|
Participation and media consumptionD
Participation in Community Organizations and Consumption of TV and Newspaper
News
Esther Thorson
Glenn Leshner
University of Missouri-Columbia
All Correspondence to
Glenn Leshner
University of Missouri
283 Gannett Hall
Columbia, MO 65211
voice: 573/884-6676
email: [log in to unmask]
Running Head: Participation and Media Consumption
Thorson is a professor and Leshner is an assistant professor.
Paper submitted to the Mass Communication and Society Division of the
Association for Education in Journalism & Mass Communication, April 1, 1998
Participation in Community Organizations and Consumption of TV and Newspaper
News
Abstract
This study examined how habitual and "special" news consumption was related to
three measures important for indexing how people are integrated with their own
communities: participation in community organizations, voting, and knowledge
about issues and occurrences important to their community. The hypothesis that
news consumption is an important variable in how people relate to their
communities was tested in five U.S. cities, each of which had experienced some
kind of news media intervention (i.e., the "special news"). The data consisted
of approximately 600 random-digit-dialed phone interviews in each city. The
results showed that indeed the classic belief that news makes a difference was
clearly supported, even after stringent statistical controls for demographic
variation were introduced. Perhaps more importantly, the special news
contributed to the dependent variables even after habitual news consumption was
controlled. There was little widening of sociocultural gaps that typically occur
as a function of such variables as race, gender, and time a person has lived in
the city, but there were a few examples of both gap widening and gap
diminishing. Of course the gap diminishing effects are especially intriguing as
they suggest a socially positive role that "special news" can play.
Participation in Community Organizations and Consumption of TV and Newspaper
News
When people consider community activity and its antecedents, they usually begin
the discussion by referring to one of American's most prestigious visitors,
Frenchman Alexis de Tocqueville. Tocqueville wrote in the 1830's that
Americans of all ages, all stations in life, and all types of
disposition are forever forming associations. There are not only
commercial and industrial associations in which all take part, but others
of a thousand different types--religious, moral, serious, futile, very
general and very limited, immensely large and very minute. . . . Nothing,
in my view, deserves more attention than the intellectual and moral
associations in America. (de Tocqueville, 1969, pp. 513).
Researchers from a great variety of areas have pointed to the importance of
this tendency to form associations, and indeed, the idea of participation in
associations has come to form a central feature of what has been termed social
capital. Friedland, Sotirovic and Daily (in press) define social capital as
"those stocks of social trust, norms and networks that people can draw on to
solve common problems." They go on to note that:
Networks of civic engagement, such as neighborhood associations,
sports clubs and cooperatives are essential forms of social capital. The
denser these networks, according to social capital theory, the more likely
that members of a community will cooperate for mutual benefit. (p. 6).
Unfortunately, considerable evidence has been accumulating that participation
in community activity has been on the decline for the last thirty years (e.g.,
Burnstein, 1972; Teixeira, 1987, Rosenston & Hansen, 1993; Putnam, 1995a;
1995b). On the other hand, some argue that civic life is not declining. Verba,
Schlozman, and Brady (1995) reported a massive study of civic participation in
which they looked at a range of political participation: voting, getting
involved in campaigns, making political contributions, working informally in the
community; attending protests or demonstrations, serving without pay on local
boards, and being politically active through voluntary organizations.
Interestingly, they found that almost half of their interviewees reported being
associated with an organization that takes stands in politics, and almost 20%
reported having working with people in the neighborhood on some problem. Almost
15% attended local board meetings. Their study seems to indicate a rather high
level of participation in community activities. Friedland et al have discussed
a number of other arguments against the notion that social capital is declining
(in press).
For our purposes, however, the issue is not whether community participation is
declining, but rather what antecedent variables are associated with various
levels of community participation. Putnam has argued that television is the
cause of diminished participation (1995a and 1995b). Merritt (1995) argues that
traditional journalism uses definitions of news that primarily come from outside
a community and that this causes feelings of detachment from the news and the
issues in one's own community. He and others argue that "civic journalism,"
which has been called a "reform movement within journalism institutions"
(Friedland et al., in press), "invites people to become a public" and "calls on
the press to help revive civic life and improve public dialogue (Rosen, 1995, p.
1). Thus the news has also been argued to be relevant to community
participation.
But is there any evidence that news consumption is related to civic
participation? The relationship between local news use and community involvement
has been examined for a long time, although the number of studies is quite
limited and review of them shows we really do not know much about the linkage.
In the early 1950's, Janowitz (1952) conceptualized the community press as an
instrument for integrating the individual into society. The news media were
thought of as providing social significance for local communities whose identity
and cohesion were threatened by industrialization.
