Although free speech and a free press have been guaranteed rights in the
United States for more than two centuries, little is known about how the
public perceives the strength of the relationship between these rights.
Journalists and communications scholars sometimes assume that the public
needs to be educated about the inherent links between the two freedoms,1
but
no research has been conducted to date that attempts to examine public
opinion on this issue. Sporadic research on expressive rights issues has
estimated public support for topical rights, but none has yet compared
directly
support for media practices with support for similar kinds of speech
rights.
This study presents the development of parallel media rights and personal
expressive rights scales to attempt to measure the perceived relationship
between the two types of rights.
In the past few years, certain kinds of expressive rights have been the
focus of much public debate. Court cases involving flag-burning, cross-
burning, abortion protests and offensive song lyrics have fired controversy
across the country. Public support, or a lack thereof, for First
Amendment
rights does not in itself have the power to effect substantive legal
changes in
those rights. It is necessary, however, to understand the effect that
public
opinion can and frequently does have on the practice of expressive rights.
Letters to the editor, telephone trees and boycotts can chill what certain
societal groups say or what the media publish -- arguably as much as any
libel
law. What effect these kinds of negative messages have on their audiences'
support for other expressive rights has yet to be determined. In other
words,
the extent to which people generalize from one type of expression to
another
is unknown. Because some members of the public are willing to exercise
their own freedom of speech to quell the rights of others, however, it
appears
that the basic premises of the First Amendment are often misunderstood.
As
Robert Wyatt noted, "Americans believe that they believe in the First
Amendment, until they are confronted with expression that runs counter to
the grain of their beliefs" (Wyatt, 1991, p. 35; emphasis in original).
This
statement is a broad generalization, but research to date has not provided
enough insight to narrow it.
To be sure, journalists perceive the public to be less than supportive o
First Amendment rights. A reading of the opinions of some prominent
journalists serves to exemplify the attitudes of the profession. Columbia
Journalism Review editor James Boylan writes that journalists "see the
public
as a weak and vacillating guardian of that First Amendment right in which
the press is most interested" (Boylan, 1991, p. 42). (Ironically, this
comment
suggests that journalists are not necessarily any more consistent than the
public in their support for expressive rights.) Similarly, John
Seigenthaler --
former editor and publisher of the Nashville Tennessean and currently the
chairman of The Freedom Forum First Amendment Center -- differentiates
journalists and the general public:
"Those of us in journalism recognize that our Founding Fathers
had a vision of a society that welcomed controversial speech. We
must recognize, however, that this society today does not understand
or appreciate that vision" (Seigenthaler, 1992, p. 35).
Research has suggested that people are indeed more supportive of their
own individual rights to free speech than of media rights (Miller,
Andsager,
& Wyatt, 1992). (The differential cannot be accurately assessed because
the
items used in this survey were not parallel.) It is likely that people
perceive
free speech rights to be more personally relevant than media rights.
Research into support for free expression and other civil liberties date
back more than 50 years, and it remains a hot topic in social science
research.
It is difficult to draw accurate conclusions about American support for
expressive rights over time because the data lack uniformity (Erskine, 1970).
An extensive study of national public opinion surveys conducted over a 50-
year period found that very few questions on support for civil liberties
remained consistent for more than a couple of survey periods (Page &
Shapiro, 1992). Studies suggest, however, that the public tends to be
inconsistent in supporting personal rights, perhaps because surveys have
most often addressed expression issues relevant to the social and political
milieu of the time. Again, it seems likely that when an issue is debated
hotly
-- especially in a political setting, as was the flag-burning issue during
the 1988
presidential campaign -- people may have less opportunity to elaborate on
that issue for themselves without being distracted and misled by rhetoric.
A clearer picture emerges from comparison of the only survey questions
that have been consistent over time. Support for freedom of speech for
extremists is the one item that has remained a salient issue in survey
after
survey, undoubtedly due in large part to the classic Stouffer (1954) study
on
communism. For example, more than three-fourths (77%) of the
respondents in a survey conducted during World War II wanted to curtail the
rights of people to make speeches against certain races in the United
States
(Cantril, 1951). Before 1950, 49 percent was the highest portion of
respondents
who would support the expressive rights of extremists (Erskine, 1970).
