|
Talk Radio as Forum and Companion: Listener attitudes
and uses and gratifications in Austin, Texas
John Beatty
Assistant Professor
Department of Mass Communications
University of North Carolina at Pembroke
Old Main 232
PO Box 1510, Pembroke, NC 28372-1510
(910) 521-6598
[log in to unmask]
Paper submitted to the Mass Communication and Society Division
of the Association for Education in Journalism and Mass Communication
annual conference, Baltimore, Md., August, 1998.
Talk radio as forum and companion
Talk Radio as Forum and Companion: Listener attitudes and uses
and gratifications in Austin, Texas
There are those who say talk radio is dangerous, and I am probably
the sort of listener they have in mind. I am addicted. I only consume
the
good stuff, mind you, and there isn't that much of it around; but
when I find
it my days are rearranged, my friendships, work and study put aside,
and for
weeks I binge on tirade and debate (Margolis, 1990; pp. 6-7).
I have recently talked with lots of talk radio buffs, and it seems
to me that people are getting tired of listening to angry middle-aged
white
guys ranting about the latest headlines and shouting down similarly
disgruntled callers (Herndon, 1994; p. Onward 17).
America is awash in talk. Loud talk. Angry talk. Conspirational
talk. Raunchy talk, smug talk, self-serving talk, funny talk,
rumor-mongering
talk. A cacophony of chat fills the airwaves from coast to coast,
from dawn
to dusk and beyond, all talk all the time (Kurtz, 1996; p.3).
Love it or hate it, talk radio came of age in the 1990s Dand some say it may
already be on the decline (Paige et al., 1998) as analysts rushed to understand
the phenomenon. The number of news/talk stations rose from 308 in 1989 to more
than 1,330 at the end of 1997 (Paige et al., 1998). It has been the subject of
recent books by a critical theorist (Munson, 1993), a television critic (Kurtz,
1996) and at least two talk show hosts (Scott, 1996; Laufer, 1995). It has been
analyzed in several national surveys (Kohut et al., 1993; Benchmark Company,
1994; Adams Research, 1995; Cappella et al., 1996) and in numerous research
papers.
Cappella et al. (1996) found that 18 percent of the adult population reports
listening to at least one call-in political talk radio show at least twice a
week. In Austin, Texas, two news talk stations have a combined rating of 8.1
percent in the fall 1996 Arbitron survey (Herndon, 1996), with several other
popular talk programs on music stations, often in morning drive time.
Much of the analysis of talk radio points out the contradictory nature of the
medium. It is reviled for its negativity, even described as an unhealthy
addiction, yet it also is seen as a powerful political force only recently
unleashed. It has been credited with stopping a mass transit tax proposal in
Denver (Green, 1997), and, at least temporarily, a proposed Congressional pay
raise, and cited as a factor in: Zoe Baird (Page and Tannenbaum, 1996) and Kimba
Wood's unsuccessful nominations (Kurtz, 1996); the Exxon boycott; the backlash
against gays in the military; the cellular phone/cancer controversy; Clinton's
health care plan; and the "three strikes and you're out" crime bill provision
which was co-authored by a talk-show host (Wright, 1995). Rush Limbaugh rose to
become an important figure in Republican politics and in 1993 President Clinton
held a briefing on health care for 200 talk show hosts at the White House
(Laufer, 1995).
Can all this be true of a medium that, for all the urgent attention afforded
it, has been around for about 70 years and may be no more than an electronic
version of the town crier, the newsboy (Burd, 1988), the lyceum movement and
women's magazines (Munson, 1993)?
Those who react with alarm are mostly authors and critics who are reacting to
reports of on-air interactions and self-promotional claims of talk show
programs, rather than examining the shows themselves, or audience members'
interpretations. This paper uses the audience in an attempt to evaluate the role
of talk radio in society. It is motivated by the suspicion that much of what is
claimed about talk radio is overblown, overly emotional and overly selective.
Review of the Literature
Some histories of talk radio isolate a psycho-social predisposition said to
derive from the alienating, monolithic conditions of modernity D the talk show
as therapist for the lonely, and as creator of social networks (Bronstein,
1983), although Bick (1987) denies that it is anything more than another form of
mass communication, claiming that "the very existence of the radio transmitter
belies interpreting talk radio formats as interpersonal communication" (p. 109).
Talk radio provides a format that argues strongly for the falseness of the
mass/interpersonal dichotomy simply on the face of its simultaneous combination
of mass media with a one-on-one conversation, to say nothing of the
interpersonal uses that audiences make of the format, including the possibility
of parasocial interactions (Horton and Wohl, 1956).
Clearly, there are elements of interpersonal communication in talk radio
programming and listening. The parasocial and interpersonal functions of talk
radio are labeled "Companion" based on the paper "Talk Radio D Forum and
Companion" (Tramer and Jeffres, 1983). Dependency theory (Ball-Rokeach and
DeFleur, 1976) could address the social end of this factor, and parasocial
interactions (Horton and Wohl, 1956) the more psychological. Wilkinson and Grant
(1995) suggest that "the Gulf War attracted many people to tune in (as
dependency theory would predict)" (p.17) and then the behavior became a habit
for many.
Talk may also have important ties to traditional communicatory roles such
as public oral communication, for example town criers, (Burd, 1988) which a
postmodern perspective would suggest are returning to vogue.
Avery and Ellis (1979) provide evidence for the hypothesis that talk radio
serves as an interpersonal surrogate, mentioning studies by Turow (1974) and
Bierig (later published as Bierig and Dimmick, 1979). These two papers and one
by Armstrong and Rubin (1989) have titles of the sort "Talk Radio as
Interpersonal Communication." They cite each other, back to Turow, but seem to
suffer from poor operationalizations of loneliness in the earlier works. Bierig
and Dimmick (1979), for example simply asked respondents if they lived alone.
A second line of thought emphasizes the political role of the talk show, a
form of social discourse also tied to the degeneration of premodern "back fence"
or commons interactions (Levin, 1987; Rehm, 1993; Kohut et al., 1993; Craft
cited in Paige et al., 1998). This factor will be labeled "Forum." This view
also attaches the growth of talk radio to disenchantment with traditional
electoral politics, a view apparently seized upon by politicians in the 1992
elections (Waluszko, 1995; Hollander, 1995).
Some support for the political forum hypothesis was gained in an early
study of a half-hour overtly political call-in show (Crittenden, 1971), when the
public was arguably more political, and in analysis of recorded talk shows from
1977 and 1982 (Levin, 1987), although Avery and Ellis (1979) claim that
Crittenden's results derived from the brief, targeted nature of the program in
question. Surlin and Soderlund (1993) found a mix of interpersonal advice-type
talk and political-community discussion in Jamaica, although they did not
analyze callers' or listeners' perceived uses of the programs.
A third factor in the attention to talk radio may be as entertainment
beyond the social- psychological and political. Uses and gratifications studies
have found a link between entertainment and parasocial gratifications (Dobos and
Dimmick, 1988), which suggests that companion-seekers will be more likely to
report entertainment as a gratification. Ultimately, entertainment is assumed to
be one of the reasons for listening, something that is taken for granted D
otherwise there would be little or no listening, by definition.
One of the key questions examined in the study remains, then, the extent to
which talk radio acts as a surrogate for interpersonal relations as opposed to a
forum for the politically interested and, perhaps, involved. In the years since
Tramer and Jeffres (1983) posed and popularized the terms, few surveys have
addressed the relationship between the two concepts. The Times-Mirror survey
(Kohut et al., 1993) Hofstetter et al., (1994) and Hollander (1996) looked at
the political, Rancer et al. (1994) examined the psychological, and Benchmark
(1993) looked at simple audience preferences and demographics. Zerbinos (1993)
came closest to re-iterating the two concerns ("Talk Radio: Motivation or
Titillation?") although she emphasized the political functions.