Nearly 30 years later, Tichenor, Donohew and Oilen (1980) looked at how 19
communities in Minnesota dealt with issues of conflict (e.g., whether to allow
the location of a nuclear power plant in their town) and how that behavior
related to the news media in their own community. The research asked whether
citizens in isolated rural communities would have the same amount and quality of
information as citizens in an urban area. Weekly newspapers were not seen to
provide the same kind of information as dailies and to suppress citizen use of
the newspaper as a source of critical information that would enable them to make
decisions about controversial issues. In other words, the idea was that
different kinds of media provide different quality of news and the quality of
news has a major impact on citizen activity and effectiveness in the face of
important decision-making.
Stamm and Fortini-Campbell (1983) looked directly at the correlations between
various measures of community ties and newspaper use. They found that
subscribing to a newspaper and the amount of reading an individual reported were
correlated with how closely people perceived themselves as tied to communities.
For example, homeowners were more likely to subscribe than non-homeowners.
Homeowners also perceived newspapers as being more useful than did
non-homeowners. Stamm and Fortini-Campbell also considered the idea that
newspaper consumption might foster the development of community ties. They
argued that if this were so, there would be growth in community ties after
individuals became subscribers. The researchers found that, indeed, the longer
people had been subscribing, the more they perceived themselves as close to and
involved in their community.
Last year, Stamm and his colleagues (Stamm, Emig, & Hesse, 1997) returned to
this issue by looking at the linkage between news media consumption and
interpersonal communication, and involvement in the community measured by
variables like paying attention to community matters, having ideas for improving
community, working for change, and getting people together to talk about issues.
They found that newspaper and TV news correlations with community involvement
were higher for those who were drifting or setting into a community than for
those who were settled. Interestingly, however, the correlations between
community activity and interpersonal communication were high for all the
categories of community linkage (drifting, settling, settled and relocating),
and Stamm et al conclude that perhaps interpersonal communication is therefore
more important than news media consumption at all stages of involvement in a
community.
On the other hand, Martinelli and Chaffee (1996) showed that the political
knowledge of those who had just become US citizens and were coming up on their
first election was very well predicted by newspaper and television news exposure
and attention, and by their level of recall of political ads. Although it
should be noted that there was only a 20% response rate in this study, these
results show lend credence to the idea that media affect at least the cognitive
aspects of community involvement.
It should be noted, however, that most of the studies examining the link
between media consumption and community participation look at perceptions and
knowledge. They do not involve measures of behavioral participation, even as
self-reported. It is that gap that the present paper attempts to begin to fill.
The main purpose of this paper is to examine in five very different
communities, the linkage between activity in community organizations and two
measures of consumption of the news. The first is best called "habitual
consumption," which is how often people report watching and reading their
broadcast and print local media. The second might be termed special news
awareness. For our purposes, "special news" is simply any domain or issue that
is talked about a lot. It could be something that is a news event that gets a
lot of time or ink, or it could be a "project" that the news media do, here,
some of our cases involving civic journalism projects. In fact, however, what
the special news is, is not critical. It simply has to get enough coverage that
a sizable enough number of people report awareness of it that those people can
be studied. In the secondary analysis that we report here, data from three of
the cases resulted from attempts to measure the impact of civic journalism
special projects; data from two of the cases resulted from attempts to measure
the impact of a community health campaign.
It should be noted prominently that we know that correlations are only one
element of a cause-effect relationship. The data reported here are
cross-sectional, so we cannot answer questions about time-order, nor can we
fully rule out (even with control measures that are quite elaborate) alternative
explanations. On the other hand, it is important to note that if media
consumption and community activity are uncorrelated, the question of time order
and causal linkage are irrelevant. A cross-sectional survey that measures all
variables the theorist considers relevant is a good way to test the viability of
a number of effects hypotheses. Those that survive can then be subjected to
methodologies that allow causal inferences to be made (Chaffee, McDevitt, &
Thorson, 1997).
Hypotheses
The basic approach here is to examine via multiple regression analyses in which
stringent demographic controls are first applied, the contributions of news and
special news consumption to (1) how many community organizations people report
belonging to; (2) self-reported voting behavior (either being registered or
having actually voted in the last election); and (3) knowledge of the content of
the special news manipulation. These linkages are reported in five quite
different cities (shown in Table 1). As can be seen, there is variation in
population, percentage of minorities (and the ethnic background of these
minorities), area of the country they are located in, and the kind of "special
news" that their community was exposed to.
Table 1. Descriptive Characteristics of the Five Cities
City
Population
(in thousands)
Race
(white)
% Voted
Special News Topic
Binghamton
50
89.9
87.2
community problems
Kansas City
150
85.6
89.4*
second-hand smoke
Madison
215
94.1
88.2
juvenile delinquency & urban growth
San Francisco
750
65.2
85.2
mayoral race
Wichita
304
80.7
80.7*
second-hand smoke
Note: * registered to vote
Because there has not been any study of the amount of community participation
that people engage in as a function of what city they live in, we had no apriori
expectations about finding differences in amount of community participation in
our five cities. We also had no particular expectations about the amount of
self-reported voting. Finally, we had no way to compare the effectiveness of
the "special news" that occurred in each city, and thus, we did not expect
specific differences in the amount of knowledge people had about the special
news content.