Between 1950 and 1960, that number decreased to 29 percent, and it fell to
21
percent after 1960. This is a stark contrast to the 97 percent of
respondents
who said in 1940 that they believed in freedom of speech (Erskine, 1970).
One survey designed to measure the relationship between political
tolerance and support for civil liberties found that people differentiate
sharply
between different types of speech, with speech critical of the American
system
of government receiving most support and speech designed to incite an
audience to violence getting the least support (Gibson & Bingham, 1982).
Other surveys have asked anywhere from three to 58 questions designed to
measure support for specific rights (i.e., Immerwahr & Doble, 1982; McLeod
et
al., 1991, 1992; Wilson, 1975; Wyatt, 1991); each of these studies found a
wide
range of support among the rights. The same is true for media rights
(Andsager, 1992).
Sex has been a problematic variable in predicting support for freedom of
expression because it has not been consistent. The lack of conclusive
findings
about the relationship between sex and support for expressive rights is due
primarily to the fact that only a few of the studies to this point have
included
this variable. As for support of individual rights of free speech or
personal
freedom, some research has found negligible differences between the sexes
(Christenson & Dunlap, 1984; McLeod, Steele, Chi, & Huang, 1991; Prothro &
Grigg, 1960). More recent studies have shown that men are significantly
more
supportive of expressive rights than women (Miller, Andsager, & Wyatt,
1992), and there are differences in the way the sexes support various media
practices (Andsager, 1992).
Several studies have found that political conservatism significantly,
negatively affected support for civil liberties (Bobo & Licari, 1989;
Miller,
Andsager, & Wyatt, 1992). Wilson (1975) also notes that people who
consider
themselves liberal are more likely to allow "unpopular minority views" to
be
expressed.
These examples suggest that it is not enough merely to ask about support
for rights in a few scenarios, a practice that only blurs the picture of
public
opinion because it fails to provide respondents with any cognitive context.
The problem with the traditional approach to measuring support for
expressive rights is the fact that a few survey items assessed separately
fail to
reflect an individual's attitude toward expressive rights. Any number of
intervening variables can affect responses to single items. To better
measure
support for expressive rights, a reliable scale needs to be developed to attempt
to elicit the construct.
Scales for personal rights and media rights would allow a broad array of
items to be summed into indexes, which would provide insight into the
attitudinal component behind support for free speech and a free press.
Moreover, such scales that are composed of parallel items will allow for
direct
comparison of support for media rights with support for personal rights; in
order to measure the perceived links between free speech and free press, it
is
necessary to use parallel situations. Otherwise, the relationship may be
obscured by the artifacts of dissimilar rights.
Although the previous research comparing support for media and
personal rights dealt with nonparallel items, it is possible to extrapolate
from
the findings. As mentioned above, Miller, Andsager, and Wyatt (1992) found
that personal rights received much more public support than media rights.
The same finding, to a lesser extent, should emerge from items on parallel
rights. The smaller differential is expected because both media and
personal
expression encompass a wide variety of activities, and when people are
asked
about (nearly) identical slices of those activities, the gap that normally
results
from disparate media and personal rights items should narrow. That
personal rights should receive more support than media rights is a
reflection
of their personal relevance.
Origin of the parallel scales
Two expressive rights scales, which are composed of 13 items each, have
been written to be as parallel as possible in order to directly compare
support
for media and personal expressive rights. Some items in the scale are
derived
from Robert Wyatt's (1991) freedom of expression survey; others have been
added to reflect concerns regarding expressive rights that have been
addressed
recently by the courts and public opinion. For example, the topic of
political
correctness is included in three items (on each scale) -- the right to
offend
people of different races, religions and sexes. Protesting, a hot topic in
the last
few years due to the Persian Gulf War, flag-burning and the continuing
battle
over abortion rights, is also included. Although these activities are
generally
done by individuals, the media items are phrased as "the right to
editorialize"
about such expression. The two expressive rights scales are attached as
Appendix A.