Forum/Companion: An organizing framework
The Tramer and Jeffres (1983) paper makes a good jumping-off point, but is
severely limited in its scope and operationalizations, and over-zealous in its
conclusions. Much of that is due to the fact that it was an intercept survey of
callers with a single question ("Are you calling to give/receive information or
did you want to chat?") serving as a filter for the three categories of callers
(seeking information, companionship, or a forum).
Callers constitute a small segment of the audience (typically 2 to 5
percent) and there is research to suggest (Kohut et al., 1993; Hofstetter et
al., 1994; Rancer et al., 1994; Hollander, 1996) that they are different from
the rest of the audience (more politically aware, active, less likely to seek
companionship). There is little basis on which to conclude, as Tramer and
Jeffres did, that
Just as Herzog found radio soap operas functional outlets for
escape some 40 years ago, today's talk radio formats are providing a
similar outlet for today's isolated listeners
(Tramer and Jeffres, 1983; p. 300; emphasis added).
Tramer and Jeffres do not even have much of a basis for making such a
conclusion about the callers, as 44 were said to seek a forum, 49 companionship,
and 61 information. The difference between the forum and companion motivations
is not significant, and the 61 information-seekers are ignored in this
conclusion. It is also too simplistic to code callers, based on a single filter
question, into mutually exclusive categories.
hypotheses
Relationships among the variables can be summarized as in the following
diagram (Figure 1). Some elements of the figure are discussed elsewhere (Beatty,
1996). Single-headed arrows are drawn to indicate probable causal pathways (e.g.
higher income "causes" forum usage), double-headed to indicate correlates.
Parasociability (Auter and Palmgreen, 1993) is a personality trait said to
predict parasocial interaction with the media.
[--- ??? Graphic Goes Here ---]
Figure 1. Relationships of study variables.
Comments in this study's pre-test raised the issue that the site of
listening was related to the gratifications sought. In particular, respondents
volunteered that they listen to talk radio for companionship primarily in the
car.
H1: Those whose primary listening site is the car will be more
likely to agree with companionship-seeking as a reason for listening to
talk radio than will those whose primary listening site is at home or
work.
Forum-seekers would be lower on parasocial involvement with a favorite
host, with talk show hosts in general (what Auter and Palmgreen (1993) call
"parasociability"), and lower on loneliness (from the UCLA loneliness scale,
Russell et al. (1980), not merely "living alone" as in Bierig and Dimmick
(1979)). Since loneliness and parasociability are personality traits, they would
predict companion-seeking (but not forum-seeking), whereas parasocial
interaction is an aspect of radio listening that is coincident with the
obtaining of gratifications.
Counter to Tramer and Jeffres' (1983) claim, Zerbinos (1993) found
companion-seeking to be the least reported gratification among six offered to
respondents, although she adds "another way of asking this question might
produce more truthful results" (p. 15). Items on parasociability and loneliness
can complement the Likert item "I listen to talk radio for companionship" to
examine both the relationship between listening and the two personality traits,
and the relationship between the Likert item and the traits.
H2: The more listeners to talk radio are lonely, the more they will
seek companionship gratifications.
H3: Parasocial interaction will be positively correlated with
companion-seeking.
H4: The more listeners to talk radio possess parasociability, the
more they will seek companionship gratifications.
Surlin's research (Surlin, 1986; Surlin and Soderlund, 1993) suggests that
ethnicity and culture help determine the nature of uses of talk radio, as
Albarran and Umphrey (1994) found for television viewers. We would expect
members of a minority culture who are new to the community to use the medium
more for information and orientation to community issues, and less to satisfy a
need for interpersonal communication. Surlin (1986) found that lower SES
listeners in Jamaica sought primarily information and surveillance (forum-type)
gratifications rather than interpersonal or companionship gratifications. Those
newer to community are also more likely to use the format for surveillance
(forum-type) purposes in that they are more marginal.
H5: Non-white listeners and those new to the Austin area are more
likely to seek forum gratifications than are whites and those residing
longer in the area.
The tendency to use talk radio for interpersonal gratification is
conceptually related to affect toward the host. "Surrogate companionship" could
derive from the experience of hearing a conversation without close listening, as
well as from the hosts who,
come into the home as companions who are willing to listen to
personal problems, share ideas about social and political issues, remind
listeners that they are not alone, and provide assurance that their
beliefs and attitudes are shared by others (Avery and Ellis, 1979; p.
114; emphasis added).
In other words, companion-seekers are likely to come to like hosts who
fulfill their needs, and through positive feedback (listening most to those they
like) develop a favorite host. Forum-seekers would not be expected to name a
favorite host.
H6: The more talk radio listeners seek companionship, the more
likely they are to name a favorite host.
Zerbinos (1993) asked listeners how important it was "To hear points of
view expressed that are similar to mine" (p. 8). Listeners were divided although
this was the only fifth-ranked of six reasons for listening. Kohut et al. (1993)
found that most listeners preferred hearing opposite points of view:
RQ1: Are forum- and companion-seeking related to a preference for
hearing an opposite or own point of view?
On the other hand, Kohut et al. (1993) found several indicators of distrust
and disaffection with government (lower favorability for President Clinton, the
U.S. Congress) to be associated with talk radio listening, as did Cappella et
al. (1996). Listeners are less likely to approve of the efforts of local
politicians (assuming they are less media-dependent). Forum-seekers might be
even less approving but they could just as easily support the government:
RQ2 : What is the relationship between forum-seeking and opinion
about the job the local government is doing?
Given the claim that the format has become overly mean-spirited, it is
likely that the forum-seekers who are more interested in issues and debate would
listen more than would companion-seekers. Seeking companionship implies that the
format is perceived as friendly, rather than mean-spirited. Put another way, a
high degree of companion-seeking would be evidence that the critics are
over-rating the hatefulness or mean-spiritedness of the medium.
H7: The more listeners to talk radio seek companionship the less
likely they will be to perceive talk radio programs as mean-spirited.
Talk radio's role as a political forum is somewhat antithetical to the idea
that talk radio is mean-spirited, although there are likely to be instances of
both types of content. Theorists (e.g. Rehm, 1993) and media critics (Austin
American-Statesman, 1992; Burns, 1994) have trumpeted the positive political
forum, or "town meeting" role in the face of the more extensive criticism of the
programs' negative tone. Among listeners, those who come to the format seeking a
political forum are more likely to see it as fulfilling a town meeting function.
H8: The more listeners to talk radio seek a political forum, the
more likely they are to perceive talk radio programs as town meetings of
the air.
Regular listening to local talk radio reveals frequent discussion of
politics. Through repetition of names and issues by the hosts and callers,
listeners are likely to become more familiar with political information. This
familiarity would seem likely to be correlated with forum-seeking but not with
companionship-seeking:
H9: The more listeners to talk radio seek political forum
gratifications, the more likely they will be to be able to identify local
political figures and abbreviations.
Zerbinos (1993) found that talk radio listeners were slightly more likely
than non-listeners to be aware of current political issues, and to be members of
community associations. Cappella et al. (1996) found more participation reported
by listeners, and Hofstetter et al. (1994) also found talk radio listeners
higher in political participation, calling this "the most surprising and
significant finding" of their study (p. 476). Extending those results to the
forum/companion uses, it would seem that forum-seeking is positively correlated
with communicating with elected officials, going to political meetings, and with
membership in associations:
H10: The more listeners to talk radio seek political forum
gratifications, the more likely they will be to have participated in
political activities.