We hypothesized, however, regardless of community participation and voting
levels, we'd find similar patterns in the:
H1: linkage between habitual and special TV and newspaper news
consumption and community participation;
H2: linkage between habitual and special TV and newspaper news
consumption and voting;
H3: linkage between habitual and special TV and newspaper news
consumption and topic-specific knowledge.
It should be noted that there were nine knowledge items in the San Francisco
questionnaire, 13 in the Binghamton and 15 in the Madison one. But because
there were only four knowledge items in the Kansas City questionnaire and five
in the Wichita, we had less reason to expect that we would successfully be able
to link that variable with news media consumption.
In view of the literature that has suggested that those who consume both
television and newspaper get the advantage of having their attention caught by
television and then getting the details from newspaper (e.g., Neuman et al,
19xx), we thought there would be an interaction between special TV and special
newspaper consumption in that if people consumed both, the impact would be
greater. Thus:
H4: Special TV and special newspaper news consumption will show a
significant interactive effect even after stringent controls for
demographics and main effects of special TV and special newspaper news
consumption are applied.
Finally, because an increasing social structural gap so often occurs in
response to mediated messages targeted to a whole population, we wondered if
media exposure would interact with demographic features such as how long
respondents had lived in the city, their gender, race, income, education. The
social structural gap literature suggested that:
H5: There will be significant indication that the linkage between
media consumption and measures of community involvement depends on
demographic features of individuals. Findings of a widening gap
between
those with higher and lower levels of socioeconomic status are likely.
The Five Cities
The five hypotheses were tested in each of five cities. Although we contend
that the exact nature of the "special news" in each of the cities is not
critical, we briefly describe here the situation in each of the five cities.
Madison
Madison is the capital of Wisconsin, a mid-sized city of approximately 215,000
in the Midwestern home of Progressivism. Madison was recently ranked the
America's most livable city by Money magazine. It has one of the lowest
unemployment rates in the United States (hovering around 2 percent). Its
predominantly white, middle-class population is employed in a wide and stable
variety of industries: state government, education, insurance, and,
increasingly, high tech and bio-tech research. There are relatively few
manufacturing jobs, but agriculture is still a major industry in Dane County.
"We the People" has been a joint initiative from its beginning in late 1991.
Stimulated in part by the Kettering Foundation report Citizens and Politics,
State Journal Editor Frank Denton and Associate Editor Tom Still and Wisconsin
Public Television News and Public Affairs Director Dave Iverson began planning
joint election coverage for 1992 using the idea of deliberative citizens'
forums. The State Journal and WPT teamed up in April 1992 with Wood
Communication, a local public relations firm and important coordinating partner,
to broadcast an upper midwest presidential primary debate using citizen panels.
Citizens sent in newspaper coupons and were selected to form panels
demographically representative of Wisconsin as a whole. Although the Madison
project had involved 16 previous projects in total, the research reported here
looked only at two, one on land use and the other on juvenile crime.
San Francisco
San Francisco is a city of roughly 750,000 in a metropolitan region of two and
a half
million. San Francisco is a major hub for banking in the Pacific Rim, a city
that personified the term "yuppie," but is also stratified by layers of class
and race and marbled by immigrant groups of almost every nationality. It is the
most expensive city in the United States, but still retains large pockets of
middle-class home owners struggling to maintain their foothold.
The San Francisco Chronicle, is the dominant newspaper in the region, with a
daily circulation of about 490,000, Sunday about 650,000. It competes with both
the afternoon Examiner, and to a lesser extent with Bay Area rivals the San Jose
Mercury News and the Oakland Tribune. KRON-TV, an NBC affiliate, was the
television partner.
San Francisco's "Voice of the Voter" project was primarily deliberative,
seeking to inform voters about a series of elections, beginning with the
California Democratic gubernatorial primaries in 1994 and continuing with the
general election that fall. The most recent project preceding the collection of
the data reported here was on the San Francisco mayoral election of October
1995, which took the form of a focus on neighborhoods. While it was formally
about city-wide politics, it targeted and followed politics in specific
neighborhoods, focusing on those neighborhoods' needs in relation to the mayoral
campaign.
Binghamton
Binghamton, a small city of just over 50,000 in upstate New York's southern
tier, can best be described as a community in disarray. (In fact the Binghamton
case represents a 13 communities with a population total of just over 200,000).
The rapid decline of the area's defense and electronics industry in the eighties
and early nineties left a community in shock, trying to redefine its core
economic identity while simultaneously trying to rebuild in every way.
Binghamton's media structure was typical of small cities. The Press & Sun
Bulletin, with a daily circulation of 68,000 and Sunday of 88,600, is the only
major paper in the area. WBNG-TV, the CBS affiliate was the leading station in
the market, and the commercial television partner. WSKG-TV, the PBS affiliate
was a relatively strong partner, but did not produce daily or weekly local news.
