Content-Type: text/html
Television Credibility Revisited: A Longitudinal Study
Prepared for the Annual AEJMC Conference
Radio/Television Journalism Division
August 9-12, 1995
Washington, D.C.
by Zhang,Kewen
PhD student
School of Journalism
University of Missouri-Columbia, USA
Address: University Terrace 12-F
Columbia, MO 65201
Phone (H): (314)884-3772
Email: [log in to unmask]
Hao, Xiaoming
Nanyang Technological University, Singapore
Address: School of Communication Studies
Nanyang Technological University
Nanyang Avenue, Singapore
Phone: (65) 799-5012 Fax: (65) 791-3082
Email: [log in to unmask]
Television Credibility Revisited:
A Longitudinal Study
(Abstracts)
Based on reliable national survey data, this study challenges the common
belief that television is more trustworthy than the print media,
concluding
that at least since the early 1970s, television has fallen behind the
print media in receiving Americans' confidence rating. It also shows
that
the decline of television credibility over the past 25 years could not
just
be attributed to the emergence of a more critical audience.
Television Credibility Revisited:
A Longitudinal Study
For long in the United States, television has been widely
considered a medium more credible than the print media,
especially in the filed of news reporting. Most studies comparing
television with the print media on the credibility issue came to
conclude that television was more favorably rated (Chang &
Lemert, 1968; Abel & Wirth, 1977; Atkin & Elwood, 1978; Lee,
1978; Reagan & Zenaty, 1979; Gaziano & McGrath, 1985a & 1985b).
Even the American Society of Newspaper Editors admitted that
Americans prefer television over newspapers by a wide margin,
"not only because of its ease but also because they find
television more believable and less biased" (Los Angeles Times,
1985).
The conclusions of these studies, supporting one another,
are so forceful that over the last ten years or so, scholars
seemed to have ceased to explore whether television is really
more trustworthy than the newspaper. To many people, the question
is virtually settled. If a picture is worth a thousand
words, how can the print media compete with television with all
its visual and audio effects?
Most studies comparing television and newspapers on
credibility, however, were based on forced comparison. That is,
they asked respondents to name their preference for a particular
medium. It was suggested that such forced comparison might not
have led to reliable answers because the criteria people used to
judge television credibility could be different from those used
to judge newspapers (Newhagen & Nass, 1989). Past researchers
studying the media credibility issue were probably
right when they concluded that television was considered more
credible in news reporting or under specified conditions.
However, the answer might have been different had the respondents
been asked to express their confidence in the two types of media
separately while using their own criteria for judgment. A
hypothesis like this, nevertheless, has never been tested.
In addition, most of these earlier studies were one-shot
tests. Because different questions were asked and different
criteria employed, they could not be used for comparative
purposes to monitor longitudinal changes although the statistics
produced in these studies tended to show a continuous decline in
media credibility.
Because the fundamental reality of mind is socially based
(Scheibe, 1970), earlier studies found that people with high
education and income tended to be more critical and place less
trust in the media (Einsiedel & Casey, 1979; Gaziano & McGrath,
1985a; Stamm, 1987). Since education and income are generally on
the increase in various societies, we are not sure whether the
continuous decline in the public perception of media credibility
should be attributed to the increasingly more discerning public
alone, or to the media's own problems as well.
This study is a continuation of past efforts to examine
media credibility. Using reliable national survey data, it seeks
to find out if Americans really have more trust in television
than in the print media. It also attempts to answer the
question of whether the decline in television credibility, should
this be found true, could be explained by demographic changes
of American society alone.
Two basic approaches have been employed in past studies
to tackle the issue of media credibility. One focuses on how
internal factors, i.e., factors related to how media present news
and other information to the public, affect the audience's
perception of media credibility (Chang & Lemert, 1968; Culbertson
& Somerick, 1976; Abel & Wirth, 1977; and Gaziano & McGrath,
1985a, etc.). The second approach focuses on various demographic
variables that affect people's perception of media credibility
(Einsiedel & Casey, 1979; Pratt, 1981; and Stamm, 1987, etc.).