Summary research questions: The personal characteristics and media use
variables in Figure 1 can be used to test for the ability to predict the
outcomes using regression models. The point here is to recognize that the
hypotheses developed above are easily testable and that multivariate models
reduce to research questions of the form: How do personal characteristics and
media use predict political knowledge (attitudes toward talk radio content,
attitudes toward political institutions, political participation)? Answering
those research questions allows another look at the role forum- and
companionship-seeking play in the outcomes of the model.
Methods
Contest entry blanks were obtained from KVET/KASE radio in Austin. KASE is
a country station, KVET-AM is news/talk and KVET-FM carried a popular morning
talk show ("Sammy and Bob").A set of local names and telephone numbers was
generated from the entry blanks. The intent was to obtain with limited resources
a rich sample of radio listeners many of whom would be familiar with call-in
talk programs broadcast on KVET AM and FM.
Interviews were completed with 174 talk radio listeners in May and June of
1996.The upper-bound response rate was 80 percent. Respondents were asked seven
uses and gratifications items, five derived from the Times-Mirror study (Kohut
et al., 1993) and two from Zerbinos (1993). Four knowledge items tested local
political knowledge. One item asked about attitude toward local politics ("In
general, how good a job do you think the city council is doing?"). Five items
from Kohut et al., (1993) asked about political involvement (whether they had
called or sent a letter to the White House, to a Congressman, participated in an
opinion poll, attended a civic meeting or joined a community group).
Parasocial interaction items (Auter and Palmgreen, 1993) asked respondents
about hypothesized interactions with a favorite host. Some items were adapted
from Auter and Palmgreen's (1993) Identification With Favorite Characters
subscale for soap operas; others are from Rubin and Perse (1987) adapted for
radio, excluding those specific to television such as "My favorite host seems
attractive."
Parasociability items asked about "any talk show host that you enjoy
listening to" and provide a measure of a personality trait, the tendency to
interact parasocially. Four loneliness items are a subscale from a longer
psychological instrument developed at UCLA (Russell et al., 1980) that improves
on indirect measures that merely asked about the number of people in the
household (Armstrong and Rubin, 1989).
Radio attitude questions consisted of one asking whether respondents liked
to hear their own or opposite point of view and two Likert items adapted from
media coverage ("Call-in talk radio shows are town meetings of the air": "Talk
radio call-in shows are mean-spirited."). Follow-up interviews were conducted
with a sample of the 40 percent who agreed to them.
results: Factor analysis, scaling
Table 1: Factor analysis of seven reasons for listening to talk radio.
Reason
Factor 1
Factor 2
Keep up on issues of day
.58
.40
Companionship
.30
-.14
To learn things can't find elsewhere
.76
.00
To learn how people feel about issues
.69
.17
Like the host
.06
.94
Know what's going on in community
.72
.02
Pick up information to use in conversation
.60
.07
Table 1 does show part of the expected pattern (seen in the pre-test) in
which what could be called the "non-interpersonal" reasons for listening fall
strongly on one factor with liking the host on a second. Liking the host loads
highly on the second factor and has close to a zero loading on the first, but
companionship loads negatively on Factor 2 and moderately on Factor 1, counter
to the hypothesis. It is interesting that a reason cited as an interpersonal
utility by Dobos and Dimmick (1988) ("To pick up information that I use in
conversation with other people") loads close to zero on the second "host"
factor. This suggests that the interpersonal nature of talk radio lies mainly in
liking a host (but not in attaining some companionship from the host), and that
the interpersonal communication utility functions with the knowledge and
surveillance gratifications (Factor 1). The nature of this factor might be
described as community information-seeking D finding out the issues, getting
knowledge and being able to use that in the community.
There does not seem to be a strong association between the two
"interpersonal" reasons for listening (host, companionship).
To further assess the inter-item relationship, a reliability analysis of
the two items was performed. This produced a very low Cronbach's alpha of .11.
On the other hand, the five community information items (loading strongest on
Factor 1) produced a Cronbach's alpha of .72, sufficient to proceed with a
single scale constructed from the combined scores on the five items. Further
analysis proceeded with examination of hypotheses that predicted relationships
with a "forum" variable employing the community information construct. Those
hypotheses that predicted relationships with a "companion" variable were
examined against the scores on the host item and the companion item
independently.
Parasocial interaction, parasociability: Respondents who reported a
favorite talk show host were given seven Likert-type parasocial interaction
items adapted from television research (Rubin and Perse, 1987; Auter and
Palmgreen, 1993).
Reliability analysis produced a Cronbach's alpha of .58. This is low for a
seven-item scale, but shows that the items almost act together well enough to
form a single scale. The lowest correlations are between items from the two
studies. The item: "If he/she appeared on another radio station, I would listen
to that program," correlated at .08 or below with Auter and Palmgreen's three
identification items (and at .09 with another Rubin and Perse item, "I would
like to meet him/her in person").
The four Likert-type parasociability items from Auter and Palmgreen (1993)
were analyzed for reliability and produced a Cronbach's alpha of .77. All
inter-item correlations were at .40 or above. These data are high enough for a
four-item scale and so a parasociability measure was created by summing the
scores to the items (possible range from 4 to 20, mean = 14.30, s.d. = 2.64).
Political knowledge, participation: Three items measuring local political
knowledge were combined into a single political knowledge score. This measure
was considered to be a simple sum (score) as opposed to a scale requiring a
check of inter-item reliability. Political activity, or participation, was
treated in a similar manner.
Listening patterns
Those who reported listening to talk radio (N=174) devoted an average of 37
percent of their radio listening time to talk. Almost half of the listeners
devote less than 20 percent of their radio listening to talk radio, mostly 10 to
20 percent. Another fifth listen about half the time to talk and one-seventh
might be characterized as talk radio "junkies" (as Margolis, 1990, described
himself) all of whom devote 90 percent or more of their radio listening to talk.
The distribution is not continuous and seems to define three distinct listening
patterns.
Time spent with talk per day = (average radio minutes per day)(percent of
radio time devoted to talk). The mean for 168 listeners who responded to both
questions was 62.9 minutes (s.d=72.2) per day.
A surprising percentage of talk radio listening D 70 percent D was reported
as being done in the car, followed by at (17%) and at home (13%).
Many people listen to talk radio in their cars while they are working
(traveling salespeople, for example), which may cloud interpretation of the
primary listening site question.
Listener attitudes toward talk radio content
Three items measured listeners' attitudes toward talk radio content. The
first was taken from Kohut et al. (1993). This study found 17 percent preferring
their own point of view versus 59 percent preferring an opposite opinion,
reflecting the pattern found in Kohut et al. (1993) and Zerbinos (1993).
Apparently, the confrontational nature of talk as portrayed in much of the press
(Herndon, 1994; Cappella et al., 1996) is an attractive rather than a repulsive
quality.
Listeners also do not agree that talk radio is mean-spirited despite the
fact that the phrases of that type have been used in reference to the popular
"Sammy and Bob Show." Only 20.1 percent agree that talk radio is mean-spirited,
whereas 60.3 percent disagree This finding dovetails with the preference for an
opposite point of view to provide a picture of listeners who enjoy lively
discussion and are perhaps aware of the hosts' ironic or sarcastic attempts at
stirring up the pot for the sake of a good show. It suggests that the average
talk radio listener is a more discriminating or even oppositional audience
member in Stuart Hall's (1980) sense than is the average newspaper critic with
regards to talk radio. This conclusion was borne out by follow-up interviews.
The response to the town meeting statement (taken directly from headlines
to an editorial (Austin American-Statesman, 1992) and an op-ed column (Burns,
1994)) indirectly supports this interpretation of the audience's reaction of
talk radio content. The proportion agreeing that talk radio shows are town
meetings (43.7 percent) was about equal to the proportion disagreeing (38.5
percent). Listeners did not agree with the statement overall, but they did not
disagree, which implies that they are not overly taken with the discussions'
importance, even though the other attitude items show that they enjoy hearing
opposing viewpoints and don't find them mean or upsetting.