The core of the project was eleven action teams that had been organized
directly out of a partnership survey in Fall 1995 by Binghamton University
asking citizens about the most important problems facing the area. The eleven
areas included such topics as business development and venture capital; tax
burdens; government cooperation; job training; transportation; energy; needs of
seniors; community morale and beautification; tourism; youth; and needs of
working parents. The teams were charged with coming up with a specific set of
recommendations that would be presented to the entire community in September,
1996.
Each of the areas was the subject of extensive reporting by the paper. Because
economic collapse had been the major community issue for a number of years,
there had been several cycles of major enterprise reporting on Binghamton's
economic future. This was the first to involve citizen teams in the development
of solutions.
Kansas City, KS and Wichita
These two cities were part of a public health campaign created and funded by
the Kansas Health Foundation. Kansas City has a population of 150,000, and
Wichita's population is 304,000. The topic of the campaign was second hand
smoke's danger to children. There was a considerable media advertising
component to the program, but our focus was on the newspaper and television
coverage of the topic and the campaign. In both cities, the news coverage was
extensive and ongoing. The stories and editorials reported many of the
statistics concerning the damage second hand smoke does to children, how many
people in the cities smoke, and how many children are estimated to be exposed.
In addition, the news coverage talked about the campaign itself. In both cities
the survey reported here was administered approximately two months after the
campaign began.
Method
This study analyzed data from five separate samples previously gathered by a
survey research center at a large midwestern university in 1997. Although the
questionnaires had many similarities, there were several differences in question
wording, which is noted below. Households were sampled by random digit dialing.
The response rates were between 55% and 65% across the five cities.
Dependent variables
The dependent variables analyzed in this study were community participation,
voting behavior, and knowledge about a community problem that received
substantial coverage in local media before the surveys were conducted.
Community participation was an index of the number of organizations in which
respondents reported to be members. Organizations included business, religious,
charity, ethnic, neighborhood associations, PTA, political social and youth
groups. Respondents were given one point for each group they reported belonging
to. Community participation was somewhat consistent across the five surveys.
Voting behavior varied among the five surveys. In three of the cities,
Binghamton, Madison, and San Francisco, respondents were asked if they had voted
in the most recent election. In Kansas City and Wichita, respondents were asked
if they were registered to vote. As shown in Table 1, respondents reported a
high level of both activities, ranging between 80 and 90%.
Knowledge was an index of the number of correct answers respondents gave to
questions about the items obtained from the special news coverage in each
community. The Kansas City and Wichita questionnaires had only 4 and 5
knowledge questions respectively, while Binghamton, Madison, and San Francisco,
had between 9 and 15 questions.
Independent Variables
Demographics: The demographic variables used in this analysis are typical of
studies of this kind, and included gender of respondent, education level, income
level, race, and length of time respondents lived in the community. Gender and
race were coded as two-category nominal variables, such that female and white
were given the value of 1, and male and non-white given the value of 0. The
other three demographic variables were measured as continuous variables and
coded such that a higher value indicated more education, income, or living in
the community longer.
Habitual media consumption: Two measures were employed to indicate habitual
media consumption. In each survey, respondents were asked the number of days
they typically read the local newspaper and the number of days they typically
watched local television news.
Special media consumption: To measure consumption of the special media
projects in each city, respondents were asked two questions, one for TV news and
one for newspapers. For TV news, respondents were asked if they saw news
stories about the specific topic of coverage on the specific television station
that carried them. For example, respondents in Binghamton were asked "Did you
see any stories or programs about 'Facing Our Future' on WBNG-TV Channel 12?" A
similar question was asked for newspapers, "Did you read any stories about
'Facing Our Future' in the Press & Sun-Bulletin?" These items were coded so
that a "yes" received a score of 1 and a "no" received a score of 0.
Results
Descriptive aspects
Community participation levels: As can be seen in Table 2, there appears to be
a fairly consistent distribution of people reporting various participation
levels in the five cities. The percent of people who reported participating in
no community groups varied from a low of 16% in Kansas City to a high of 24% in
San Francisco. Wichita had the second highest percent of non-participants with
23%. Those who reported participating in 1, 2, or 3 organizations varied
between a low of 55% in Wichita to a high of 58% in Binghamton and Madison. The
last quarter of individuals in each city reported participating in four or more
organizations. However, there was a statistical difference in number of
organization respondents reported belonging to and the city in which they live.
Table 2: Community Participation in Organization by City
Number
of
Organizations
City
0
1-3
4+
Binghamton
21
58
21
Kansas City
16
56
28
Madison
18
58
24
San Francisco
24
57
19
Wichita
23
55
22
Chi-square (8) = 27.10, p < .001. Note: Cell entries are percentages.
Voting: It should be noted again that in Kansas City and Wichita the question
was whether people were registered; whereas in the other cities they were asked
if they had voted in the last election. The percent of non-voters in each city
varied from 19% in Wichita to 11% in Kansas City.
Knowledge: The knowledge scales for each city varied so much that comparing
the distributions is probably not informative. From a statistical perspective,
however, it is important to note that there was considerable variance in the
number of questions that people got correct and therefore it is appropriate to
try to parse the variance in these scores.
