|
Personal Involvement as a Mediating Variable in the Agenda-setting Process
Introduction
In agenda-setting research, issues on the media and public agendas can be placed
on an "obtrusiveness-unobtrusiveness continuum" (McCombs et al., 1995, p. 288).
Obtrusive issues are those with which the public has direct or personal
experience while unobtrusive issues are those with which the public has little
or no direct or personal experience. However, there appeared to be disagreement
on whether the media had a stronger agenda-setting effect on obtrusive issues or
on unobtrusive issues. When Zucker (1978) first raised the hypothesis of issue
obtrusiveness, he argued that the media had less agenda-setting effect on
obtrusive issues than on unobtrusive issues, because people did not need the
media for information about an issue which was obtrusive in their daily lives
and were more likely to turn to the media for information on unobtrusive issues.
Subsequent studies (Weaver et al., 1981; Winter et al., 1982; Behr & Iyengar,
1985) confirmed that it was with the unobtrusive issues that the media had a
strong agenda-setting effect. Gonzenbach (1996) argued that the media not only
had little effect on obtrusive issues but were sometimes influenced by the
public on these issues.
However, some studies produced findings which are contrary to Zucker's
hypothesis. Lasorsa and Wanta (1990) asked survey respondents about the
obtrusiveness of thirteen issues on the media agenda and found that personal
experience with the issues was positively correlated with "media agenda
conformity" (p. 812). They concluded that the media did indeed have a stronger
agenda-setting effect on issues with which people had a high level of personal
involvement. Dearing and Rogers (1996) reasoned that personal experience with an
issue
Personal Involvement
sensitized people who would then seek further information about the issue in the
media; therefore, "personal experience with an issue might enhance the media
agenda's influence on the public agenda" (p. 53). In other words, the stronger
the need for information or orientation, the more likely people were susceptible
to media's agenda-setting effects (Weaver, 1980; Wanta, 1997).
Obviously, the contradictory findings were based on the same antecedent:
personal involvement. In a review of theories on personal involvement, Thomsen
et al. (1995) maintained that personal involvement determined whether or not
people approached an object, how they approached it, and what happened after
such an interaction. Specifically, two theories delineated how people with high
or low personal involvement might behave toward a communication message. The
elaboration likelihood model (Petty & Cacioppo, 1981; Petty & Priester, 1994)
and the heuristic-systematic model (Chaiken, 1980; Eagly & Chaiken, 1993) stated
that high personal involvement with an issue led to active cognitive processing
of verbal information while low involvement led to the reliance on peripheral or
heuristic cues, such as visual images.
These two models were supported by research in hemispheric lateralization of the
human brain (Hansen, 1981), which indicated that high involvement with an issue
activated the left brain, which is responsible for processing verbal
information, such as newspaper reading, while the right brain controls, among
other things, the processing of visual images, such as television viewing. This
raised the question of whether the differential agenda-setting effects on
obtrusive and unobtrusive issues were related to the difference between
newspapers and television. In other words, newspapers might have a stronger
agenda-setting effect on obtrusive issues while
Personal Involvement
television might have a stronger effect on unobtrusive issues.
Literature Review
The argument that the media have a stronger agenda-setting effect on obtrusive
issues (Lasorsa & Wanta, 1990; Wanta, 1997) has support in traditional theories
concerning personal involvement or personal relevance. Allport (1943) pointed
out that the presence or absence of personal involvement made "a critical
difference in human behavior," because people behaved in a "neutral, impersonal"
manner in the absence of personal involvement while they behaved "quite
differently" once they were "seriously committed to a task" (p. 459). Festinger
(1957) maintained that people were considerably motivated to seek out
information about an object that was "relevant to some impending or possible
future behavior for the person" (p. 125). Markus (1977) pointed out that people
usually activated their self-schemata in processing information about
themselves. Others (Marx & Tombaugh, 1967) emphasized that "motivational
conditions must be present" before the process of information-seeking began (p.
69). Only when individuals are personally involved with an issue, event or
person, will they "care about that entity and perceive it as important" (Thomsen
et al., 1995, p. 191).
The social judgement theory (Sherif & Hovland, 1961) posited that an
individual's involvement with an issue of intrinsic importance was "a main
anchor for evaluation" (p. 174). A message "closely approximating" one's own
attitude or experience would be assimilated while a message "diverging markedly"
from one's attitude or experience would be contrasted and rejected (p. 188).
Personal Involvement
Eagly (1967) found that attitude change was more with high involvement than with
low involvement when pro-attitudinal messages were provided to subjects, i.e.,
in a "positive-discrepancy condition" (p. 3). Pallak et al. (1972) also found
that people of high involvement with an issue were "more likely to consider an
extreme consonant communication, assimilate its contents, and change toward the
position advocated than subjects under low commitment" (p. 433). Petty and
Cacioppo (1979) agreed that increased involvement was associated with more
cognitive processing, during which people "assimilated the proattitudinal
information, producing acceptance, and contrasted the counterattitudinal
information, producing resistance" (p. 1920).
In agenda-setting research, personal involvement or personal motivation has been
found to be embedded in the very first study (McCombs & Shaw, 1972), which was
conducted on 100 voters who "had not yet definitely decided how to
vote--presumably those most open or susceptible to campaign information" (p.
178). Dearing and Rogers (1996) pointed out that "individuals with high
uncertainty" were more likely to "become oriented about an issue," and that such
a high need for orientation would "result in greater agenda-setting effects" (p.
52). Erbring et al. (1980) made a conclusion that echoed Festinger's theory of
cognitive dissonance: People perceived media messages selectively to the extent
that they sought out and attended to "relevant" messages but casually bypassed
or forgot irrelevant ones, thus "routinely apportioning a limited attention
budget" (p. 28).
Weaver (1980) introduced the psychological theories of motivation, including
Tolman's theory of cognitive mapping, and developed a typology containing two
dimensions: relevance and uncertainty. He found that a high level of interest in
an issue, together with a high level of
Personal Involvement
uncertainty, led to a strong "need for orientation" (p. 365). People with a high
need for orientation about political matters were found to be more susceptible
to the agenda-setting influence than those with a moderate need for orientation.