In other words, they may take an ironic stance which recognizes that much
of what goes on is fabricated for the sake of lively interactions .
Uses and gratifications
Table 2 gives the percentages of talk radio listeners in each response
category for the seven uses and gratifications as well as comparisons to the
other two talk radio studies. It is notable that of 1,218 questions asked there
were only three "Don't know" responses, suggesting that these listeners have
already formed some opinions about why they listen.
Items are listed in order of the mean score for each Likert item for which
5= "Strongly agree." Because of the differentiations among gratifications
sought, gratifications gained (Dobos and Dimmick, 1988) and uses are unaddressed
in this study, the less specific term "reasons" (as in Zerbinos, 1993) will be
used in the remainder of this paper.
Table 2: Percentages of talk radio listeners by reasons for listening to talk
radio with comparisons to Times-Mirror and Zerbinos.
Reasons for listening
SD
D
N
A
SA
Mean
To learn how people feel about different issues1
2.3
7.5
7.5
62.1
20.7
3.91
Because I like the host of the show6
4.0
13.2
17.2
46.0
18.4
3.62
To know what's going on in my communityii
1.7
17.2
17.2
52.3
11.5
3.57
To keep up on issues of the day2
2.9
18.4
17.8
51.7
9.2
3.47
To pick up information that I can use in conversation with other people5
1.7
23.6
22.4
44.8
7.5
3.35
Because it is a good way to learn things I can't find out elsewhere3
4.6
26.4
12.1
43.7
12.6
3.35
For companionshipvi
16.7
52.3
17.8
11.5
1.7
2.29
Times-Mirror (Kohut et al., 1993) rankings indicated by numerical
superscripts. Their fourth-ranked reason was "Because it is entertaining."
Zerbinos (1993) rankings indicated by Roman numeral superscripts. *
Despite some comparability problems, patterns emerge. Companionship was the
lowest-rated of six reasons for Zerbinos (1993) and the lowest of the seven
reasons in this study, with 13.2 percent agreeing that it was a reason they
listened. This is further evidence that the early claims of a lonely, isolated
audience (Avery and Ellis, 1979; Bierig and Dimmick, 1979; Tramer and Jeffres,
1983; Armstrong and Rubin, 1989) no longer apply, if they ever did.
This study and Kohut et al. (1993) found a surveillance reason to be the
top-ranked reason for listening, and another surveillance reason ranked high in
this study (third) and in Zerbinos (second). Knowledge-based reasons were also
important but less so (fourth and tied for fifth in this study, second and third
in Kohut et al., 1993) and communication utility was still less important (tied
for fifth in this study and fifth in Kohut et al., 1993).
The difference in rankings between liking the host (second) and seeking
companionship (seventh) suggest that the two items are not acting together or
are not part of a single psychological construct for listeners. In other words,
there are many listeners who listen because they like the host of a program or
programs, but they do not seek or attain companionship from the host(s). This
was born out in the factor analysis and associated reliability check of the
factors.
Uses and gratifications and relations to other variables
The three reasons derived from the factor analysis reported previously D
community information, host and companionship D were tested in separate one-way
ANOVAs against the primary listening site.
The use of talk radio for companionship did vary depending on primary
listening location with an (F = 5.06; p=.007). This seems to confirm the
expectation that companionship was the use most likely to be tied to site of
listening, although multiple range tests showed the pattern to be different than
expected. The companionship use for those listening primarily at home was
greatest (mean=2.86; N=22) and differed significantly from the companionship
ratings of "car listeners" (mean=2.22; N=121) and "work listeners" (mean=2.13;
N=30) at p= .05. Companionship ratings did not differ significantly between work
and car listeners. Thus Hypothesis 1 is not supported in that
companionship-seeking is not associated primarily with listening in the car. It
seems that respondents found it to be associated more with "the comforts of
home."
The use of talk radio for community information or because the host was
liked did not vary greatly across the three listening sites, and no two groups
differed significantly at the .05 level according to multiple range tests.
Respondents' scores on the three reasons for listening were compared with
their scores on the loneliness scale, recalling the relatively low alpha score
obtained for the scale (.40). Companionship had an almost-zero correlation with
loneliness (r=.04; p=.61), contrary to Hypothesis 2, while liking the host had a
slight negative correlation (r=-.04; p=.60, N=167). Surprisingly, the community
information scale score had a statistically significant correlation with the
loneliness score (r=.20; p=.01, N=168). The community information scale does
have elements of interpersonal communication (e.g. getting information to use in
conversation with others) that could explain a correlation with loneliness.
The lack of a positive relation between liking the host and loneliness is
not surprising in the light of results reported earlier. Liking the host and
having a favorite are not strongly related to companionship nor is liking the
host strongly related to loneliness. Those listeners who express positive affect
toward hosts may do so because they find the host(s) enjoyable or entertaining,
perhaps because they too are outgoing, communicative people. It is clear that
the opposite conclusion is not supported by this study. Talk radio listeners in
general, and even those who relate positively to talk radio hosts, are not
listening to compensate for a lonely life.
The positive correlation between seeking information about the community
and loneliness is unexpected, but may be evidence that finding out the issues,
how others feel and what's going on, and using it in conversation are more
attractive to those who are lonely because they can address their loneliness
through feeling connected to the community, more than through feelings for the
host. Those who seek such community information also show some affect toward the
host (r=.14; p=.07) and tend to seek companionship (r=.13; p=.10). It may be
safer to say that there is not a strong relationship between loneliness and the
reasons people listen to talk radio, but loneliness is tied more to community
information items than it is to affect for the host.
The three reasons for listening were compared to parasocial interaction and
parasociability. Parasocial interaction (a media use behavior) and
parasociability (a psychological tendency) correlated at r= .56 (p=.001; N=61)
which suggests that the operationalization of parasociability is of some use in
predicting parasocial interaction. Auter and Palmgreen (1993), who devised the
parasociability items, found that parasociability and parasocial interaction
correlated at r= .33.
As with loneliness, companionship-seeking had almost no relationship to
parasocial interaction (r=-0.07, p=.56, N=63; Hypothesis 3 not supported), an
element of media exposure, nor to parasociability (r=0.02, p=.78, N=165;
Hypothesis 4 not supported), a personality trait. In other words,
companionship-seeking is not related to parasocial interaction , which is an
interaction with the host, and the tendency to interact parasocially with the
host (parasociability) does not predict companionship-seeking.
In contrast to the null results for companionship-seeking, listening
because of liking the host correlated with parasocial interaction (r=0.26,
p=.04, N=63) and was predicted by parasociability (r=0.29, p=.000, N=163),
lending some reliability to these three host-related measures. Unexpected
interactions between community information-seeking and parasocial interaction
(r=0.31, p=.01, N=63) and parasociability (r=0.33, p=.000, N=165) suggest that
community information-seekers have some affective relationship with the host,
despite the distinct factor loadings reported above. There seems to be an
interpersonal character to community information-seeking in that it is related
to these two measures and to loneliness.
The study found talk radio listeners more likely than non-listeners to be
white and they also are likely to have been in Austin for a shorter period than
non-listeners. Those newer to the community are more likely to seek out talk
radio for whatever reasons, although language barriers may keep the percentage
of non-white listeners down.
White and non-white listeners were compared in terms of their mean scores
on companionship, liking the host and community information. Whites and
non-whites do not differ in the uses they make of talk radio in Austin, in
contrast to the findings of Surlin (1986) and Surlin and Soderlund (1993).