Exposure to the news media: Table 3 shows the self-reported habitual exposure
to television and newspaper news. Table 4 shows self-reported exposure to the
special news. In Binghamton only 4% (Table 3) of the people habitually consumed
neither of the media, but 58% (Table 4) never heard of the special project.
Table 3: Habitual Media Exposure by City
Neither TV nor Newspaper
TV Only
Newspaper Only
Both
Binghamton
4
11
18
67
Kansas City
3
11
6
79
Madison
9
11
20
59
San Francisco
12
15
29
44
Wichita
5
13
8
74
Chi-square (12) = 248.0, p < .001. Note: Cell entries are percentages.
In Kansas City, only 3% habitually consumed neither of the media, but 53% never
heard of the special project. In Madison, only 9% habitually consumed neither
of the media, but 55% never heard of the special project. In San Francisco,
only 12% habitually consumed neither of the media, but 73% never heard of the
special project. Finally, in Wichita, only 5% habitually consumed neither of
the media, but 54% never heard of the special project. In general, then, about
88-97% of the people had the opportunity to be exposed, but only half or less
reported being exposed to the special news.
Table 4: Special Media Exposure by City
Neither TV nor Newspaper
Only One Medium
Both
Binghamton
58
24
18
Kansas City
53
37
10
Madison
55
32
13
San Francisco
73
16
11
Wichita
54
36
10
Chi-square (8) = 110.0, p < .001. Note: Cell entries are percentages.
Tests of the hypotheses
Hypothesis 1 predicted that there would be a significant association between
habitual and special media consumption and community activity. To test this
hypothesis, a hierarchical multiple regression analysis was conducted for each
of the five communities. Five demographic variablesDeducation, income, race,
gender, and length of time respondents reported living in their communityDwere
entered in the first block. Then, habitual media consumption, both TV news and
newspapers, were entered in the second block. Finally, the special mediaDTV and
newspapersDwere entered in the third block.
Table 5: Demographics, Habitual Media Consumption, and Special Media
Consumption as Predictors of Community Activity for Five Communities
Community
Participation
Binghamton
KC
Madison
SF
Wichita
Demographics
R-sq ch
8.23***
8.55***
13.16***
9.78***
11.94***
TV News
.00
.09
.09
.07
.19**
Newspaper
.08
.18*
.12*
.01
.23**
R-sq ch
2.10**
2.78***
0.99*
0.86*
4.05***
S-tvnews
.099*
.09*
.02
.09*
-.01
S-newspaper
.11*
.12**
.08
.12**
.15***
R-sq ch
2.69***
2.25***
0.72
3.05***
1.93***
Tot R-sq
13.02***
13.58***
14.87***
13.69***
17.92***
Note: *p < 05; **p < .01; ***p < .001. Cell entries are standardized
regression coefficients (beta weights), except where R-square change is noted.
R-square change and Total R-square are in percentages.
This permits the analysis of the R-square change from one block to the next,
while controlling for the variables in the previous equations. The results of
this analysis are shown in Table 5.
For the first block, the R-square values across the five cities for the five
demographic variables are somewhat consistent. Demographics accounted for the
least amount of variance in community participation in Binghamton (8.23%) while
they accounted for the most in Madison (13.16%). For the second block, habitual
exposure to media account for significant amounts of variation in community
participation in each of the five cities. When the special media items were
entered into the equation, additional variance in community participation was
explained in four of the five cities, above that accounted for by habitual media
consumption. This shows a significant and positive relationship, even under
tightly controlled conditions, between seeing special news stories on television
and in the newspapers and community participation. Thus, there is ample
evidence to support Hypothesis 1.
Table 6: Demographics, Habitual Media Consumption, and Special Media
Consumption as Predictors of Vote for Five Communities
Vote
Binghamton
KC
Madison
SF
Wichita
Demographics
R-sq ch
17.62***
3.65**
19.33***
20.80***
7.56***
TV News
.14 *
.16*
.14*
-.11 *
.01
Newspaper
.22***
.11
.16**
.07
.08
R-sq ch
3.73***
2.00**
1.75**
2.00***
2.54***
S TV News
.05
.04
.04
.11*
.01
S Newspaper
.06
.06
.08*
.06
-.02
R-sq ch
0.74
0.54
0.87*
1.90***
0.04
Tot R-sq
22.09***
6.19***
21.95***
24.70***
10.14***
Note: *p < 05; **p < .01; ***p < .001. Cell entries are standardized
regression coefficients (beta weights), except where R-square change is noted.
R-square change and Total R-square are in percentages.