The Support for Personal Rights Scale (SPRS) was first written as a 12-
item scale. Testing among a relatively homogeneous sample produced a
coefficient alpha of .86 for the original scale. Estimates of divergent
construct
validity were provided by correlating the SPRS with the Need for Cognition
Scale (Petty & Cacioppo, 1986) and with the Right-wing Authoritarianism
Scale (Altemeyer, 1981), which should be negatively related to support for
expressive rights because individuals high in right-wing authoritarianism
are
resistant to change, nontraditional or threatening ideas. As expected, the
SPRS was negatively correlated with the right-wing authoritarianism scale
(r=
-.46). The relationship was significant at the p<.05 level. The
correlation
between the NFC and SPRS scales was positive and low (r= .22). While this
value indicates a slight relationship between SPRS and NFC, it is not high
enough to suggest convergent validity, nor is it quite low enough for
discriminant validity. When the thirteen-item scales were developed,
seventy-six subjects enrolled in junior-level communications classes at a
southeastern university participated in an early pretest for reliability.
Even
with this relatively homogeneous sample of communications majors who
should be sensitized to First Amendment issues, the scales met minimum
standards for reliability. For the media rights scale (SMRS), the
coefficient
alpha was .75, and the coefficient alpha for SPRS was .70. The two scales
were
correlated at .68 (p < .01).
Validation with established scales
In order to validate the SMRS and the SPRS, two established scales were
offered to a group of subjects along with the newly developed scales. One
was
the twelve-item social criticism scale (Jessor & Jessor, 1977). The social
criticism construct is defined as the "degree of acceptance or rejection of
the
values, norms, and practices of the society" (Shaver, Robinson, &
Wrightsman, 1991, p.355). It has been found to be positively related to
problem behavior and, thus unconventionality. When tested with a
longitudinal college sample, the social criticism scale produced a
Cronbach's
alpha of .85. The scale includes items that deal with economic
opportunity,
governmental activities and performance, education and the environment.
Because people who tend to be highly critical of the society in which th
live should also be especially interested in obtaining more information
about
that society, social criticism and the two expressive rights scales should
be
positively correlated. To be able to criticize society, one must also be
able -- in
most cases -- to access information that is both good and bad. Therefore,
people who are high in social criticism should also be more likely to
support
expressive rights than those who are low. The notion that people who score
high on the social criticism scale are also frequently involved in problem
behavior (including, according to Jessor and Jessor, marijuana use),
however,
suggests that the positive correlation may be moderate at best. Because
personal expressive rights, perhaps including behavior or expression that
some may consider problematic, are more personally relevant than media
rights, the correlation between SPRS and social criticism should be
slightly
higher than between SMRS and social criticism.
The second established scale was the independence portion of
Wrightsman's philosphies of human nature scale (1974). The fourteen
independence items produced a corrected split-half reliability of .68, with
a
test-retest reliability of .75 after three months (Shaver, Robinson, &
Wrightsman, 1991). Independence was highly correlated with the
trustworthiness (r= .64) and altruism (r= .61) portions of the philosophies
of
human nature scale; it was slightly negatively correlated with complexity
(r=
-.16). Wrightsman conceptualizes independence as the extent to which
people can maintain their convictions in the face of pressures to conform.
Independence should be divergent from the expressive rights scales
because no research to date has suggested a link between the trait and
support
for free expression. That independence is negatively related to
complexity,
and support for personal expressive rights was found in a pretest to be
positively related to need for cognition -- a trait that entails some
cognitive
complexity -- further suggests that the independence construct is unrelated
to
support for expressive rights.
Method
The four scales -- SPRS, SMRS, social criticism and independence -- were
presented to 231 juniors and seniors enrolled in marketing courses at a
large
southeastern university. To reduce obtrusiveness of the expressive rights
measures, items from the two scales were mixed together in a random order.
Seven-point Likert-type scales were used to gather responses for each item.