Yagade and Dozier's (1990) classification of issues as concrete and abstract had
an element of personal involvement. Part of their operational definition of a
concrete issue was whether survey respondents found an issue "real" and
understood "all the basics" of the issue (p. 7). Their findings showed that the
media had stronger effects on concrete issues than on abstract ones, which were
difficult to conceptualize.
However, low personal involvement with an issue was not without its place in the
agenda-setting process. Petty and Cacioppo (1981) posited in their elaboration
likelihood model (ELM) that highly-involved people tended to take the central
route and engaged in "diligent consideration of issue-relevant arguments," while
people of low involvement tended to take the peripheral route and rely on
non-verbal cues, such as source credibility or visual images, which allowed a
person to "decide what attitudinal position to adopt without the need for
engaging in any extensive issue-relevant thinking" (p. 847).
Proposing a heuristic-systematic model (HSM), Chaiken (1980) also maintained
that highly involved people would take a systematic route by which they devoted
considerable cognitive effort to processing a message, while people who did not
consider an issue to be of much personal relevance would "employ a more economic
heuristic strategy" and rely on non-content cues (p. 762). Her experiments
showed that subjects of high involvement were significantly affected by
arguments in a message but were unaffected by communicator
Personal Involvement
likability, while subjects of low involvement were significantly influenced by a
likable communicator but unaffected by the arguments.
The dichotomy in approaching a communication message was termed by Langer and
Imber (1980) as mindful and mindless ways of processing information. Their study
found that the "mindless" people received incoming informational stimuli in a
"passive, reactive fashion" while the "mindful" people were engaged in "active"
processing and "conscious manipulation" of incoming messages (p. 360).
Further empirical evidence came from Borgida and Howard-Pitney (1983), who found
that high-involvement subjects were more influenced by the arguments of a
message, regardless of the manipulation of the visual salience of the source. In
contrast, the low-involvement subjects were easily influenced by the visual
salience of the message source. Other studies (Rossiter & Percy, 1983; Edell &
Staelin, 1983) confirmed that visual images were an effective peripheral cue in
changing people's attitudes. Other non-verbal cues included music (Gorn, 1982;
Middlestadt et al., 1995), emotions or moods (Gardner, 1995), source expertise
(Johnson & Scileppi, 1969), and source credibility (Rhine & Severance, 1970).
The underlying rationale for the function of central or systematic and
peripheral or heuristic routes was provided by bio-psychological studies on
human responses to objects appearing in the left or right visual field (Sperry,
1974) and studies on patients sustaining gunshot wounds to the left or right
hemisphere, as well as studies employing electroencephalograms and radioisotopes
(Weinstein, 1982). These studies demonstrated that the left hemisphere
controlled verbal information processing while the right hemisphere controlled
Personal Involvement
the processing of pictorial, musical, and other nonverbal information (Hansen,
1981). Applying the theory of hemispheric lateralization, Krugman (1977)
explicitly pointed out that reading newspaper was a "left-brain" and "high
involvement" activity while watching television was a "right-brain" and "low
involvement" activity (p. 8). Singer (1980) held similar views that reading
engaged the brain "in a more complex way" than "the more passive television
viewing" (p. 57).
A closer examination of previous agenda-setting studies indicated that the
stronger agenda-setting effect of newspapers was related to issues that the
public found to be more relevant or important, such as economic issues (Benton &
Frazier, 1976) or political issues (McClure & Patterson, 1976). On the contrary,
a stronger agenda-setting effect for unobtrusive issues was related to the low
involvement medium of television (Zucker, 1978; Behr & Iyengar, 1985; Iyengar,
1988). Weaver et al. (1976, cited in Winter, 1981) found that different types of
audience--those relying on newspapers and those relying on
television--demonstrated "distinct differences in issue agendas" (p. 238). For
people with a high level of need for orientation, newspaper reading was more
strongly associated with seeking "political information and political
knowledge," while for those with a low need for orientation, television was a
source of "reinforcement or excitement" (Weaver, 1980, p. 371).
McClure and Patterson (1976) found that television news viewers "could describe
a candidate's appearance on television in great detail" but could "remember
nothing that he said" (p. 26). Their analysis (Patterson & McClure, 1976) of
television viewers' recall of the 1972 campaign stories showed that the viewers'
memories were dominated by what their eyes had seen and that only 20 percent of
their recall was about what their ears had heard. More
Personal Involvement
pertinently, a Swedish study (Asp, 1983) found that highly involved voters were
"in greater agreement with the agenda" of their party newspapers, while voters
who were not "particularly interested in politics" showed "a greater agenda
agreement with the TV news" (p. 345).
Research Question
Newspapers and television are similar in news topics they selected (Stempel,
1985) or in "news priorities" they presented (Dearing & Rogers, 1996, p. 32).
This has been attributed to the socialization of journalists (Atwood & Grotta,
1970), the role of wire services (Breed, 1955; Whitney & Becker, 1982), and the
common definitions of news (Clyde & Bucklaw, 1969; Dearing & Rogers, 1996).
However, agenda-setting researchers recognize that there exist "clear
print-broadcast differences" (Tipton et al., 1975, p. 19), because television is
apparently "more visually oriented" than newspapers (McCombs, 1977, p. 98).
Some (Eyal, 1981; Berkman & Kitch, 1986; Wanta, 1997) have proposed that
newspapers have a stronger effect because newspaper readers can process
newspaper information at their own pace and at multiple time points while
television viewers have little time to process and comprehend the incoming
information, which is presented in rapid succession. Others (Palmgreen & Clarke,
1977) disagreed and argued that television has a stronger effect because it has
the advantage of being a visual medium and provides the audience with a sense of
participation. These arguments have ignored the possible link between media's
differential effects and the personal involvement of the audience.
