 2 (Perceptions of Majority Opinion => Opinion) was partially
supported. The
main effect was found to be significant, F(df 4, 425) = 24.24 (p <
0.0001). An analysis
of mean differences however, revealed interesting results. Since
media coverage typically
reflected positive support for homosexuals in the military, theory would
predict that
those who perceived themselves to be in the majority should have
extremely strong pro
homosexuals in the military opinions. What was actually found,
however, was that those
who believed about one-half or a only a slight majority of the
public supported their
opinion typically held the most positive opinions. In line with the
hypothesis, however,
those who perceived themselves to be in the minority typically held the most
negative
opinions.
---Table 3 about here---
Hypothesis 3 (Controlling for Opinion, Perception of Majority => Assessment of
Future
Trend) was not supported. Although the main effect was significant,
F(4, 392) = 3.59 (p <
0.0068), mean differences do not entirely follow the expected pattern.[4] Those
who
perceived themselves to be in the majority were not the most likely to
believe that their
position was gaining strength. Again, it was the people who
believed themselves to be in
the mid-range of public sentiment who were more likely to believe
their opinions were
gaining strength. In contrast to predictions, those who believed
themselves currently in
agreement with a large majority were least likely to believe that
their opinions were
gaining strength.
---Table 4 about here---
Finally, Hypothesis 4B (Perception of Majority Opinion => Speaking Out) was not
supported,
however Hypothesis 4A (Opinion => Speaking Out) was strongly supported, F(4,
297) = 11.95
(p < 0.0001) and Hypothesis 4C (Assessment of Future Trend => Speaking Out)
received
marginal support, F(4,297) = 2.08 (p < 0.0834). Hypothesis 4A
followed predictions.
Individuals with the most polarized positions (positive or negative)
were more willing to
speak out.
---Table 5 about here---
The main effect for Hypothesis 4C was significant at the .1 level. However,
predictions
were only partially supported, with those who believed their
position was rapidly gaining
strength more likely to speak out than those who believed their
position was remaining
about the same.
---Table 6 about here---
Analysis
The results offer partial support for the spiral of silence. The findings,
however, also
suggest interesting twists that are possibly explained by the
media's treatment of the
issue. Based on the test of Hypothesis 1, the respondents' exposure
to the media is
negatively, rather than positively, related to their perception of
majority opinion; high
media exposure leads to decreased perceptions that one is in the
majority. An examination
of the means, however, indicates that while those with low media exposure are
likely to
believe that they are in the majority, those with high exposure do
not think that they are
in the minority, but rather believe that about half of the people agree with
their
position. Recall, however, that only 8.22% of media coverage (26
articles) primarily
concerned public opinion. Independent analysis of this category
reveals that an equal
number of articles (7 each) suggested either primarily favorable
(For) or unfavorable
(Against) public attitudes regarding gays in the military, while a
plurality of the
articles (12) presented a mixed view of public opinion. It stands to
reason, therefore,
that such a mixed picture of public opinion would lead individuals
with high media
exposure to perceive only partial public support regardless of one's
initial disposition.
The relationship between majority opinion and one's own opinion might be
similarly
explained. Theory suggests that individuals who believe they are
clearly in the majority
should subsequently express more extreme opinions. In this study,
however, it was found
that individuals who believed that only about one-half or a slight
majority of the public
agreed with them expressed the most extreme opinions in support of
homosexuals in the
military. These results may be explained by an alternative model
where media exposure
influences both perception of majority opinion and one's own opinion
directly. This would
produce a deceptive correlation between majority opinion and personal opinion
that might
be falsely interpreted as a causal relationship.
It also is extremely likely that one's opinions are not influenced as much by
perceptions
of "mass" opinion, as they are influenced by perceptions of the opinions of
respected
sub-groups (eg. Salmon and Kline, 1985; Glynn and McLeod, 1985).
Since the majority of
stories concerned governmental groups and official support for gays
in the military,
viewers opinions may have been influenced more by them than by their
perceptions of
"public" opinion.
The findings regarding the impact of perceptions of the majority on future
trends are
more difficult to interpret, but also might be largely explained by
the media picture. It
was predicted that those who perceived current strong public support for their
opinion
would also perceive strong future support. This prediction was not
supported.
The fact that stories regarding governmental action focused primarily on
opposition to
the ban and support for gays in the military may have acted as a cue
for future trends.
People may interpret these actions as indicative of an emerging
change in public
sentiment. Therefore, as the findings suggest, those who believed that
their position
(strong opposition to the ban as indicated by the relationship
between majority and
opinion) was, at that time, only supported by about half or a slight
majority of the
people may have sensed a future shift in their favor.
Finally, both opinion and assessment of future trends, but not perceptions of
majority
opinion, were significantly related to one's willingness to speak
out on the issue. The
lack of effect for perceptions of the majority might be explained by
the fact that people
are not attempting to conform to an undifferentiated majority, but
as suggested, are only
concerned about and influenced by certain sub-groups. As predicted,
those with more
polarized opinions (positive or negative) were more likely to speak out
on the issue. The
impact of the assessment of future trends was only partially anticipated.