Hypothesis 5 is not supported, perhaps because of the influence of the "country
music" sample used. Taken together, race and length of residence did not appear
to explain differences in reasons for listening to talk radio, contrary to what
is stated in Hypothesis 5.
Uses and gratifications and outcome variables:
Outcome variables (see Figure 1) encompass: attitudes toward talk radio
content, attitudes toward political institutions, political knowledge and
political participation.
Mean scores for the three reasons for listening to talk radio were compared
for those naming a favorite host and those who did not.
Those naming a favorite host scored higher on the reason "Because I like
the host of the show" (mean difference = .51, p=.002) as hypothesized using the
original "companion" concept (Hypothesis 6). The two items acted almost as a
reliability check on whether or not respondents liked a host (had a favorite).
On the other hand, there was almost no difference between the two types of
listeners (favorite host or none) on the degree to which they listened for
companionship (mean difference = .05, p=.76; contrary to Hypothesis 6). This is
further evidence that positive affect toward hosts (in this case, having a
favorite) is not related to using talk radio for companionship.
The three reasons for listening to talk radio were analyzed with respect to
preferring to hear the respondent's own point of view or an opposite one. This
outcome variable and responses to the "mean-spirited" and "town meeting" items
are considered measures of attitudes toward talk radio content.
Research question 1 asked about the relationship between preferred point of
view and reasons for listening. There does not seem to be an important
relationship between the point of view preferred and the uses made of talk
radio.
Of the three reasons for listening only companionship-seeking predicted a
(positive) attitude toward city council (r=.16; p=.03; N=168; Research question
2).
Hypothesis 7 stated that forum-seeking would predict disagreement with the
statement, "Talk radio shows are mean-spirited," and Hypothesis 8 predicted that
forum-seeking would predict agreement with the statement, "Talk radio shows are
town meetings of the air." The community information (forum) scale correlated at
r = -.24 (p=.001; N = 169) with the "mean-spirited" Likert item and at r = .35
(p=.001; N = 169) with the "town meeting" item, supporting both hypotheses.
Those seeking community information may be finding what they seek D a
reasoned political forum D or these results may suggest that those paying close
attention to the discussion of political issues recognize that there is passion
involved in debate and welcome a "good argument." Rancer et al. (1994) found
that talk radio listeners were more likely to be "argument prone" and less
likely to avoid arguments than non-listeners. This study supports that
characterization of talk radio listeners, in particular if they come to the
format seeking community information.
None of the correlations between reasons for listening and political
knowledge (Hypothesis 9) and between reasons for listening and political
activities (Hypothesis 10) attained statistical significance at p<.05; in fact,
the hypothesized relationship between community information-seeking and
political activities approached significance but in the opposite direction
specified by Hypothesis 10 (the more listeners seek community information the
less likely they are to report engaging in political activities).This is
surprising in the light of this study's finding that talk radio listeners were
more likely to have local political knowledge and to report participating in
political activities than were non-listeners. Apparently, gaining political
knowledge and engaging in political activities are a result of listening to talk
radio but not of coming to the format seeking community information.
Summary multiple regressions
Hierarchical multiple regressions were performed to summarize the ability
to predict the cognitive, attitudinal and behavioral outcomes among talk radio
listeners. Predictor variables were entered in blocks with the first block
consisting of demographic, psychological and media use variables. These included
time spent listening to talk radio. The second block consisted of the reasons
for listening to talk radio. The regressions shown do not include parasocial
interaction since inclusion of that variable would eliminate all listeners who
did not report a favorite host.
Table 3 shows the hierarchical regression Beta weights for the cognitive
outcome variable, local political knowledge. Regression 2 added the reasons for
listening as a second block.
Table 3: Predictors of local political knowledge
Predictor
Regression 1
Regression 2
Time listening to radio
-.08
-.07
Days reading newspaper
.07
.07
Time reading newspaper
.04
.05
Time reading magazines
.03
.03
Time watching TV
-.12
-.12
Education level
.12
.13
Age
.008
-.01
How long in area
.18**
.19**
Income
.11
.13
Ideology
.18**
.19**
Loneliness
-.04
-.05
Parasociability
.15*
.13
Time listen to talk radio
.01
-.006
Race (white/other)
-.15*
-.14
Gender
-.31**
-.31**
Companionship
.08
Like the host
-.004
Community information
.11
Multiple R (F significance)
.57 (.0002)
.59 (.0003)
R2
.33
.35
R2 change
.02
Adjusted R2
.23
.23
* p<.10 ** p<.05 *** p<.01
The non-significant contribution of community information-seeking means
that Hypothesis 9 is again not supported. As a related result, adding the
reasons for listening increased the variance explained by the predictors by only
.02. Significant predictors of local political knowledge were length of
residence in the area, ideology (more liberal -> more knowledge) and gender
(males -> more knowledge). Race (white) and parasociability were moderately
significant in the first regression but dropped below significance after the
reasons for listening were added. Education was not significantly related to
local political knowledge, nor were any of the media use variables. In a
re-analysis of American National Election Study data, Waluszko (1995) found that
education was the strongest predictor of political knowledge among radio and TV
talk show audiences, although media use variables including talk show exposure
were among the weakest, similar to the results of this study.
It is not surprising that the longer respondents lived in the area the more
likely they are to have local political knowledge. There is no prima facie
reason to associate males with greater political knowledge. Waluszko, in fact,
(1995) found that females were more likely to have political knowledge.
The regressions did explain 35 percent of the variance in political
knowledge with F significant at p<.0001. The conclusion is that liberal males
who are long-term residents of the area are more likely to have local political
knowledge and that reasons for listening to talk radio and other exposure to
other media do not help to predict that knowledge. Talk radio listening does
appear to aid in local knowledge acquisition but the effect is slight and not
attributable to reasons for (gratifications sought from) listening.
Table 4 gives the results of regressions predicting attitude toward the
city council.
Table 4: Predictors of attitudes toward city council.
Predictor
Regression 1
Regression 2
Time listening to radio
.04
.06
Days reading newspaper
.17*
.15
Time reading newspaper
-.07
-.05
Time reading magazines
.04
.03
Time watching TV
.008
.02
Education level
.08
.09
Age
-.08
-.09
How long in area
-.12
-.12
Income
-.01
-.003
Ideology
-.27**
-.27**
Loneliness
-.02
-.03
Parasociability
.12
.14
Time listen to talk radio
-.24**
-.23**
Race (white/other)
-.11
-.09
Gender
.19*
.20*
Companionship
.10
Like the host
-.08
Community information
-.004
Multiple R (F significance)
.51 (.008)
.52 (.017)
R2
.26
.27
R2 change
.01
Adjusted R2
.15
.14
*p< .10 **p< .05 ***p< .01
The reasons for listening to talk radio did not contribute significantly to
prediction of attitudes toward local government. The R2 value increased only by
.01 when the three reasons were added to the regression equation. The second
equation produced an R value of .52 (F=.02) and explained 27 percent of the
variance in attitude. Significant predictor variables were ideology (more
conservative -> more negative attitude; p= .006), time spent listening to talk
radio (more time spent -> more negative; p= .03) and gender (female -> more
positive; p= .05).
It may be that attributing cognitive and attitudinal outcomes to the
reasons for listening is attempting to fine tune the analysis too much, at least
for a sample of this size, and one drawn non-randomly. Taking listeners as a
group, it was found that talk radio listening predicted a more negative attitude
toward local government, in accordance with the findings of Kohut et al. (1993)
and Hollander (1996). The regressions in Table 4 above extend that finding by
revealing that negativity toward the government increases with time spent
listening. It is also understandable that, at least in Austin, conservative
ideology predicts a more negative attitude toward the government. Tapes made of
KVET talk shows during the survey period, primarily "Sammy and Bob," include
examples attacking the city council on elections, taxes, legislation pertaining
to bicycle helmets, traffic and road maintenance, superfluous city positions and
pro-environmentalism.