The sample included majors from several colleges (business, arts and
sciences, communications, engineering). Subjects were told they were
participating in a survey of their attitudes toward society. They were
asked to
complete the four scales, along with recording their sex and political
orientation. As noted above, sex and political orientation have been found
to
predict levels of support for expressive rights, with men typically being
more
likely to support rights than women and people who identify themselves as
liberal more likely to support rights than those on the right. The sample
consisted of 112 women and 121 men. In terms of political orientation,
23.7
percent identified themselves as liberal or far left, 34.6 percent
identified
themselves as moderate and 32.5 percent said they were conservative or on
the far right. The remaining 9.2 percent did not know their political
orientation.
When the data were coded and entered into an SPSS file, each scale was
analyzed for reliability. Minimum reliability is .70 in order for a scale
to be
internally consistent, such that each item is contributing toward the
assessment of the construct in question (Nunnally, 1967). The coefficient
alpha for the SPRS scale was .83, and the coefficient alpha for SMRS was
.82.
These figures indicate that the scales are internally consistent, without
being
so high that the constructs are too narrow. For the social criticism
scale,
coefficient alpha was a low .68. This may reflect in part the moderate to
conservative nature of the sample. The coefficient alpha for the
independence scale was .79.
Correlations were obtained to gauge the relationships between each pair
of scales in the form of a multitrait correlation matrix. To further
validate
the expressive rights scales, t-tests were run to determine whether men and
women differentiated in their levels of support for the two kinds of
rights.
An analysis of variance was also used to look for differences in levels of
support by political orientation. Finally, for both the SPRS and the SMRS,
item-total and inter-item correlations were obtained to further document
the
reliability of the scales; to test for relationships between the two
scales, items
from each scale were correlated with the sum of the items from the other
scale. Principal-components factor analyses were conducted on each
expressive rights scale to check for unidimensionality.
Results
A multitrait correlation matrix was constructed to determine whether
the expressive rights scales were indeed divergent from the social
criticism
and independence scales (see Table 1). As predicted, both expressive
rights
scales were positively correlated with social criticism, but only lowly
(rSPRS= .14;
rSMRS= .11). This indicates that the two traits are divergent. The
personal rights
scale was significantly correlated with the social criticism scale at the p
< .05
level, suggesting that perhaps those who are interested in protecting their
personal expressive rights may want to do so in part to voice social
criticism.
The independence trait is even more divergent from the SPRS and SMRS,
with virtually no correlation (rSPRS= .05; rSMRS= .04).
The two expressive rights scales are, as might be expected, highly
correlated (r= .76; p < . 01). People who are willing to support one form
of
rights are fairly willing to support the other. The difference between the
two
scales lies in the mean levels of support. The SPRS had a mean level of
support of 61.6 (standard deviation = 12.2), while the mean support for
SMRS
was 51.1 with a standard deviation of 12.6. This disparity suggests that
while
the constructs are closely related, they are not entirely convergent.
Previous research mentioned above indicates that men and women
differ in support for expressive rights, and the greatest disparity lies in
support for media rights (Miller, Andsager, & Wyatt, 1992). T-tests to
further
validate the expressive rights scales support this finding. For the SPRS,
women's mean level of support was 58.9 and men's was 64.1. The difference
was significant at the p < .001 level (t229= -3.33). Mean support for SMRS
was
lower for both groups, with women's mean at 47.9 and men's mean 54.1 (t231=
-3.87; p < .0001). These mean levels of support are consistent with
previous
research that has attempted to examine differences between men and
women's support for expressive rights, further suggesting that the scales
are
measuring different constructs.
An ANOVA to further validate the findings relative to political
orientation found a significant difference among the political orientation
groups in terms of their support for expressive rights. For personal
rights,
subjects who identified themselves as liberal (mean support = 66.2) were
significantly more supportive than those who identified themselves as
conservative (mean support = 58.9; F5, 227= 2.58, p < .05). Moderates fell
in the
middle with a mean of 61.6. The same trends appeared for the SMRS scale
(F5,
227= 2.95; p < .05). Liberals had a mean of 56.6; moderates, 50.1; and
conservatives had a mean level of support of 49.0. Again, these
differences in
means lend support to the differences between the SPRS and SMRS
constructs.