Those who felt
their position was rapidly gaining strength expressed the greatest willingness
to speak
out. Those who believed their position was losing strength,
however, were not
significantly less likely to speak out. In fact, the only significant
difference was that
those who perceived only a moderate change in public support for their position
were less
likely to speak out than those who perceived a large favorable shift. This
might be
explained by the fact that those who believed themselves to be in the
minority on this
issue typically held extreme negative positions regarding homosexuals
in the military. The
impact of their opinions, therefore, may have that may have outweighed their
desire to
conform to public sentiments. Yet, there was no significant
interaction between Majority
and Opinion on Speak Out. These results are likely explained by
other confounding factors
not included in this model.
Conclusions
This study examined Elisabeth Noelle-Neumann's spiral of silence theory about
the
controversial issue of allowing homosexuals in the military. A content
analysis of media
coverage of the issue revealed that most articles focused on
governmental action and
policies. While most articles focused on support for gays in the
military, media coverage
suggested a mixed picture of public opinion. The results of the study do not
support a
priori predictions suggested by the spiral of silence theory. Upon
closer inspection of
the media's treatment of the issue, however, the results were found
to offer qualified
support for the model.
Most significantly, support was not found for the hypothesis that perception of
majority
opinion influences one's willingness to speak out on the issue. In
line with past
research (Salmon and Kline, 1985; Glynn and McLeod, 1985) it is suggested
that one's
opinions and willingness to speak out are not influenced by perceptions
of "mass" public
opinion, but by the opinions of respected sub-groups. It is
speculated, therefore, that
those who support homosexuals in the military may be more willing to
speak out on the
issue due to perceptions that respected groups and officials strongly
support their
opinions, even though the general public may not. To clarify these
findings, future
research should attempt to expand the model by including other
variables and assessing the
media's impact on perceptions of key subgroups as well as the subsequent impact
of these
perceptions on an individual's personal opinions and willingness to
speak out.
References
Donsbach, Wolfgang., and Robert L. Stevenson, R. L. 1984. "Challenges,
Problems and
Empirical Evidence of the Theory of the Spiral of Silence." Paper
presented to the Panel
on the Spiral of Silence Theory at the Conference of the
International Communication
Association, San Francisco, CA.
Fields, James M., and Howard Schuman. 1976. "Public Beliefs about the Beliefs
of the
Public." Public Opinion Quarterly 40:427-448.
Glynn, Carroll. J., and Jack M. McLeod. 1985. "Implications of the Spiral of
Silence
Theory for Communication and Public Opinion Research." In Political
Communication
Yearbook, 1984, eds. K. R. Sanders, Keith R., L. E. Kaid, and Dan
Nimmo, pp. 43-65.
Carbondale: Southern Illinois University Press.
Gonzenbach, William (1992). "The Conformity Hypothesis: Empirical
Considerations for the
Spiral of Silence's First Link," Journalism Quarterly, 69: 633-645.
Merten, Klaus. 1985. "Some Silence in the Spiral of Silence." In Political
Communication Yearbook, 1984, eds. K. R. Sanders, Keith R., L. E. Kaid, and
Dan Nimmo, pp.
31-41. Carbondale: Southern Illinois University Press.
Noelle-Neumann, Elisabeth. 1974. "The Spiral of Silence: A Theory of Public
Opinion."
Journal of Communication 24:43-51.
Noelle-Neumann, Elisabeth. 1979. "Public Opinion and the Classical Tradition."
Public
Opinion Quarterly 43:143-156.
Noelle-Neumann, Elisabeth. 1984. The Spiral of Silence: Public Opinion -- Our
Social
Skin. Chicago: The University of Chicago Press.
Noelle-Neumann, Elisabeth. 1985. "The Spiral of Silence: A Response." In
Political
Communication Yearbook, 1984, eds. K. R. Sanders, Keith R., L. E.
Kaid, and Dan Nimmo, pp.
66-94. Carbondale: Southern Illinois University Press.
Noelle-Neumann, Elisabeth. 1989. "Advances in Spiral of Silence Research."
KEIO
Communication Review 10:3-29.
O'Gorman, Hubert. 1975. "Pluralistic Ignorance and White Estimates of White
Support for
Racial Segregation." Public Opinion Quarterly 39:313-330.
O'Gorman, Hubert. 1979. "White and Black Perceptions of Racial Values."
Public Opinion
Quarterly 43:48-59.
O'Gorman, Hubert. 1980. "False Consciousness of Kind: Pluralistic Ignorance
Among the
Aged." Research on Aging 2:105-128.
O'Gorman, H., & Garry, S. (1976). "Pluralistic ignorance -- a replication and
extension." Public Opinion Quarterly, 40: 449-458.
Price, Vincent and Allen, Scott (1990). "Opinion Spirals, Silent and
Otherwise."
Communication Research 17:369-392.