Attitudes: None of the individual predictors of preferring the listener's
own or opposite point of view was statistically significant and the F values for
the two regressions were far from significant. The dependent variable in this
case is dichotomous and so the use of multiple regression is debatable, and
discriminant analysis may be more appropriate.
Table 5 gives results of regressions predicting responses to the item
"Call-in talk radio shows are town meetings of the air. More listeners agreed
(44 percent) than disagreed (39 percent) with the statement.
Table 5: predictors of agreement that talk radio is town meeting.
Predictor
Regression 1
Regression 2
Time listening to radio
.18*
.20**
Days reading newspaper
.03
.04
Time reading newspaper
-.16
-.16*
Time reading magazines
-.21*
-.20*
Time watching TV
-.01
-.01
Education level
-.06
-.04
Age
.06
.01
How long in area
.02
-.01
Income
.14
.19*
Ideology
-.07
-.05
Loneliness
.14
.10
Parasociability
.06
-.04
Time listen to talk radio
-.07
-.13
Race (white/other)
.13
.11
Gender
-.004
-.02
Companionship
.06
Like the host
.03
Community information
.38**
Multiple R (F significance)
.41 (.184)
.56 (.004)
R2
.17
.31
R2 change
.14
Adjusted R2
.04
.18
* p< .10 ** p< .05 *** p< .01
In this case, the addition of the reasons for listening increased the
variance explained by 14 percent, resulting in a multiple R significant at p<
.01. Most of this increase was due to the community information reason for
listening (Beta = .38; p = .0001). This is strong support for Hypothesis 8 which
held that those who sought a political forum from talk radio would then perceive
the format as a town meeting of the air. This is not surprising in that a
political forum (redefined as community information) is almost synonymous with
"town meeting." However, this is an important finding in that it adds confidence
in the worth of the community information construct, to the responses to the
town meeting item, and in the hierarchical regression method employed.
Time spent reading magazines (p< .05) and time spent reading the newspaper
(p< .10) had significant negative Beta weights, meaning that the more time
listeners spend reading, the less likely they are to agree that talk radio shows
are time meetings. Those who prefer the print media seem to disparage the
significance of talk radio, perhaps viewing it more as ironic entertainment.
Table 6 shows the results of regressions predicting response to the item
"Talk radio call-in shows are mean-spirited." The majority of listeners
disagreed (60 percent, vs. 20 percent agreeing) with the statement.
Table 6: Predictors of agreement that talk radio is mean-spirited.
Predictor
Regression 1
Regression 2
Time listening to radio
.04
.04
Days reading newspaper
.06
.05
Time reading newspaper
.12
.14
Time reading magazines
-.09
-.08
Time watching TV
-.02
.001
Education level
.09
.07
Age
.15
.16
How long in area
.001
-.00009
Income
-.06
-.06
Ideology
-.19*
-.20**
Loneliness
.16*
.18**
Parasociability
-.09
-.03
Time listen to talk radio
-.29***
-.24**
Race (white/other)
-.11
-.08
Gender
.03
.07
Companionship
.14
Like the host
.03
Community information
-.31***
Multiple R (F significance)
.47 (.031)
.55 (.003)
R2
.22
.31
R2 change
..09
Adjusted R2
.10
.18
* p< .10 ** p< .05 *** p< .01
Results in Table 6 show a similar pattern to those for the town meeting
statement. Addition of the reasons for listening in the second regression
increased the R2 by .09, with a multiple R of .55 (F = .003). This increase was
due primarily to the negative Beta weight for community information-seeking (p =
.001), supporting Hypothesis 7. The more listeners seek community information,
the less likely they are to perceive talk radio as mean-spirited.
Other significant negative predictors were time spent listening to talk
radio (p = .02) and ideology (p = .03). If the content of talk radio is indeed
not mean-spirited, as these respondents think (by 3 to 1), then it is logical to
find that the more one listens the less likely one is to perceive talk radio as
mean-spirited.
Loneliness had a significant positive Beta weight (p = .04).
Companionship-seeking also had a positive Beta weight (p = .13), suggesting that
those looking for interpersonal support or psychological comfort from talk radio
and perhaps not finding it might come to see talk radio as mean.
The fourth item categorized as attitudinal, naming a favorite host (dummy
coded; see Table 7), is distinct from the preceding three in that it does not
describe a belief or attitude about talk radio content. It is perhaps not
conceptually an outcome variable, but it is included in these concluding
regressions because it tests another hypothesis about a specific reason for
listening (Hypothesis 6).
Table 7: Predictors of naming a favorite host.
Predictor
Regression 1
Regression 2
Time listening to radio
.08
.06
Days reading newspaper
-.15
-.12
Time reading newspaper
.03
.04
Time reading magazines
-.03
.001
Time watching TV
-.03
-.02
Education level
.08
.04
Age
.17*
.18*
How long in area
.23**
.22**
Income
.08
.08
Ideology
-.05
-.05
Loneliness
-.13
-.11
Parasociability
.31***
.28***
Time listen to talk radio
.004
.02
Race (white/other)
.08
.07
Gender
-.16
-.14
Companionship
-.001
Like the host
.20**
Community information
-.13
Multiple R (F significance)
.57 (.0002)
.54 (.012)
R2
.33
.29
R2 change
.05
Adjusted R2
.23
.15
* p< .10 ** p< .05 *** p< .01
Addition of the reasons for listening to talk radio increased the explained
variance only slightly (by .05), but the initial regression was already
significant (p = .0002), mostly due to the contribution of parasociability (p =
. 002) and length of time in the Austin area (p = .02). These predictors, along
with age (p = .08) had similar Beta weights in the second regression and were
joined by the reason "Because I like the host" (p = .04) as significant
predictors of naming a favorite host. This supports Hypothesis 6, acting as a
form of reliability check on those reporting affect toward the host (listening
because the host is liked, vs. naming a favorite). Companionship-seeking once
again seemed unrelated to liking the host, with an almost-zero Beta weight (p =
.99). Cautions about the use of multiple regression with a dichotomous
dependent
variable should be born in mind, although the model did show strong support for
expected predictors parasociability and liking the host.
Parasociability, the psychological trait predicting parasocial interactions
with hosts (Auter and Palmgreen, 1993), was further supported as a viable
construct, being the indicator contributing most to prediction of those naming a
favorite host (p = .009). Since it was conceptualized by Auter and Palmgreen
(1993) as a psychological tendency to interact parasocially with a given host,
it is understandable that it helps predict having a favorite (conceptually a
precursor to developing parasocial interactions with that host).
Listeners who lived longest in the Austin area and are older are more
likely to have named a favorite host. The hosts named most often as favorites
(Sammy and Bob in 22 out of 69 cases) are local hosts who have been on the air
for a long time.
The final regression analysis examines predictors of a (reported)
behavioral outcome, political participation:
Table 8: predictors of engaging in political activities.
Predictor
Regression 1
Regression 2
Time listening to radio
-.07
-.07
Days reading newspaper
.08
.09
Time reading newspaper
.05
.03
Time reading magazines
.11
.10
Time watching TV
-.10
-.12
Education level
.13
.13
Age
.30***
.34***
How long in area
.08
.08
Income
.24***
.22**
Ideology
.02
.01
Loneliness
-.14
-.13
Parasociability
-.06
-.02
Time listen to talk radio
-.07
-.06
Race (white/other)
.002
-.02
Gender
-.01
-.04
Companionship
-.19**
Like the host
-.06
Community information
-.07
Multiple R (F significance)
.59 (<.0001)
.63 (<.0001)
R2
.35
.39
R2 change
.04
Adjusted R2
.25
.28
* p< .10 ** p< .05 *** p< .01
The second regression, incorporating the reasons for listening, increased
the explained variance slightly (R2 increased by .04) from the first regression,
which was highly significant (F<.0001). The increase was due primarily to the
negative coefficient for companionship-seeking (Beta = -.19; p = .03) and not to
that of community information-seeking (Beta = -.07; p = .40), meaning that
Hypothesis 10 was not supported. Community information, in fact, was negatively
related to political activities, as was companionship.