SPRS
For both expressive rights scales, corrected item-total correlations and
inter-item correlations were obtained to further assess the psychometric
qualities of the scales. Item-total correlations evaluate individual items
on
the basis of total variance, and those with high correlations will increase
the
scale's variance, adding to the total reliability (Nunnally, 1967). To be
valuable, each item should have an item-total correlation of at least .20;
the
scale should have an average item-total correlation of .40. Inter-item
correlations should be positive and at least moderate.
The average item-total correlation for the SPRS was .47, indicating that
the items each contribute in a positive way to the scale's variance.
Individual
item-total correlations, displayed in Table 2, range from .32 for the right
to
offend people of another race or ethnic group to .57 for the right to
protest
military actions by the government. Means were obtained for each item; the
highest mean (6.0) occurred for the right to criticize the president, and
the
item receiving the least support (mean= 3.1) was the right to burn the flag
to
protest governmental actions.
Inter-item correlations were calculated. (See Table 3.) All were
positive,
ranging from .01 (offending another race and criticizing the president) to
.72
(viewing pictures of graphic sex acts and viewing material that contains
nudity). Nearly all correlations were significant. This finding further
suggests that all thirteen items should be retained. When the SPRS items
were correlated with the sum of media rights items, all were positively
related to the SMRS items, although none of the correlations was high.
This
indicates that the personal expressive rights and media rights are not the
same construct.
A factor analysis produced communalities ranging from .27 for
discussing false information to .81 for viewing nude photos. Although
communalities below .60 are undesirable (with a population of 200 or more),
the false information item was retained because it did not decrease total
reliability, and it correlated positively with the other scale items. Four
factors
with eigenvalues above 1.0 were produced, accounting for 63.5 percent of
the
variance. A scree test suggests that the four factors should be retained.
This
factor analysis suggests that perhaps more items might help to define the
SPRS construct because people apparently do not think about all forms of
expression in the same way, but as written the scale is reliable.
Taken together, the psychometric tests performed on the SPRS indicate
that it is a reliable scale that appears to measure the construct of
support for
personal expressive rights.
SMRS
Corrected item-total correlations for the media rights scale produced an
average of .45, which also surpasses the minimum requirement and suggests
that all items are contributing to the scale's reliability. Item-total
correlations
for the SMRS have a broader range than those of the SPRS: the lowest is
.28
for discussing the sex lives of public figures, and the highest is .58 for
editorializing in favor of burning the flag. Means for the media-rights
items
are slightly lower than for their parallel SPRS items. The highest level
of
support for media rights occurred for the right to criticize the president
(5.5),
the same item that received the most support in terms of personal rights.
For
media rights, the least-supported item was that of the media reporting
false
information they believe to be true (a situation dealing with libel), which
had
a mean of 2.4. Item-total correlations, means and standard deviations for the
SMRS are displayed in Table 4.
Inter-item correlations for SMRS are all positive, ranging from .02 to
.63.
(See Table 5.) The lowest correlation appeared between discussing the
sexual
habits of public figures and editorializing for or against abortion rights;
the
highest -- not surprisingly -- occurred between offending people of other
religions and offending people of other races or ethnic origins. Again,
the
majority of the inter-item correlations were significant. When SMRS items
were correlated with the sum of items from the personal rights scale, each
correlation was positive, but only one was high -- that of the media's
right to
offend people of other religions (r=.60).
A factor analysis showed that four factors emerged from the SMRS, each
with an eigenvalue of more than 1.2. The cumulative variance for the
thirteen items was 62.5 percent, which indicates that more items might
further account for this construct. Communalities ranged from .43 for
discussing false information to .75 for publishing nude photos. Scree
tests of
eigenvalues indicate that all four factors should be retained, despite the
low
communality of two items.
These properties suggest that the SMRS is also a reliable scale; the
differing levels of support and patterns of support indicate that the SMRS
is
not the same construct as support for personal rights, although they are
highly correlated.
Discussion and conclusion
The parallel scales developed in this study are designed to evaluate the
public's level of support for media rights and personal expressive rights.