Hypothesis 2 predicted that there would be a significant association between
habitual and special media consumption and voting behavior. The analytic
strategy was the same as reported above for community participation, and is
shown in Table 6. Demographics were significantly associated with vote in all
five cities, but the associations were smallest in Kansas City and Wichita,
which both asked respondents if they were registered to vote. Habitual media
consumption was significantly associated with vote in all five cities. However,
the special media consumption was associated with vote in only two
citiesDMadison and San Francisco. In Madison, seeing the special coverage in
newspapers was associated with vote, while in San Francisco, seeing the special
coverage on TV news was associated with vote. Thus, there is support for
Hypothesis 2.
Hypothesis 3 predicted that there would be a significant association between
the media measures and knowledge about the special coverage. As shown in Table
7, demographics were significantly associated with knowledge in each of the five
cities. The habitual and special media consumption items were associated with
knowledge in three of the five citiesDBinghamton, Madison, and San Francisco.
The lack of variance explained by the media items in the Kansas City and Wichita
samples may have been due to the reduced number of knowledge items in the
questionnaire. Thus, there is support for Hypothesis 3.
Table 7: Demographics, Habitual Media Consumption, and Special Media
Consumption as Predictors of Knowledge for Five Communities
Knowledge
Binghamton
KC
Madison
SF
Wichita
Demographics
R-sq ch
6.75***
3.48**
10.36***
6.56***
3.75**
TV News
.11
-.01
.08
.10
.08
Newspaper
.18**
.01
.13*
.10
.06
R-sq ch
8.43***
0.27
2.95***
2.23***
0.33
S TV News
.23***
.07
.12**
.08
-.01
S Newspaper
.36***
.03
.06
.12**
.06
R-sq ch
20.99***
0.60
1.90**
2.65***
0.30
Tot R-sq
36.17***
4.35**
15.21***
11.44***
4.38**
Note: *p < 05; **p < .01; ***p < .001. Cell entries are standardized
regression coefficients (beta weights), except where R-square change is noted.
R-square change and Total R-square are in percentages.
Hypothesis 4 predicted that the special media consumption would show a
significant interaction even after controlling for demographics, habitual media
use, and main effects of special TV and newspaper consumption. Table 8 shows
the contributions of the interactions between both special news consumption to
each of the three dependent variables. It should be noted that the controls for
demographics, habitual media consumption, and main effects for special TV and
newspaper consumption were not included in Table 8. Those effects appear in
Tables 5 through 7. Counter to Hypothesis 4, only two of the 15 tests were
significant at the p < .05 level, just about the level of chance. In the
Binghamton sample, the interaction term was negatively associated with
knowledge. In the Wichita sample, the interaction term was negatively
associated with community participation. Thus, in the two cases where the
interaction between special TV news and special newspaper consumption were
significant predictors, the impact was negative, not positive, as prior
literature would predict.
To test Hypothesis 5, a series of hierarchical multiple regressions were run.
All possible interactions between each of the five demographic variables and the
two special media consumption items were examined, on each of the three
dependent variables for each city. First, one demographic variable and one
special media consumption variable were entered in a multiple
Table 8: Special TV News and Newspaper Consumption Interactions as Predictors
of Community Participation, Voting, and Knowledge
Community Participation
Binghamton
KC
Madison
SF
Wichita
STV News by SNewspaper
-.06
-.07
.01
.02
-.16**
R sq ch
0.01
0.23
0.00
0.01
1.02**
Vote
STV News by SNewspaper
.01
-.02
-.04
-.13
.01
R sq ch
0.00
0.01
0.01
0.43
0.01
Knowledge
STV News by SNewspaper
-.33***
.00
-.08
-.05
.03
R sq ch
2.72***
0.00
0.24
0.06
0.03
Note: *p < 05; **p < .01; ***p < .001. Cell entries are standardized
regression coefficients (beta weights), except where R-square change is noted.
R-square change and Total R-square are in percentages.
regression equation. Then, an interaction term was created by multiplying those
two variables together. Of these 150 tests (50 per 3 dependent variables), 25
interactions were significant at the
Table 9. Significant Interaction Effects of Specific Media Consumption in Five
Cities on Community Participation
Community
Participation
Binghamton
Kansas City
Madison
San Francisco
Wichita
race (white) by STV News
-.15*
R sq ch
0.68*
race (white) by SNewspapers
-.17*
R sq ch
0.81*
gender by STV News
.33**
.24*
R sq ch
1.17**
0.62*
income by STV News
.41
.33**
R sq ch
1.90***
1.54**
income by SNewspaper
.24*
.48***
.28*
R sq ch
0.61*
2.18*
0.83*
education by STV News
.29**
R sq ch
1.55**
Note: *p < 05; **p < .01; ***p < .001. Cell entries are standardized
regression coefficients (beta weights), except where R-square change is noted.
R-square change and Total R-square are in percentages.
p < .05 level, greater than would be expected by chance. The significant
interactions for community participation are shown in Table 9. Similarly, the
significant interactions for vote are shown in Table 10; and for knowledge in
Table 11.