There is no clearly obvious mechanism that would explain why
companionship-seeking would lead to fewer political activities. Two of the five
items that formed the activity scale have a social component (attending a
council meeting, joining a community organization) which would predict a
positive relationship between companionship and those sorts of activities. The
fact that an activity item referring to polls and one referring to
letter-writing were not explicitly local might explain the weak relationship
between companionship and political activities as operationalized in this study.
The "forum" concept was reconceptualized in this study as "community
information," a construct that emphasizes the community utility of the
information sought from talk radio. It was expected under Hypothesis 10 that
those who sought and obtained useful information about their community from talk
radio would be moved to act on that information by joining an organization,
attending a meeting, responding to polls, and writing letters. The fact that an
item referring to polls and one referring to letter-writing were not explicitly
local might explain also the weak relationship between community information and
political activities as operationalized in this study. It might be advisable in
future to create distinct measures for local and non-local political
participation.
The strongest predictors, as found in the bivariate tests, of reported
political activities in both regressions were age and income. The greater the
age and income of listeners, the more likely they are to participate in
political activities. Explanations for this relationship seem straightforward,
although tangential to the concerns of this study.
Discussion
One of the strongest conclusions of this study is that talk radio is not
perceived of by its listeners as mean-spirited, despite the fact that listening
to opposite points of view is preferred. This finding suggests an audience that
can "handle" debate, sees that much of it is set up for entertainment value, and
is listening with an ironic ear to any emotions that heat up. At the same time,
listeners are aware that talk radio is not supremely powerful politically, as
seen in the neutral response to an item portraying talk radio as a town meeting.
Tabulation of the uses and gratification items shows that political forum
uses are highly rated, as is one companion-type use (liking the host) and
companionship is by far the least agreed-with reason for listening. On its own
this last finding also runs counter to the portrayal of the marginal audience
member.
Another line of research found that reasons for listening to talk radio are
related to race and ethnicity, the reasoning being that those new to and in the
minority of a community will feel the need to orient themselves to the
mainstream. This was not found in this study.
It was further discovered that liking the host and seeking companionship
are conceptually distinct for this audience, so reasons for listening were split
into companion, host and what is termed here community information (having
elements of interpersonal utility as well as surveillance and information).
It is assumed that the three reasons for listening adequately capture the
essence of listening motivations. However, identifying predictors of reasons for
listening was generally unsuccessful. Psychological characteristics such as
parasociability and loneliness did not predict companion- (or host-) seeking,
and in fact both did predict community-information seeking. There seems to be an
element to community information-seeking that is associated with these
psychological needs in a way that suggests the needs are satisfied by trying to
become more connected to the community rather than to the individuals hosting
the shows. The interaction is not interpersonal, rather it could be termed
intra-community.
The validity of the community information construct is increased by its
ability to predict attitudes toward talk radio content (the media-attitudes
mentioned above). Those who seek community information are more likely to see
talk radio as a town meeting and less likely to see it as mean-spirited. These
findings support the portrayal of an audience that does not react superficially
to the format. Within that audience, there are some who are acutely attuned to
talk radio's ability to provide useful dialogue (the town meeting) in among the
distractions and trivialities, even though listeners were neutral overall in
their agreement with the town meeting nature of the format.
Listening because they liked the host also predicted listeners' naming a
favorite host, which serves as a form of reliability check on both measures.
Parasociability also predicted naming a favorite host as it should, a more
important finding in that parasociability has not been extensively tested.
Multiple regression models generally reaffirmed the stronger relationships
mentioned above. Certain demographic variables emerged as predictors of
outcomes, some explicable (conservatives have a more negative attitude toward
city council) some not (females are more likely to have political knowledge).
Interesting patterns among other media use variables included a sort of
anti-talk stance among heavier readers of magazines and newspapers (seeing talk
radio as mean and not as a town meeting).
Parasocial interaction was also tested as an intervening media experience
variable. This was only possible for about one-third of the listeners, and so it
is not surprising that patterns changed and the equations often were not strong
predictors. Nevertheless, this resulted in some interesting findings (not shown
above). It was a significant predictor of political knowledge, of positive
attitude toward government, and of preferring to hear opposite points of view
(it could not be tested as a predictor of liking the host). Taken together,
these results imply that those who identify (interact parasocially) with hosts
gain political knowledge while enjoying the experience of listening to debate.
In almost the reverse of the set of findings that community information-seeking
was predicted by affective characteristics (loneliness, parasociability), in
this case an affect leads to more knowledge, perhaps through "being in a good
mood" while listening.
Clearly, sorting out the interactions among the cognitive and affective
components of reasons for listening and listening outcomes is as problematic as
separating mass from interpersonal communication. There do seem to some
counter-intuitive results which suggest that the interpersonal elements of
community information (getting information to communicate to others) complicate
the interpretations. Nevertheless, it is a useful construct in predicting
attitudes toward talk radio, suggesting also that further use be made of items
that directly ask listeners about their perceptions of the format. It should be
well established that listeners attend to the format in ways that are more
sophisticated, and that they themselves are less marginal, than the critics
would have it.
In addition to conclusions reported above, this study has potential use for
those in radio broadcasting unfamiliar with audience reasons for attending to
the format and with any psychological or interpersonal interactions with it. If
most listeners do not use the format for companionship but often interact
parasocially with the host and find liking the host to be important, then hosts
need not attempt to coddle or be overly friendly to the listeners. Listeners
seem to enjoy the sense of community that listening to talk radio provides, and
this should be emphasized. The strong relationship between listening and
political activities suggests that hosts should provide encouragement to
listeners to get involved. Follow-up interviews pointed out that listeners
should not be talked down to, nor should controversy or argumentation be avoided
for fear of turning off the audience. Heavy listenership in cars with the
potential for button-setting and "surfing" among channels suggests that
listeners are likely to listen to brief portions of programs, often without keen
awareness of the host.
This study is limited in part by the nature of the self-selected sample.
While the use of a list of radio contest entrants produced respondents
apparently willing to be interviewed, it precluded strong conclusions about the
talk radio audience in general, despite apparent parallels between this study's
respondents and those of other national surveys.
Further research could benefit from replicating the study with a random
sample of respondents, and with a census or random sampling of content.
The picture of the talk radio audience provided here is much different from
that of earlier radio research and from many of the portrayals in the press.
Most likely, changing times, changing uses of the format and biases of the
critics all contributed to this difference. As the format continues to mature,
society evolves, and more attention is paid to it by researchers and
journalists, the most important research agenda may simply be replication of
studies such as this one to allow for a better understanding of a complex and
dynamic communication interaction.
References:
Adams Research. 1995. "Talk Daily" 1995 Talk Radio Survey. Adams Research
Inc., Arlington, Va.
Albarran A.B. and Umphrey, D. 1994. Revisiting Ritualistic and
Instrumental Television Use: Comparing Ethnic Audiences. Manuscript
submitted to Journal of Broadcasting and Electronic Media .
Armstrong, C.B. and Rubin, A.M. 1989. Talk Radio as Interpersonal
Communication. Journal of Communication 39(2): 84-94.