To
date, public support for the two kinds of rights has never been directly
compared, although the public itself, scholars and journalists seem to make
assumptions about the variation in support based on surveys that address a
number of different and incomparable forms of expression. Use of parallel
scales can help public opinion and freedom of expression scholars to better
assess where -- or if -- disparities in public support for the media or
personal
rights lie. Such information would be useful for educators, especially
journalism educators, who might be able to better pinpoint how to increase
support for expressive rights among their students. Public information
campaigns designed to increase (or decrease) public support for certain
kinds
of expression would also find the parallel scales useful.
While the SPRS and SMRS have been found to be reliable and valid in
this study, they need to be tested among a larger, more heterogeneous
sample.
The sample used in this study was fairly broad in terms of interests and
political orientation, but limited by education, age and geography.
However,
because the scales had better-than-adequate reliabilities even in such a
homogeneous sample, they should remain reliable among better samples.
The question that needs to be addressed is validity. A known-groups
validity
test, perhaps among members of the American Civil Liberties Union and the
American Family Association (which intends to promote further media
regulation, especially for entertainment television and movies), would
provide further evidence of construct validity. Finally, a sample broader
in
terms of geography should also be studied because certain kinds of
expression
may be perceived differently in various areas depending on what is topical
or
relevant.
More items on each scale would only add to the reliability and variance
accounted for, but the scales currently meet psychometric requirements.
Whether the items included in these parallel scales will stand the test of
time
-- and for how long -- is unknown. Historically, survey items on public
support for free expression have fluctuated quite a bit to measure reaction
to
whatever topic is currently in the spotlight. The items on the SPRS and
SMRS are for the most part, however, fairly general. Over time, some may
have to be replaced or modified, but that is typical for most scales that
focus
on social attitudes. The main bodies of the scales should remain viable for
the foreseeable future because they are broad enough to encompass a number
of concerns that have been historically debated.
Although more study needs to be conducted on the parallel scales
presented here, the evidence to date suggests that these scales are
reliable and
divergent from other, somewhat related constructs. The assessment of
public
support for personal expressive rights and media rights to the same kinds
of
expression can be done quickly and easily using the SPRS and SMRS.
References
Altemeyer, B. (1981). Right-wing authoritarianism. Manitoba, Canada:
University of Manitoba Press.
Andsager, J.L. (1992, August). Differentiating media practices: how
demographic variables relate to support for freedom of the press. Paper
presented at the annual convention of the Association for Education in
Journalism and Mass Communication, Montreal.
Bobo, L., & Licari, F.C. (1989). Education and political tolerance:
testing the
effects of cognitive sophistication and target group affect. Public
Opinion
Quarterly, 53, 285-308.
Boylan, J. (1991, November/December). Pleading the First. Columbia
Journalism Review, 30, 41-43.
Cantril, H. (1951). Public opinion 1935-1946. Princeton, N.J.:
Princeton
University Press.
Christenson, J.A., & Dunlap, R.E. (1984). Freedom and equality in
American
political ideology: race and gender differences. Social Science
Quarterly,
65, 861-867.
Erskine, H. (1970). The polls: freedom of speech. Public Opinion
Quarterly,
34, 483-496.
Gibson, J.L., & Bingham, R.D. (1982). On the conceptualization and
measurement of political tolerance. The American Political Science
Review, 76, 603-620.
Immerwahr, J., & Doble, J. (1982). Public attitudes toward freedom of the
press. Public Opinion Quarterly, 46, 177-194.
McLeod, J.M., Guo, Z., Huang, K., Rzeszut, A.K., & Voakes, P.S. (1992,
August). Evaluating models of public support for First Amendment
rights. Paper presented at the annual convention for the Association
for Education in Journalism and Mass Communication, Montreal.
McLeod, J.M., Steele, C.A., Chi, H., & Huang, K. (1991, August).
Understanding political tolerance: alternative models of public support.
Paper presented at the annual convention of the Association for
Education in Journalism and Mass Communication, Boston.