According to the elaboration likelihood model or the heuristic-systematic model,
verbal
Personal Involvement
information in newspapers is actively sought out by people who are highly
involved with an issue while visual images on television are more appropriate
for people of low involvement. This study aimed to find out whether newspapers
has a stronger effect on issues of high personal involvement while television
has a stronger effect on issues of low personal involvement.
This study also attempted to address a few methodological problems. In most
agenda-setting studies, "issue obtrusiveness was researcher-defined rather than
defined by the individual respondent" (Einsiedel et al., 1984, p. 132). As a
result, different researchers have placed issues at different points on the
continuum of obtrusiveness. Zucker classified the crime issue as obtrusive, but
Weaver et al. (1981) treated it as an unobtrusive issue. Zucker treated
unemployment as an obtrusive issue, but Winter et al. (1982) treated it as
unobtrusive. Zucker considered drug abuse as an unobtrusive issue, but
Gonzenbach (1996) believed it was an obtrusive one. Apparently, these studies,
which came to the same conclusion that the media have stronger effects on
unobtrusive issues than on obtrusive issues, are inherently contradictory to one
another.
This study examined issues individually so that the methodological problem of an
aggregated issue agenda would be avoided (Erbring et al., 1980; Becker, 1982).
Such an individual approach was based on the argument that it was hardly
possible that "the entire list of issues" (Zucker, 1978, p. 227; Weaver et al.,
1981, p. 99) would be transferred from the media agenda to the public agenda.
Evidence (Culbertson & Stempel, 1984; Wanta et al., 1989; Shaw & Martin, 1992)
showed that at times there existed no correlation between a "wholesale
rank-ordering of issues" on the media agenda and the same rank-ordering of
issues on the public
Personal Involvement
agenda because negative correlations for the obtrusive issues canceled out the
positive correlations for the unobtrusive issues (Winter et al., 1982, p. 1).
This study examined the media individually so that inter-media difference could
be found (Becker, 1982). Some important studies had limitations because they
were based on the examination of only one type of media. Zucker (1978), who
found a stronger agenda-setting effect associated with unobtrusive issues, drew
his conclusion from examining television, a low involvement medium which might
differ sharply from a newspaper, a high involvement medium. Behr and Iyengar
(1985), who made the same conclusion that public concern about obtrusive issues
such as employment was hardly affected by media, also drew their conclusion from
examining television coverage. On the other hand, Yagade and Dozier (1990), who
found that media had a stronger effect on concrete or obtrusive issues, made
their conclusion from studying Time magazine, a higher involvement medium than
television.
Hypotheses
Based on the literature review and research questions, this study hypothesized:
1. Newspapers and television have similar issue agendas.
2. Newspapers provide more information than television.
3. Newspapers have a stronger agenda-setting effect than television on obtrusive
issues.
4. Television has a stronger agenda-setting effect than newspapers on
unobtrusive issues.
Personal Involvement
Methods
The methods were content analysis of the media coverage and secondary analysis
of the Gallup poll data. Content analysis was conducted by keyword searches on
Lexis-Nexis (classic version), an electronic database of news and legal
information. News programs of two news media, the New York Times and the ABC
network, were studied. Pearson correlations were run between the media and
public agendas; paired-sample t-tests were used to compare the newspaper and
television agendas.
Based on the agenda-setting theory that the media agenda set the public agenda,
the media agenda was treated as the independent variable, which was
conceptualized as issue covered by the media and operationalized in this study
as the number of stories covering an issue. Though "story length" conveys issue
"prominence" on the media agenda (Eaton, 1989, p. 944), most studies have
defined the media agenda as the number of stories devoted to an issue (Erbring
et al., 1980; Winter & Eyal, 1981; Winter et al., 1982), because the number of
stories was highly correlated with the number of column inches and therefore was
a sufficient measurement of the media agenda (Stone & McCombs, 1981).
The dependent variable, the public agenda, was conceptualized as the awareness
of media issues and operationalized as the degree of importance given to an
issue by the public. The degree of importance was expressed by the percentage of
the public considering an issue as the most important problem facing the nation.
This study used the secondary source of the Gallup polls, whose
most-important-problem (MIP) question has been widely used in measuring the
public agenda (Funkhouser, 1973; Zucker, 1978; Eaton, 1989; Ader, 1995). The MIP
question
Personal Involvement
has been considered as "the best way to measure the public agenda" (Dearing &
Rogers, 1996, pp. 83-84).
As the purpose here was to study the difference between issues of high
involvement or obtrusiveness and issues of low involvement or unobtrusiveness,
one issue from each end of the continuum of involvement was chosen: health care
and crime. This would "allow the independent examination of separate media and
distinct issues" (Eyal et al., 1981, p. 217).
Different from previous studies, which classified issues intuitively, this study
resorted to real-world indicators and classified these two issues by the
percentage of the population involved with them. The crime issue was classified
as an unobtrusive issue according to the Statistical Abstract of the United
States, which showed that the percentages of the population becoming victims of
crime from 1991 through 1996 were respectively 14, 13.2, 16.9, 16.3, 15.2, and
13.9 (1995, p. 204; 1998, p. 216). On the other hand, the health care issue was
obtrusive according to the Abstract, which showed that the percentages of the
population having health care coverage from 1990 through 1996 were respectively
86.1, 85.9, 85, 84.7, 84.8, 84.6, and 84.4 (1998, p. 125).
The New York Times was selected because it was a sufficient indicator of
newspaper coverage of national issues. Agenda-setting researchers (Dearing &
Rogers, 1996) considered the Times as "the most respected U.S. news medium" (p.
32) which plays "a particularly crucial role in media agenda-setting for many
national issues" (p. 39). The paper has been regarded as "the dominant newspaper
in America" (Berkman & Kitch, 1986, p. 27), and a national "bulletin board"
(White, 1973, p. 259) shaping the issue agenda for the nation (Crouse, 1972).
The pace-
Personal Involvement
setting influence of the Times on other papers has been termed the "arterial"
phenomenon (Breed, 1955, p. 279), meaning that the influence of the Times flows
to less influential media.