When a media consumption variable significantly interacts with a demographic
variable while predicting a desirable outcome, it has been taken to mean that
the interaction is an indication of a decreasing or increasing gap between
individuals based on the specific demographic variable, or sociostructural
characteristic. A positive association between the interactions in Tables 9,
10, and 11 and the outcome variable indicates that the specific media helped to
increase the gap and a negative association indicates a decreasing gap. As
such, not all positive associations (indicating an increasing gap) may be
considered detrimental. For example, because gender was coded for female, a
positive R-square change indicates that media were beneficial to that group.
Therefore, because there was a significant positive association between gender
and special TV news consumption such that women who reported seeing the special
coverage in San Francisco learned more than men, it can be said that the media
helped a group that is known to respond less actively to politicsDand thus,
narrows a sociostructural gap. Generally, gap-narrowing media effects are
considered favorable for creating a more equitable society.
Table 10. Significant Interaction Effects of Specific Media Consumption in Five
Cities on Vote
Vote
Binghamton
Kansas City
Madison
San Francisco
Wichita
race (white) by STV News
-.36*
.19*
R sq ch
0.71*
0.69*
gender by STV News
-.25*
R sq ch
0.53*
live by STV News
-.30***
-.11*
R sq ch
1.83***
0.67*
live by SNewspaper
-.16*
R sq ch
0.54*
A glance at Tables 9, 10, and 11 also shows that the media helped to increase
some gaps (most notably income), but helped to reduce some as well (most notably
gender). The media helped to increase the gap due to education in two cases and
decrease it in one. Also, the media helped to decrease the gap due to race in
three cases and increase it in two.
For length of time lived in a particular community, the significant
interactions show that media better served relative newcomers in four of the
cases. In one case, the media better served those whom lived in a community
longer.
Finally, Tables 9, 10, and 11 show that the fewest significant interactions
between specific media consumption and one of the sociostructural variables
occurred in the Binghamton sample (two significant interactions). The Madison
sample had the most significant interactions (seven). The San Francisco sample
had six significant interactions; Kansas City and Wichita had five each.
Table 11. Significant Interaction Effects of Specific Media Consumption in Five
Cities on Knowledge
Knowledge
Binghamton
Kansas City
Madison
San Francisco
Wichita
race (white) by STV News
.61*
R sq ch
0.88*
gender by STV News
.32**
R sq ch
1.10**
gender by SNewspaper
.49***
R sq ch
2.52***
income by STV News
.30*
R sq ch
1.00*
income by SNewspaper
.50***
R sq ch
2.48***
education by STV news
R sq ch
-.21*
0.64*
education by SNewspaper
R sq ch
-.38**
0.95**
live by STV News
R sq ch
.17*
0.68*
-.17*
0.58*
Note: *p < 05; **p < .01; ***p < .001. Cell entries are standardized
regression coefficients (beta weights), except where R-square change is noted.
R-square change and Total R-square are in percentages.
Discussion
The results of this study first of all provide strong support for Stamm's and
others' contention that news consumption is strongly related to community
activity (and here, also knowledge about issues in the news). The contribution
that habitual news consumption made even after highly stringent controls are
removed was impressive. But the present study goes beyond the classic results
to show that when "special news" appears in television and newspapers,
consumption of that material contributes to explained variance in community
activity, voting, and knowledge, even beyond demographic controls and habitual
news consumption as well. This result demonstrates that the material contained
in the "special news," which varied extensively in content in this study, has an
effect that is different or "on top of" the effects of just habitual consumption
of the news. For the producers of this special news, in three cases those
involved in "civic journalism" and in two cases presumably straight-forward
reporting of a major public health information campaign, the news is
encouraging.
An additional finding that is impressive is that the impact of the "special
news" did not always favor those higher on the socioeconomic hierarchy.
Seventy-five percent of the tests of whether exposure to the special news
interacted with individual demographic characteristics were not significant.
This means that the effects of consumption of the special news were affecting
everyone in the sample fairly consistently. On the other hand, there were
important examples of sociostructural gap widenings and narrowings (Table 12).
For community participation, there were four instances of gap narrowings, two
for race and one for gender in San Francisco, and one for gender in Wichita. In
San Francisco, which used a neighborhood-specific approach to talking about a
mayoral race, there was more impact on minorities and women than on whites and
men. In Wichita, where the concern was for children exposed to second-hand
smoke, there was more impact on women than on men. The other gap effects for
community activity were gap-wideners, and they all acted on higher income or
higher education-level people than on lower. This finding is indeed what is
very commonly seen for almost any media event, that is, having more effect on
the more well-to-do and the more highly educated. It is actually surprising
that more of this kind of effect was not observed.
For voting, there was slightly more gap-narrowing than widening. In Binghamton
there was greater vote contribution for non-whites than whites; in Madison on
those who had lived in Dane County a short time than those who had lived there a
long time. In San Francisco there was more impact on those who had lived less
time in the city. The other two significant effects involved gap widening.
For knowledge, there was just about the same number of gap-widenings and
narrowings. Women were more affected than men in Kansas City. Those of lower
education level were more affected than those with higher levels in Kansas City
as well. In Madison, those who had lived in the city less long were more
affected than those who had lived there longer. And in San Francisco, women and
those with lower education were more affected than their opposites.