Miller, M.M., Andsager, J.L., & Wyatt, R.O. (1992, August). How gender
and
select demographics relate to support for expressive rights. Paper
presented at the annual convention of the Association for Education in
Journalism and Mass Communication, Montreal.
Nunnally, J.C. (1967). Psychometric theory. New York: McGraw-Hill.
Page, B.I., & Shapiro, R.Y. (1992). The rational public: fifty years of
trends in
Americans' policy preferences. Chicago: University of Chicago Press.
Petty, R.E., & Cacioppo, J.T. (1986). Communication and persuasion:
central
and peripheral routes to attitude change. New York: Springer-Verlag.
Prothro, J.W., & Grigg, C.M. (1960). Fundamental principles of democracy:
bases of agreement and disagreement. Journal of Politics, 22, 276-294.
Seigenthaler, J. (1992). First Amendment controversy -- then and now.
The
Freedom Forum Annual Report 1991-1992. Arlington, VA: The
Freedom Forum.
Shaver, P.R., Robinson, J., & Wrightsman, L. (1991). Measures of
personality
and social psychological attitudes. New York: Harcourt Brace Jovanich.
Stouffer, S.A. (1955). Communism, conformity and civil liberties. Garden
City, N.Y.: Doubleday.
Wilson, W.C. (1975). Belief in freedom of speech and press. Journal of
Social
Issues, 31, 69-76.
Wrightsman, L.S. (1974). Assumptions about human nature: a social-
psychological approach. Monterey, CA: Brooks/Cole.
Wyatt, R.O. (1991). Free expression and the American public: a survey
commemorating the 200th anniversary of the First Amendment.
Murfreesboro, TN: the American Society of Newspaper Editors.
Table 1. Multitrait correlation matrix of the four scales.
SPRS SMRS Independence Social
criticism
SPRS 1.0
SMRS .76** 1.0
Independence .05 .04 1.0
Social criticism .14* .11 -.23** 1.0
* p < .05
** p < . 01
Table 2. Corrected item-total correlations, means and standard deviations
for
SPRS items.
item-total standard
correlation mean* deviation
Offend race .32 4.2 1.9
Foreign govt. .53 5.2 1.4
Criticize president .41 6.0 1.2
Protest military .57 5.2 1.4
Protest abortion rts. .33 5.4 1.5
Offend sex .58 4.2 1.8
Burn flag .53 3.1 2.1
Sex habits of public .45 4.9 1.6
Offend religion .56 4.7 1.6
View graphic sex acts .47 4.9 1.7
False information .40 4.1 1.7
Homosexuality .56 4.6 1.7
View nude photos .39 5.1 1.6
*On a 7-point Likert-scale, with 7 as strongly agree (or support this
right) and 1
as strongly disagree (do not support this right).
Table 3. Inter-item correlations for the SPRS.
1 2 3 4 5 6 7 8 9 10 11
1. Offend race 1.0
2. Foreign govt. .20a 1.0
3. Crit. president .01 .47a 1.0
4. Crit. military .10 .54a .43a 1.0
5. Protest abortion .15b .24a .14b .34a 1.0
6. Offend sex .44a .35a .24a .32a .22a 1.0
7. Burn flag .22a .34a .28a .47a .27a .25a 1.0
8. Sex -- public fig. .13 .25a .40a .25a .04 .36a .23a 1.0
9. Offend religion .45a .39a .17a .33a .32a .66a .26a .31
10. View sex acts .07 .26a .21a .29a .11 .27a .31a .27
11. False inform. .11 .16b .14b .29a .17a .30a .26a .33
12. Homosexual. .23a .30a .21a .37a .26a .33a .39a .31a .39
13. View nude
ph. .13b .21a .21a .21a .03 .18a .33a .23a .05 .72a
a Correlations are significant at p < .01.
b Correlations are significant at p < .05.