The rationale of selecting ABC was that it differed little from other major
television networks in covering national issues. Weaver et al. (1981) found that
the issue agendas of the three television networks were "much more similar than
dissimilar" to one another (Weaver et al., 1981, p. 94). Erikson et al. (1988)
observed that the three major networks started news programs with "the same lead
story 43 percent of the time" (p. 214). Lemert (1974) pointed out that
duplication of stories among the television networks was common. Capo (1983)
found the television networks' duplication rates at 74.6 percent in 1972 and
84.3 percent in 1973 during the Watergate coverage. The duplication rate in
business news coverage was 59 percent (Dominick, 1981). In international news
coverage, networks were similar in the number of reports, minutes of coverage
and topics (Altheide, 1982).
As one single poll's data might not be representative of public opinion, this
study collected data from ten Gallup polls conducted between 1991 and 1994. A
longitudinal design provides "a check on the stability of correlations" between
the media and the public agendas (Sohn, 1978, p. 328). Erbring et al. (1980)
argued that agenda-setting studies would benefit from "the self-anchoring
properties of a longitudinal design" (p. 20). If agenda-setting was a process
going on over time, as Sohn (1978) argued, a longitudinal design would
demonstrate "a consistency and generality of results" (p. 328).
The year 1990 was chosen as the starting point of this study, based on the fact
that the health care issue was first recognized as the most important problem by
the public in late 1990,
Personal Involvement
as indicated by the Gallup poll in November 1990. The year 1994 was chosen as
the ending point because the Gallup Poll stopped asking the
most-important-question in that year (The question was resumed in 1997).
Altogether, ten polls between 1990 and 1994 were used.
The counting of the stories on Lexis-Nexis was mainly conducted by the author. A
20-week period of media coverage prior to each of the ten polls was studied.
This time frame was larger than those in most agenda-setting studies, so that
variation in the agenda-setting effect might be captured. The starting point for
each of the 20 weeks of media coverage was the date on which a poll ended (See
appendixes for the dates of the Gallup polls). Prior to each poll, the media
coverage was divided into 20 units cumulatively from week one through week 20.
This cumulative approach was based on the assumption that each added news story
on an issue would "presumably reach an additional measure of attention among
those reached" (Erbring et al., 1980, p. 28).
A test of inter-coder reliability by Scott's pi was 95 percent. The high
agreement rate was possible because the keyword search always produced the same
story count when the commands were typed correctly. The procedure of content
analysis was to enter the "file" of a specific medium, e.g., NYT or ABCNEW, in
the "news library" of Lexis-Nexis and then type the command--"health care (or
crime) bef (a date, e.g., 3/7/91) and aft (a date, e.g., 2/27/91)". This would
produce a count of the stories in the week between March 6, 1991 and February
28, 1991.
Findings
The findings supported the hypotheses. The New York Times and ABC network had
Personal Involvement
similar issue agendas but NYT, as a print medium, provided more verbal
information and had a significantly stronger effect than ABC on the high
involvement issue of health care, while ABC, as a visual medium, provided much
less verbal information but had a significantly stronger effect than NYT on the
low involvement issue of crime.
Hypothesis One stated that newspapers and television would have similar issue
agendas. The hypothesis was supported by the data as shown in Table One, which
lists the frequencies of news stories. The frequencies of news stories by NYT
and ABC were highly correlated on both the health care issue (Pearson r = .807,
p < .01) and the crime issue (Pearson r = .799, p < .01). This indicated that
there indeed existed inter-media correlations. This finding corroborates
Stempel's (1985) finding that mass media share the same mix of news topics, and
confirms Berkman and Kitch's (1986) observation that the agenda-setting function
was possible because the issues composing the various media agendas were
strikingly similar.
Hypothesis Two, that newspapers would provide more verbal information than
television was also supported. The frequencies of news stories provided by NYT
were consistently higher than those by ABC, as shown in Table One. Paired
samples t-tests, shown in Table Two, indicated a significant difference between
NYT and ABC in the amount of coverage of the health care issue (t = 10.195, df =
9, p < .001) and the crime issue (t = 34.697, df = 9, p < .001). This finding
indicates that newspapers and television took different approaches in reporting
issues. Newspapers' role was obviously centered on in-depth coverage while the
more visually oriented television cut into reality "at a different angle"
(McCombs, 1977, p. 98) and sought "good visuals" instead of "lengthy
explanations" (Berkman & Kitch, 1986, p. 123).
Personal Involvement
Hypotheses Three and Four were that newspapers would have a stronger effect than
television on obtrusive issues and that television would have a stronger effect
than television on unobtrusive issues. Findings supported these two hypotheses.
Table Three lists the correlations between the public opinion polls and the
media coverage in 20 cumulative weeks prior to each of the ten polls. On the
health care issue, the correlations between NYT and the public were consistently
higher than those between ABC and the public, whereas, on the crime issue, the
ABC/public correlations were consistently higher than NYT/public correlations.
Paired samples t-tests on the correlations indicate significant difference on
both the health care issue (t = 21.518, df = 19, p < .001) and the crime issue
(t = -7.01, df = 19, p < .001). This finding confirms McCombs' prediction (1977)
that there are "distinct agenda-setting roles for newspapers and television" (p.
96). Newspapers seems to have superiority in setting the public agenda on issues
of high involvement while television, with its visual advantages, is superior in
affecting the public awareness of low involvement issues.
This study also reaffirmed the basic theory of agenda-setting that media set the
public agenda (McCombs & Shaw, 1972) and that the priorities of the media would
"become in time the priorities of the public" (Weaver et al., 1981, p. 6). Data
in Table Three show that the cumulative coverage by both NYT and ABC on the
health care and crime issues was highly correlated with the public awareness of
these two issues.