Table 12: Demographic Gaps Due to Interactions with Special Media Consumption
Gaps
Widen
Narrow
Neither
Education
1
2
27
Income
7
0
23
Race (white)
2
3
25
Gender (female)
1
4
25
Length lived in community
1
4
25
Total
12
13
125
In general, then, it was clear that the "special news" studied here had a
fairly consistent effect on all who consumed it. In various cases, however,
there were interactions with socioeconomic variables. Interactions with
education were just about evenly split between widening and narrowing gaps, as
were the interactions with race. For time lived in the cities, the gap effect
was primarily to narrow it, and the same held true for gender. On the other
hand, all seven interactions with income were in the direction of widening gaps
between the more and less wealthy. Thus we cannot say that there was complete
"success" in these projects, particularly if our criterion was to narrow most
socioeconomic gaps. But at the very least this analysis may demonstrate an
excellent additional way to track the impact of news, and particularly of
specific news that may be designed specifically to aid people by informing and
activating them.
The descriptive aspects of the data are also interesting. There certainly are
variations in how many community activities people engage in across our five
cities, and that fact may indeed relate to "how well cities work." It takes
human hours and dedication to work in the schools, to work with businesses, and
with youth organizations. And the more hours that are dedicated to these
activities, then presumably the better the community looks, runs, and is a
pleasant place to be. Much further work needs to be done to examine community
activity in a variety of areas, cities and townsDand of course whether the media
are related in important ways to that activity is a critical question for those
of us in the news or persuasion business.
The main caveat in this study is one typical of all cross-sectional data.
Although we made careful use of control demographics, the linkages we found
between habitual and special news with community activity, voting and knowledge
are simply correlational. They support the first requirement for causality,
that is the necessary correlations, but because there are not time differences
involved, the second criterion of inferring causality cannot be met. Thus a
next important step is to somehow look in an experimental way at how news
consumption links with an individual's integration into his/her community.
Even within the correlational realm, however, the present results are
encouraging for those who believe that the news can be structured so that it
does make a difference. In each of our five cities, sometimes to a greater
extent of course than other times, the special news contributed to the measures
we considered importantDbeing active, voting, and knowing something about what
was going on. What we must strive for are more opportunities not only to create
news that makes a difference, but opportunities to evaluate the extent and type
of that difference.
References
Burnstein, Paul (1972). Social structure and individual political
participation in five countries. American Journal of Sociology, 77(6),
1087-1110.
Chaffee, Steven H., McDevitt, Michael, & Thorson, Esther (1997). Citizen
response to civic journalism: Four case studies. Paper presented at the
Association for Education in Journalism, Chicago, August.
Friedland, Lewis A., Sotirovic, Mira, and Daily, Katie (In press). Public
journalism and social capital: The case of Madison, Wisconsin. In
(Eds.) E. Lambeth, P. Meyer, & E. Thorson, Assessing Public Journalism,
University of Missouri Press.
Janowitz, Morris (1952). The community press in an urban setting.
Glencoe, IL: Free Press.
Martinelli, Kathleen A., & Chaffee, Steven H. (1996). Measuring new-voter
learning via three channels of political information. Journalism and
Mass Communication Quarterly, 72(1), 18-32.
Merritt, Davis (1995). Public journalism--Defining a democratic art.
Media Studies Journal, 19(3), 125-132.
Putnam, Robert D. (1995a). Tuning in, tuning out: The strange
disappearance of social capital in America. PS: Political Science &
Politics, December, 664-683.
Putnam, Robert D. (1995b). Bowling alone: America's declining social
capital. Journal of Democracy, 6(1), 65-78.
Rosen, Jay (1995). Getting the connections right: Public journalism and
the troubles in the press. New York: Twentieth Century Fund.
Rosenston, S. J., & Hansen, J.M. (1993). Mobilization, participation and
democracy in America. New York: MacMillan Publishing Company.
Stamm, Keith, & Fortini-Campbell, Lisa (1983). The relationship of
community ties to newspaper subscribing and use. Journalism Monographs,
No. 84. Columbia, SC: AEJMC.
Stamm, Keith R., Emig, Arthur G., & Hesse, Michael B. (1997). The
contribution of local media to community involvement. Journalism and Mass
Communication Quarterly, 74(1), 97-107.
Teixeira, R. A. (1987). Why Americans don't vote: Turnout decline in the
United States, 1960-1984. New York: Greenwood Press.
Tichenor, Phillip J., Donohue, George A., & Olien, Clarice N. (1980).
Community conflict & the press. Beverly Hills: Sage.
Toqueville, Alexis de (1969). Democracy in America, translated by George
Lawrence. Garden City, NU: Anchor Books.
Verba, Sidney, Schlozman, Kay, & Brady, Henry E. (1995). Voice and
equality: Civic volunteerism in American politics. Cambridge MA:
Harvard University Press.
|