Scherer, Helmut (1991). "Personal Opinions and the Perception of Public
Opinion? A Test
of Two Alternative Models." Paper presented to the Political Communications
Division at
the Conference of the International Communication Association,
Chicago, IL.
Salmon, Charles. T., and F. Gerald Kline. 1985. "The Spiral of Silence Ten
Years Later:
An Examination and Evaluation." In Political Communication Yearbook, 1984,
Eds. K. R.
Sanders, Keith R., L. E. Kaid, and Dan Nimmo, pp. 3-29. Carbondale:
Southern Illinois
University Press.
Scheff, Thomas J. 1967. "Toward a Sociological Model of Consensus." American
Sociological Review, 32:32-46.
Taylor, D. Garth. 1982. "Pluralistic Ignorance and the Spiral of Silence: A
Formal
Analysis." Public Opinion Quarterly 46:310-335.
Table 1. Summary of News Coverage by Source
Source Frequency Percent
New York Times 346 83.17
ABC 19 4.57
NBC 25 6.02
CBS 22 5.28
Nightline 4 .96
Total 316 100.00
Table 2. Summary of News Coverage by Story Category
Category Frequency Percent
Government 77 24.36
Workability 53 16.78
Effect On Other Issues 48 15.20
Violence/Gay Activities 41 12.97
Lawsuits and Laws 29 9.17
Public Opinion 26 8.22
Miscellaneous 22 6.96
Discrimination 12 3.80
International 8 2.53
Total 316 100.00
Table 3 Personal Opinion by Perception of Majority Opinion
--------------------------------------------------------------------------------
----
--
Perception of
Majority Opinion Mean Personal Opinion
--------------------------------------------------------------------------------
----
--
Large Majority Supports 2.73c
Majority Supports 3.79a
Half Support 3.63b
Majority Not Support 2.51c
Large Majority Not Support 1.84c,d
--------------------------------------------------------------------------------
----
--
Note. Opinion scores could range for 1 to 5 with 1 being definitely banned and
5
being definitely allowed. Mean scores not sharing at least 1
superscript letter
differ significantly at p < 0.05 by Tukey's HSD.
Table 4 Assessment of Future Trend by Perception of Majority Opinion
--------------------------------------------------------------------------------
----
--
Perception of Mean Assessment
Majority Opinion of Future Trend
--------------------------------------------------------------------------------
----
--
Large Majority Supports 2.73c
Majority Supports 3.79a
Half Support 3.63b
Majority Not Support 2.51c
Large Majority Not Support 1.84c,d
--------------------------------------------------------------------------------
----
--
Note. Assessment scores could range for 1 to 5 with 1 being opinion rapidly
losing
strength and 5 being rapidly gaining strength. Mean scores not
sharing at least 1
superscript letter differ significantly at p < 0.05 by Tukey's HSD.
Table 5 Speak
Out About Issue by Personal Opinion
--------------------------------------------------------------------------------
----
--
Personal Opinion Speak Out
--------------------------------------------------------------------------------
----
--
Definitely Banned 3.89a
Probably Banned 2.98b,c
Not Sure 2.93b,c
Probably Allowed 2.87c
Definitely Allowed 3.61a,b
--------------------------------------------------------------------------------
----
--
Note. Speak Out scores could range for 1 to 5 with 1 being never speak out and
5
being definitely speak out. Mean scores not sharing at least 1
superscript letter
differ significantly at p < 0.05 by Tukey's HSD.
Table 6 Speak Out About Issue by Assessment of Future Trend
--------------------------------------------------------------------------------
----
--
Future Trend Speak Out
--------------------------------------------------------------------------------
----
--
Position Rapidly Losing Strength 3.29a,b
Position Losing Strength 3.55a,b
Position Staying the Same 3.27a
Position Gaining Strength 3.61a,b
Position Rapidly Gaining Strength 4.11b
--------------------------------------------------------------------------------
----
--
Note. Speak Out scores could range for 1 to 5 with 1 being never speak out and
5 being
definitely speak out. Mean scores not sharing at least 1
superscript letter differ
significantly at p < 0.05 by Tukey's HSD.
[1] This category may appear diverse, however,
these topics were almost always found in
the same article. This problem
might have been resolved by cross referencing articles
into more than one
category. However, for convenience sake and because people are usually
left with only a
n overall impression of an article rather than an understanding of
details
, exclusive categorization was used.
[2] Analysis was also considered isolating only t
hose with extremely low coverage (0
-1) or high coverage (14) but extreme
ly skewed cell sizes compared to variable level made
such analysis difficu
lt.
[3] Again, analyses was tried with collapsed variable categories, with no
significan
t
differences. Control variables including education, church
attendance,
and political
ideology were originally included in the analysis. While bo
th education and church att
endance had significant main effects in portion
s of the analysis, they did not
significantly add or detract from the ove
rall models and so were excluded from the final
analysis.
[4] Analysis w
as also run separately for those opposed and in favor of gays in the
milit
ary, but no significant differences from the full model were found.
|