In addition, this study found that newspapers and television had different
effect spans. On the health care issue, correlations between NYT coverage and
public opinion were significant across all 20 weeks, whereas there was a gap of
insignificant correlations appeared between ABC
Personal Involvement
coverage and public opinion from week eight to week eleven. On the crime issue,
significant correlations were found between ABC coverage and public opinion
across all 20 weeks, while a gap of insignificant correlations appeared between
NYT coverage and public opinion from week two to week eight. This finding
indicates that the selection of the time lag is crucial in agenda-setting
research, because the agenda-setting effect might not occur at certain time
points.
Conclusion
Congruent with theories of personal involvement, such as the elaboration
likelihood model (Petty & Cacioppo, 1981), the heuristic-systematic model
(Chaiken, 1980), and the hemispheric lateralization (Sperry, 1974; Krugman,
1977), which posit that people actively process verbal information in high
involvement situations but rely on non-verbal cues in low involvement
situations, this study confirms that different degrees of personal involvement
with issues lead to differential agenda-setting effects of print media and
visual media. Newspapers are verbally oriented and provides substantially more
information than the visual medium of television, and therefore newspapers have
a stronger agenda-setting effect on issues of high personal involvement or high
degrees of obtrusiveness. On the other hand, the visual images projected by
television are more likely to influence those who consider an issue to be of
little personal relevance.
Therefore, the traditional argument that media have a stronger effect on issues
with which people have little or no direct experience seems to result from
examining the visual or low-involvement medium--television (Zucker, 1978; Behr &
Iyengar, 1985). On the other hand, the
Personal Involvement
argument that media have a stronger effect on issues of personal involvement or
personal relevance seems more valid for print media (Yagade & Dozier, 1990;
Lasorsa & Wanta, 1990).
This study shows that newspapers and television might share the same issue
agendas but differ in their effects on the public agenda. Therefore, it is
necessary that differentiation be made between newspapers and television in
agenda-setting research. As Wanta (1997) recently pointed out, television is
visually oriented and involved "a different type of information processing in an
individual" than verbally or textually oriented newspapers (p. 62).
The finding on the different effect spans of newspapers and television also has
significant implications for agenda-setting research. The selection of the time
lag in previous studies ranged from one week (Kaid et al., 1977), two weeks
(Weaver et al., 1992), three weeks (Erbring et al., 1980), one month (Wanta &
Wu, 1992), two months (Sohn, 1978), three months (Ader, 1995), four months
(McCombs, 1977), six months (Einsiedel et al., 1984), to one year (Funkhouser,
1973). Eyal et al. (1981) pointed out that the selection of the time lag in most
agenda-setting studies was based on "intuition," "guesswork," "conjecture," or
"speculation" (pp. 213-214). The study indicated that the agenda-setting effect
might or might not appear at a certain time point for one particular medium and
that the selection of the time lag is crucial in determining whether a research
hypothesis might be supported.
This study analyzed two issues, which were at the opposite ends of the continuum
of obtrusiveness. Issues in the middle part of the continuum might demonstrate
different characteristics. Also, the obtrusiveness of an issue is not a static
property. Issue obtrusiveness might increase or decrease at different time
points with different segments of the public. The
Personal Involvement
crime issue was unobtrusive at the national level but could be an obtrusive
issue in certain geographic areas. Likewise, the obtrusiveness of the
unemployment issue might increase during an economic recession and decrease in a
period of economic growth.
References
Ader, C. (1995). A longitudinal study of agenda setting for the issue of
environmental pollution. Journalism Quarterly, 72, 300-311.
Allport, G. (1943). The ego in contemporary psychology. Psychological Review,
50, 451-478.
Altheide, D. (1982). Three-in-one news: Network coverage of Iran. Journalism
Quarterly, 59, 482-486.
Asp, K. (1983). The struggle for the agenda: Party agenda, media agenda, and
voter agenda in the 1979 Swedish election campaign. Communication Research, 10,
333-355.
Atwater, T., Salwen, M., & Anderson, R. (1985). Interpersonal discussion as a
potential barrier to agenda-setting. Newspaper Research Journal, 6 (4), 37-43.
Atwood, L., & Grotta, G. (1970). Socialization of news values in beginning
reporters. Journalism Quarterly, 47, 296-302.
Atwood, E., Sohn, A., & Sohn, H. (1978). Daily newspaper contributions to
community discussion. Journalism Quarterly, 55, 570-576.
Behr, R.L., & Iyengar, S. (1985). Television news, real-world cues, and changes
in the public agenda. Public Opinion Quarterly, 49, 38-57.
Benton, M., & Frazier, P. (1976). The agenda-setting function of the mass media
at three levels of "information holding." Communication Research, 3 (3),
261-274.
Borgida, E., & Howard-Pitney, B. (1983). Personal involvement and the robustness
of perceptual salience effects. Journal of Personality and Social Psychology,
45, 560-570.
Breed, W. (1955). Newspaper "opinion leaders" and processes of standardization.
Journalism Quarterly, 32, 277-284, 328.
Brewer, M., & McCombs, M. (1996). Setting the community agenda. Journalism
Quarterly, 73, 7-16.
Brosius, H.B., & Kepplinger, H.M. (1990). The agenda-setting function of
television news: Static and dynamic views. Communication Research, 17, 183-211.
Brown, L. (1971). Television. New York: Harcourt Brace Jovanovich.
Capo, J. (1983). Network Watergate coverage patterns in late 1972 and early
1973. Journalism Quarterly, 60, 595-602.
Chaiken, S. (1980). Heuristic versus systematic information processing and the
use of source versus message cues in persuasion. Journal of Personality and
Social Psychology, 39 (5), 752-766.
Clyde, R., & Bucklaw, J. (1969). Intermedia standardization: A Q-analysis of
news editors. Journalism Quarterly, 46, 349-351.
Crouse, T. (1972). The boys on the bus. New York: Random House.
Culbertson, H.M., & Stempel III, G.H. (1984). Possible barriers to agenda
setting in medical news. Newspaper Research Journal, 5(3), 53-60.
Dearing, J., & Rogers, E. (1996). Agenda-setting. Thousand Oaks, CA: Sage.
Dominick, J. (1981). Business coverage in network newscasts. Journalism
Quarterly, 58, 179-185, 191.