Austin American-Statesman. 1992. Talk shows are the town meeting of the
'90s. Austin American-Statesman (June 10), p. A18.
Auter, P.J. and Palmgreen, P. 1993. Development of Parasocial Interaction
as a Function of Reported Viewing of a Television Program. Paper
presented to the Mass Communication and Society Division of the
Association for Education in Journalism and Mass Communication, Kansas
City, Mo., 1993.
Avery, R.K. and Ellis, D.G. 1979. Talk Radio as an Interpersonal
Phenomenon. In Gumpert, G. and Cathcart, R. (eds.) Inter/Media.
Interpersonal Communication in a Media World. New York, N.Y.: Oxford
University Press. pp. 108-115.
Ball-Rokeach, S.J. and DeFleur, M. 1976. A Dependency Model of Mass Media
Effects. Communication Research 3(1):
Beatty, J.E. 1991. Talk radio content, routines and organization: A case
study of KLBJ-AM, Austin. Unpublished manuscript, University of Texas at
Austin.
Beatty, J.E. 1996. Talk radio as forum and companion: Listener uses and
gratifications in Austin, Texas. Unpublished Ph.D. dissertation, University of
Texas at Austin.
Benchmark Company, The. 1994. Talk Radio in America: An Audience Survey.
The Benchmark Company, Austin, Texas.
Bick, J.H. 1987. The Development of Two-Way Talk Radio in America.
Unpublished Ph.D. dissertation, University of Massachusetts. Amherst,
Mass.
Bierig, J. and Dimmick, J. 1979. The Late Night Talk Show as
Interpersonal Communication. Journalism Quarterly 56(1): 92-96.
Bronstein, E.M. 1983. Local Call-in Radio: A Study of the Search for
Advice and Social Networks. Unpublished Ph.D. dissertation, City
University of New York.
Burd, G. 1988. Communication Culture Adapts to Media Technology: Town
Criers, Newsboys, Radio Commentators and TV Anchors. Paper presented to
the Qualitative Studies Division, Association for Education in Journalism
and Mass Communication, Portland, Ore.
Burns, G. 1994. Talk radio shows are town meetings of the airwaves.
Austin American- Statesman (Aug. 29), p. A11.
Cappella, J.N., Turow, J. and Jamieson, K.H. (1996). Call-In Political
Talk Radio: Background, Content, Audiences, Portrayal in Mainstream Media
. Annenberg Public Policy Center, Philadelphia, Pa.
Crittenden, J. 1971. Democratic Function of the Open Mike Radio. Public
Opinion Quarterly 35(2): 200-210.
Dobos, J. and Dimmick, J. 1988. Factor Analysis and Gratification
Constructs. Journal of Broadcasting and Electronic Media 32(3): 335-350.
Green, C. 1997. "Talk radio helped kill RTD plan." Denver Post (Nov. 7).
p. 2D
Hall, S. 1980. Encoding/Decoding. In Hall, S. et al. (eds.) Culture,
Media, Language: Working Papers in Cultural Studies, 1972-1979. London:
Hutchinson. pp. 128-138.
Herndon, J. 1994. All talk may get no more action. Similar hosts, styles,
have potential to stem growth in format's popularity. Austin
American-Statesman (Aug. 18), p. Onward 17.
Herndon, J. 1996. Good news for locals, not bad for Stern. Austin
American-Statesman (Nov. 7) p. XL 21.
Hofstetter, C.R., Donovan, M.C., Klauber, M.R., Cole, A., Huie, C.J. and
Yuasa, T. 1994. Political Talk Radio: A Stereotype Reconsidered.
Political Research Quarterly 47(2): 467-479.
Hollander, B.A. 1995. The New News and the 1992 Presidential Campaign:
Perceived vs. Actual Political Knowledge. Journalism and Mass
Communication Quarterly 72(4): 786-798.
Hollander, B.A. 1996. Talk Radio: Predictors of Use and Effects on
Attitudes about Government. Journalism and Mass Communication Quarterly 73
(1): 102-113.
Horton, D. and Wohl, R.R. 1956. Mass Communication and Para-Social
Interaction: Observation on Intimacy at a Distance. Psychiatry 19(3):
215-229.
Kohut, A., Zukin, C. and Bowman, C. 1993. The Vocal Minority in American
Politics. Washington, D.C.: Times-Mirror Center for The People & The
Press.
Kurtz, H. 1996. Hot Air. All Talk, All the Time. New York, N.Y.: Random
House.
Laufer, P. 1995. Inside Talk Radio. America's Voice Or Just Hot Air? New
York, N.Y.: Birch Lane Press.
Levin, M.B. 1987. Talk Radio and the American Dream. Lexington, Mass.:
Lexington Books.
Margolis, M. 1990. Talk-radio host Rush Limbaugh and the aesthetics of
insensitivity. Images, supplement to The Daily Texan (Oct. 25), p. 6,7.
Munson, W. 1993. All Talk. The Talkshow in Media Culture. Philadelphia,
Penn.: Temple University Press.
Page, B.I. and Tannenbaum, J. 1996. Populistic Deliberation and Talk
Radio. Journal of Communication 46(2): 33-54.
Paige, S., Danitz, T., Hickey, J.G. and Russell, K. 1998. Talking the
Talk. Washington Times (Feb. 9) p. 09.
Rancer, A.S., Miles, T.T. and Baukus, R.A. 1994. "Communication
predispositions and demographic characteristics of listeners and
non-listeners to talk radio." Paper presented to the annual meeting of
the Eastern Communication Assoc. (April 29), Washington, D.C.
Rehm, D. 1993. Talking Over America's Electronic Backyard Fence. Media
Studies Journal 7(3): 63-69.
Rubin, A.M. and Perse, E.M. 1987. Audience Activity and Soap Opera
Involvement. A Uses and Effects Investigation. Human Communication
Research 14(2): 246-268.
Russell, D., Peplau, L.A. and Cutrona, C. 1980. The Revised UCLA
Loneliness Scale: Concurrent and Discriminant Validity Evidence. Journal
of Personality and Social Psychology 39(3): 472-480.
Scott, G.G. 1996. Can We Talk? The Power and Influence of Talk Shows.
New York, N.Y.: Insight Books.
Surlin, S.H. 1986. Uses of Jamaican Talk Radio. Journal of Broadcasting
and Electronic Media 30(4): 459-466.
Surlin, S.H. and Soderlund, W.C. 1993. Jamaican Call-In Radio: An
Analysis of Callers and Hosts. Paper presented to the International
Communication Division, Association for Education in Journalism and Mass
Communication, Kansas City, Mo.
Tramer, H. and Jeffres, L.W. 1983. Talk Radio D Forum and Companion.
Journal of Broadcasting 27(3): 297-300.
Turow, J. 1974. Talk Show Radio As Interpersonal Communication. Journal
of Broadcasting 18(2): 171-179.
Waluszko, L. 1995. Radio and Television Call-in Shows and Their Impact on
the Public in the 1992 Presidential Campaign. Paper presented to the
Radio-Television Journalism Division, Association for Education in
Journalism and Mass Communication (August), Washington, D.C.
Wilkinson, J.S. and Grant, A.E. 1995. Seasonality and talk radio
audiences. Paper presented to the Media Management and Economics division
of the Association for Education in Journalism and Mass Communication
(August), Washington, D.C.
Wright, R. 1995. Hyper Democracy. Washington isn't dangerously
disconnected from the people; the trouble may be it's too plugged in. Time
145(3): 14-21.
Zerbinos, E. 1993. Talk Radio: Motivation or Titillation? Paper presented
to the Radio-Television Journalism Division of the Association for
Education in Journalism and Mass Communication, (August), Kansas City, Mo.
|