Table 4. Corrected item-total correlations, means and standard deviations
for
SMRS items.
item-total standard
correlation mean* deviation
Offend race .51 3.9 1.7
Foreign govt. .55 4.4 1.7
Criticize president .34 5.5 1.5
Protest military .54 4.7 1.6
Protest abortion rts. .43 4.4 1.8
Offend sex .43 3.6 1.8
Burn flag .58 3.2 2.0
Sex habits of public .28 4.1 1.6
Offend religion .55 4.1 1.7
View graphic sex acts .46 2.9 1.8
False information .33 2.4 1.5
Homosexuality .44 3.9 1.7
View nude photos .44 4.0 1.8
*On a 7-point Likert-scale, with 7 as strongly agree (or support this
right) and 1
as strongly disagree (do not support this right).
Table 5. Inter-item correlations for the SMRS.
1 2 3 4 5 6 7 8 9 10 11
1. Offend race 1.0
2. Foreign govt. .29a 1.0
3. Crit. president .19a .33a 1.0
4. Crit. military .31a .61a .35a 1.0
5. Protest abortion .19a .35a .14b .43a 1.0
6. Offend sex .55a .26a .05 .19a .19a 1.0
7. Burn flag .21a .48a .25a .41a .41a .25a 1.0
8. Sex -- public fig. .22a .18a .21a .13b .02 .21a .15b 1.0
9. Offend religion .63a .31a .24a .27a .20a .57a .32a .19
10. View sex acts .29a .22a .18a .19a .15b .28a .31a .19
11. False inform. .21a .12 .03 .17a .25a .16b .28a .09
12. Homosexual. .26a .29a .09 .30a .36a .17a .36a .04 .31a
13. View nude
ph. .09 .23a .31a .21a .23a .05 .38a .30a .15b .48
a Correlations are significant at p < .01.
b Correlations are significant at p < .05.
Appendix Support for Personal Rights Scale
disagree agree
1. An individual's right to protest for or against abortion rights should
be
protected by law.
2. An individual should have the right to publicly say that a foreign
government is better than the United States government.
3. An individual's right to make statements that would offend people from
a different racial or ethnic group should be protected by law.
4. *There should be a law against allowing an individual the right to buy
or
view material that features nude pictures.
5. By law, an individual should have the right to burn the American flag to
protest the actions of the government.
6. An individual's right should be protected by law when he or she makes
statements that offend people of different religious groups.
7. *There should be a law preventing an individual from criticizing the
president.
8. An individual's rights under the law should be protected when he or she
makes statements in public that advocate homosexuality.
9. An individual should have the right by law to make statements that
would offend people of a certain gender.
10. By law, an individual should have the right to buy or view material
that
includes pictures of sexual acts between consenting adults.
11. *There should be a law preventing an individual from discussing the
sexual habits of public figures.
12. The law should allow an individual the right to protest against
military
actions by the government.
13. An individual should have the right to discuss false information about
other people when he or she believes it is true.
*Indicates items that should be reverse-coded.
Support for Media Rights Scale
disagree agree
1. Under the law, the media should have the right to print or broadcast
statements that would offend people of a certain gender.
2. By law, the media should have the right to print or broadcast material
that includes pictures of sexual acts between consenting adults.
3. The media should have the right to print or broadcast false information
when they believe it is true.
4. By law, the media should have the right to editorialize in favor of
burning the American flag to protest the actions of the government.
5. The media's right should be protected when they publish or broadcast
statements that offend people of various religious groups.
6. *There should be a law preventing the media from discussing the sexual
habits of public figures.
7. By law, the media should have the right to editorialize for or against
abortion.
8. The law should protect the media's rights when they print or broadcast
public statements that advocate homosexuality.
9. *There should be a law preventing the media from criticizing the
president.
10. The media's right to print or broadcast statements that would offend
people of various racial or ethnic groups should be protected by law.
11. The law should allow the media the right to editorialize against
military
actions by the government.
12. The media should have the legal right to publicly say that a foreign
government is better than the United States government.
13. *There should be a law against allowing the media the right to print or
broadcast material that features nude pictures.
*Indicates items that should be reverse-coded.
1The need for newspapers to inform the public about the relationship
between personal expressive
rights and media rights guaranteed by the First Amendment was a major topic
of discussion at The
Freedom Forum First Amendment Center's first seminar on Nov. 4, 1992
(Nashville, Tenn.).
|