Eagly, A. (1967). Involvement as a determinant of response to favorable and
unfavorable information. Journal of Personality and Social Psychology Monograph,
7 (3), 1-15.
Eagly, A., & Chaiken, S. (1993). The psychology of attitudes. Fort Worth, TX:
Harcourt Brace.
Eaton, H., Jr. (1989). Agenda-setting with bi-weekly data on content of three
national media. Journalism Quarterly, 66, 942-948, 959.
Edell, J., & Staelin, R. (1983). The information processing of pictures in print
advertisements. Journal of Consumer Research, 10, 45-61.
Einsiedel, E., Salomone, K., & Schneider, F. (1984). Crime effects of media
exposure and personal experience on issue salience. Journalism Quarterly, 61,
131-136.
Erbring, L., Goldenberg, E., & Miller, A. (1980). Front-page news and real-world
cues: A new look at agenda-setting by the media. American Journal of Political
Science, 2 (1), 17-49.
Eyal, C. (1981). The roles of newspapers and television in agenda-setting. In G.
C. Wilhoit & H. DeBock (Eds.), Mass communication review yearbook 2 (pp.
225-234). Beverly Hills, CA: Sage.
Eyal, C., Winter, J., & DeGeorge, W. (1981). The concept of time frame in
agenda-setting. In G.C. Wilhoit & H. DeBock (Eds.), Mass communication review
yearbook 2 (pp. 212-218). Beverly Hills, CA:Sage.
Festinger, L. (1957). A theory of cognitive dissonance. Stanford, CA: Stanford
University Press.
Funkhouser, G.R. (1973). The issues of the sixties: An exploratory study in the
dynamics of public opinion. Public Opinion Quarterly, 37, 62-75.
Gardner, M. (1995). Responses to emotional and informational appeals: The
moderating role of context-induced mood states. In E. Clark, T. Brock & D.
Stewart (Eds.), Attention, attitude, and affect in response to advertising (pp.
207-221). Hillsdale, NJ: Lawrence Erlbaum.
Gorn, G. (1982). The effects of music in advertising on choice behavior: A
classical conditioning approach. Journal of Marketing, 46, 94-101.
Hansen, F. (1981). Hemispheral lateralization: Implications for understanding
consumer behavior. Journal of Consumer Research, 8, 23-36.
Iyengar, S., Peters, M., & Kinder, D. (1982). Experimental demonstrations of the
"not-so-minimal' consequences of television news programs. American Political
Science Review, 76 (4), 848-858.
Johnson, H., & Scileppi, J. (1969). Effects of ego-involvement conditions on
attitude change to high and low credibility communicators. Journal of
Personality and Social Psychology, 13 (1), 31-36.
Kaid, L., Hale, K., & Williams, J. (1977). Media agenda-setting of a specific
political event. Journalism Quarterly, 54, 584-587.
Krugman, H. (1977). Memory without recall, exposure without perception. Journal
of Advertising Research, 17 (4), 7-12.
Langer, E., & Imber, L. (1980). Role of mindlessness in the perception of
deviance. Journal of Personality and Social Psychology, 39 (3), 360-367.
Lasorsa, D., & Wanta, W. (1990). The effects of personal, interpersonal and
media experience on issue salience. Journalism Quarterly, 67, 804-813.
Lemert, J. (1974). Content duplication by the networks in competing evening
newscasts. Journalism Quarterly, 51, 238-244.
Markus, H. (1977). Self-schemata and processing information about the self.
Personality and Social Psychology, 35 (2), 63-78.
McClure, R.D., & Patterson, T.E. (1976). Setting the political agenda: Print vs.
network news. Journal of Communication, 26 (2), 23-28.
McCombs, M. (1977). Newspapers versus television: Mass communication effects
across time. In D. Shaw & M. McCombs (Eds.), The emergence of American political
issues: The agenda-setting function of the press (pp. 89-105). St. Paul, MN:
West.
McCombs, M., Danielian, L., & Wanta, W. (1995). Issues in the news and the
public agenda: The agenda-setting tradition. In T. Glasser & C. Salmon (Eds.),
Public opinion and the communication of consent (pp. 281-300). New York:
Guilford Press.
McCombs, M., & Shaw, D. (1972). The agenda-setting function of mass media.
Public Opinion Quarterly, 36, 176-187.
Middlestadt, S., Fishbein, M., & Chan, D. (1995). The effect of music on brand
attitudes: Affect- or belief-based change? In E. Clark, T. Brock & D. Stewart
(Eds.), Attention, attitude, and affect in response to advertising (pp.
149-167). Hillsdale, NJ: Lawrence Erlbaum.
Pallak, M., Mueller, M., Dollar, K., & Pallak, J. (1972). Effect of commitment
on responsiveness to an extreme consonant communication. Journal of Personality
and Social Psychology, 23, 429-436.
Palmgreen, P., & Clarke, P. (1977). Agenda-setting with local and national
issues. Communication Research, 14, 435-452.
Patterson, T., & McClure, R. (1976). The unseeing eye: The myth of television
power in national elections New York: Putnam.
Petty, R., & Cacioppo, J. (1981). Attitudes and persuasion: Classic and
contemporary approaches. Dubuque, Iowa: Wm. C. Brown Company.
Petty, R., & Priester, J. (1994). Mass media attitude change: Implications of
the elaboration likelihood model of persuasion. In J. Bryant & D. Zillmann
(Eds.), Media effects: Advances in theory and research (pp. 91-122). Hillsdale,
NJ: Lawrence Erlbaum.
Rogers, E., & Dearing, J. (1988). Agenda-setting research: Where has it been,
where is it going? In J. Anderson (Ed.), Communication yearbook 11 (pp.
555-594). Newbury Park, CA: Sage.
Shaw, D., & Martin, S. (1992). The function of mass media agenda-setting.
Journalism Quarterly, 69, 902-920.
Sherif, M., & Hovland, C. (1961). Social judgment: Assimilation and contrast
effects in communication and attitude change. New Haven: Yale University Press.
Singer, J. (1980). The power and limitations of television: A
cognitive-affective analysis. In P. Tannenbaum (Ed.), The entertainment
functions of television (pp. 31-65). Hillsdale, NJ: Lawrence Erlbaum.
Sohn, A. (1978). A longitudinal analysis of local non-political agenda-setting
analysis. Journalism Quarterly, 55, 325-333.
Sperry, R. (1974). Lateral specialization in the surgically separated
hemispheres. In F. Schmitt & F. Worden (Eds.), The neurosciences: Third study
program. Cambridge, MA: MIT Press.
Statistical Abstract of the United States 1995. Washington, DC: U.S. Bureau of
the Census.
Statistical Abstract of the United States 1998. Washington, DC: U.S. Bureau of
the Census.
Stempel III, G.H. (1985). Gatekeeping: The mix of topics and the selection of
stories. Journalism Quarterly, 62, 791-796, 815.
Thomsen, C., Borgida, E., & Lavine, H. (1995). The causes and consequences of
personal involvement. In R. Petty & J. Krosnick (Eds.), Attitude strength:
Antecedents and consequences (pp. 191-214). Mahwah, NJ: Lawrence Erlbaum.
Wanta, W. (1997). The public and the national agenda: How people learn about
important issues. Mahwah, NJ: Lawrence Erlbaum.
Wanta, W., & Wu, Y. (1992). Interpersonal communication and the agenda-setting
process. Journalism Quarterly, 69, 847-855.
Weaver, D. (1980). Audience need for orientation and media effects.
Communication Research, 7 (3), 361-376.
Weaver, D., Graber, D.A., McCombs, M.E., & Eyal, C.H. (1981). Media
agenda-setting in a presidential election: Issues, images, and interests. New
York: Praeger.
Weaver, D., Zhu, J., & Willnat, L. (1992). The bridging function of
interpersonal communication in agenda-setting. Journalism Quarterly, 69,
856-867.
Weinstein, S. (1982). A review of brain hemisphere research. Journal of
Advertising Research, 22 (3), 59-63.
Whitney, D., & Becker, L. (1982). "Keeping the gates" for gatekeepers: The
effects of wire news. Journalism Quarterly, 59, 60-65.
Winter, J.P., & Eyal, C.H. (1981). Agenda-setting for the civil rights issue.
Public Opinion Quarterly, 45, 376-383.
Winter, J.P., Eyal, C.H., & Rogers, A.H. (1982). Issue-specific agenda-setting:
The whole as less than the sum of the parts. Canadian Journal of Communication,
8, 8-10.
Yagade, A., & Dozier, D.M. (1990). The media agenda-setting effect of concrete
versus abstract issues. Journalism Quarterly, 67, 3-10.
Zucker, H.G. (1978). The variable nature of news media influence. In B.D. Ruben
(Ed.), Communication yearbook 2 (pp. 225-240). New Brunswick, NJ: Transaction
Books.
Table 1
Frequencies of 20 Weeks' Coverage by NYT and ABC
and the Pearson Correlation between NYT and ABC
Poll Health Care Crime
NYT ABC NYT ABC
1 493 48 1762 66
2 546 32 1542 97
3 589 41 1480 108
4 719 58 1486 97
5 877 101 1732 118
6 827 79 1774 125
7 965 87 1764 101
8 1155 102 1699 149
9 1351 171 2205 381
10 1199 271 1986 438
Pearson r .807** .799**
** p < .01
Table 2
Paired Samples t-test between 20 Weeks' Coverage by NYT and ABC
Issue Mean of Frequencies t value df Sig (2-tailed)
of News Stories
NYT ABC
Health
Care 872.1 99 10.195 9 p < .001
Crime 1743 168 34.697 9 p < .001
Table 3
Pearson Correlation Between Media Coverage and Public Opinions (N=10)
Cumulative Obtrusive: Health Care Unobtrusive: Crime
Weeks NYT/Public ABC/Public NYT/Public ABC/Public
1 .970** .704* .801** .934**
2 .914** .721* .553 .916**
3 .854** .742* .507 .879**
4 .879** .731* .479 .886**
5 .844** .671* .432 .902**
6 .851** .677* .473 .897**
7 .864** .648* .540 .911**
8 .859** .598 .598 .942**
9 .850** .595 .638* .947**
10 .869** .587 .674* .945**
11 .885** .629 .712* .942**
12 .902** .671* .710* .946**
13 .928** .696* .783** .941**
14 .937** .697* .824** .930**
15 .946** .695* .853** .935**
16 .955** .708* .847** .931**
17 .962** .688* .853** .926**
18 .971** .698* .857** .919**
19 .972** .696* .866** .912**
20 .978** .696* .872** .904**
** p < .01
* p < .05
Table 4
Paired Samples t-test between Times/Public Correlations
and ABC/Public Correlations (N=20)
Issue Mean of Correlation t value df Sig (2-tailed)
NYT/Public ABC/Public
Health
Care .9095 .6774 21.518 19 p < .001
Crime .6936 .92225 -7.01 19 p < .001
Appendix 1
Public Opinion Percentage and the New York Times Coverage of Crime and Health
Care Issues
poll year month crime health one one two two three three four
data % care % week(c) week(h) weeks(c)weeks(h) weeks(c) weeks(h)weeks(c)
1 90 11/1 2 1 89 32 195 50 313 72 429
2 91 3/7 2 1 84 30 164 50 244 77 312
3 91 5/23 3 1 71 28 152 65 215 94 293
4 91 11/21 6 6 66 33 140 90 219 139 280
5 92 3/26 5 12 95 39 209 87 305 118 409
6 92 8/28 7 12 89 41 184 79 280 126 381
7 93 1/8 9 18 87 50 167 91 247 144 329
8 93 9/10 16 28 83 72 153 126 261 179 367
9 94 1/28 49 31 119 73 211 116 306 160 413
10 94 7/15 30 21 121 67 243 123 345 199 434
four five five six six seven seven eight eight nine nine ten
weeks(h)weeks(c)weeks(h)weeks(c) weeks(h)weeks(c) weeks(h)weeks(c) weeks(h)weeks
(c) weeks(h)weeks(c)
108 540 136 637 158 737 186 829 212 892 222 961
98 379 133 459 168 537 197 612 222 682 241 737
129 379 170 444 199 543 237 605 260 678 288 747
176 374 218 437 264 507 295 571 332 644 360 718
172 497 215 585 256 685 311 761 363 850 421 941
182 464 214 548 259 632 312 705 346 784 388 883
203 426 254 507 287 600 325 699 364 778 406 875
239 441 303 538 361 621 417 723 474 796 514 889
213 497 262 597 304 725 360 842 412 930 450 1033
257 534 348 637 391 731 451 809 527 916 578 1025
ten eleven eleven twelve twelve thirteen thirteen fourteen fourteen fifteen fift
een sixteen
weeks(h)weeks(c) weeks(h)weeks(c) weeks(h)weeks(c) weeks(h)weeks(c) weeks(h)week
s(c) weeks(h)weeks(c)
234 1050 254 1132 282 1220 324 1302 346 1373 371 1455
282 804 309 879 344 952 372 1035 400 1118 424 1196
314 798 348 882 378 962 398 1042 425 1110 446 1177
391 800 419 866 454 936 484 1020 519 1106 563 1177
471 1017 517 1090 551 1156 587 1230 622 1316 679 1426
424 977 477 1067 521 1147 556 1221 583 1309 621 1410
479 970 529 1089 594 1184 650 1266 695 1337 757 1427
556 977 593 1049 644 1143 704 1221 756 1305 816 1385
512 1156 584 1251 656 1422 734 1559 810 1688 892 1796
631 1122 682 1209 742 1298 795 1398 854 1491 908 1612
sixteen seventn seventn eighteen eighteen nineteen nineteen twenty twenty
weeks(h)weeks(c) weeks(h)weeks(c) weeks(h)weeks(c) weeks(h)weeks(c) weeks(h)
403 1539 428 1606 456 1705 474 1762 493
448 1267 467 1347 496 1436 528 1542 546
481 1257 516 1335 545 1410 570 1480 589
599 1270 633 1334 672 1433 697 1486 719
710 1497 751 1592 787 1658 820 1732 877
659 1508 701 1584 744 1663 782 1774 827
797 1523 847 1589 885 1675 924 1764 965
884 1451 946 1522 1029 1601 1085 1699 1155
967 1909 1054 2000 1178 2105 1270 2205 1351
967 1700 1022 1792 1094 1893 1156 1986 1199
Note: c = crime; h = health care.
Appendix 2
Public Opinion Percentage and ABC Coverage of Crime and Health Care Issues
poll year month crime health one one two two three three four
data % care % week(c) week(h) weeks(c)weeks(h) weeks(c) weeks(h)weeks(c)
1 90 11/1 2 1 6 3 11 5 15 8 19
2 91 3/7 2 1 6 2 13 4 16 5 21
3 91 5/23 3 1 5 2 12 5 18 8 22
4 91 11/21 6 6 11 2 15 7 22 13 25
5 92 3/26 5 12 7 5 15 9 22 15 28
6 92 8/28 7 12 3 3 6 10 11 13 19
7 93 1/8 9 18 6 2 9 4 17 6 23
8 93 9/10 16 28 4 6 16 9 29 22 40
9 94 1/28 49 31 39 26 65 30 79 34 106
10 94 7/15 30 21 24 8 65 21 98 41 132
four five five six six seven seven eight eight nine nine ten
weeks(h)weeks(c)weeks(h)weeks(c) weeks(h)weeks(c) weeks(h)weeks(c) weeks(h)weeks
(c) weeks(h)weeks(c)
11 22 14 25 19 30 22 33 24 37 28 37
6 25 8 35 11 40 11 42 11 48 12 54
13 30 17 34 19 39 22 43 25 48 27 52
19 28 24 31 24 35 24 39 26 45 31 52
18 32 22 39 27 46 35 49 48 53 53 60
19 23 24 27 29 31 36 37 38 42 42 48
11 24 11 30 14 34 18 41 23 45 28 49
29 50 33 61 36 67 39 74 43 79 47 85
41 123 47 132 51 156 56 185 69 210 74 232
54 144 67 156 71 174 86 185 123 203 134 226
ten eleven eleven twelve twelve thirteen thirteen fourteen fourteen fifteen fift
een sixteen
weeks(h)weeks(c) weeks(h)weeks(c) weeks(h)weeks(c) weeks(h)weeks(c) weeks(h)week
s(c) weeks(h)weeks(c)
30 40 30 41 31 42 31 46 33 48 35 51
13 56 13 61 14 64 17 69 18 72 20 76
29 60 29 66 31 73 33 76 34 81 35 85
34 55 34 58 39 60 39 66 42 70 48 76
61 68 65 71 66 77 67 82 70 91 76 97
44 53 47 59 51 66 55 73 58 83 59 89
36 57 41 61 47 67 55 74 58 77 65 80
48 94 50 96 59 103 64 112 69 116 72 121
78 252 92 277 106 300 112 311 120 331 133 346
145 251 153 268 165 297 170 326 183 338 200 359
sixteen seventn seventn eighteen eighteen nineteen nineteen twenty twenty
weeks(h)weeks(c) weeks(h)weeks(c) weeks(h)weeks(c) weeks(h)weeks(c) weeks(h)
36 56 38 60 40 61 44 66 48
21 80 22 86 27 92 30 97 32
37 95 40 100 40 102 40 108 41
50 79 52 84 53 89 55 97 58
82 98 89 103 94 114 96 118 101
61 101 66 112 69 118 74 125 79
72 86 76 93 80 98 84 101 87
80 130 83 135 92 142 97 149 102
133 358 137 366 145 374 162 381 171
205 380 222 398 235 418 258 438 271
Note: c = crime; h = health care.
|