This study follows the second approach.
Literature Review
Roughly in the 1960s, when television had gained
increasingly popularity in the United States, communication
scholars and the media industries began interested in comparative
study of television and other media for credibility. A favorite
topic for their studies is which medium fares better in terms of
credibility.
Chang and Lemert (1968) noted the relative anonymity of the
newspaper reporter and described this as a reason for people to
assign greater credibility to television than the newspaper.
Abel & Wirth (1977) said that in terms of local news coverage,
television was perceived to be a more credible, truthful, and
important news source than the newspaper. Reagan & Zenaty (1979)
reached a similar conclusion. Citing their survey results,
Atkins & Elwood (1978) pointed out that high school students
favored television as a news source by a wide margin over
newspapers, radio, and news magazines in three areas: general
preference, believability, and preference should the consumer be
limited to one news source. In a survey of college students, Lee
(1978) found that 76 percent of the polled said they
would believe television news while only 24 percent said they
would believe newspaper news in case of conflicting or different
reports of the same news story.
The advantages of television over the print media in
winning people's confidence were clearly stated by Gaziano &
McGrath (1985b) in a study on attitudes towards the credibility
of individual media and news media in general. They found that
television was chosen by most respondents as the medium (1) they
would believe concerning reports of conflict; (2) they would
choose if they were limited to just one source for local, state,
or national/international news; and (3) they considered most
reliable for local, state, and national/international news.
These studies have one thing in common. That is, they forced
respondents to make a direct comparison of the two different
types of media in researcher-defined situations. As Newhagen and
Nass (1989) noted, "many respondents base their perception of
credibility or confidence in a newspaper on its performance as an
institution, while they base their perception of credibility on
the standards and trustworthiness of television news on the
performance of an aggregate of on-camera personalities . . . .
The comparison thus imposes different criteria and different
levels of analysis on the media consumer to answer the question."
Another area of study on media credibility is the impact of
demographic differences of audiences on their confidence in
various media. In a study of African and U.S. students, Pratt
(1981) attempted to test if those who used media more intensively
were more likely to trust the media, but the hypothesized
positive relationship between media exposure and perceived media
credibility was generally not supported in either sample groups.
Einsiedel and Casey (1979) found that people with less
education tended to select television as the most credible
medium. Gaziano and McGrath (1985a) identified two segments of
the population who were relatively more critical of the media.
One segment is composed of people with relatively high education
and income and greater propensity to act when angered by media
content. The other is made up by people with relatively low
education and income, less knowledge of media, and less
likelihood of taking action when provoked by media content. The
second segment was also identified in a survey report (Miami
Herald, 1985), which noted that less educated people are more
likely to distrust the media. It claimed that at one end of the
social-economic spectrum of people who most distrust the media
are those who are less well-informed and suspicious. These were
people with less education and lower incomes than the majority.
They were said to be bound together by their demographic
characteristics and by alienation and suspicion. A Gallop poll
(Los Angeles Times, 1986) found that critics of the news media
generally exhibit greater knowledge about the press, greater
interest in press issues than those who are supportive. The
majority of critics also share a conservative ideology and a
college education. In his study, Stamm (1987) concluded that
there was a negative correlation between respondents' education
and PSI -- perception of a face-to-face relationship between
spectator and media performer.
In addition to the above-mentioned factors, other variables
were also found to be associated with people's confidence in the
media. A survey (Los Angeles Times, 1993) concluded that blacks,
liberals, women as well as people with less education and lower
incomes generally gave the news media (especially television
news) higher ratings than other groups.
Methods
To re-examine the question of media credibility on a
longitudinal basis, cumulative data gathered in the General
Social Survey (GSS) by the National Opinion Research Center
at the University of Chicago was used in this study.
Since 1972, the GSS has been conducted annually to find out
the attitudes of Americans towards various issues. Each
survey involves about 1,500 interviews. By 1993, the total
number had accumulated to nearly 30,000. Respondents in
each yearly survey are composed of a national cross-section of
non-institutionalized persons over 18 years of age in the
continental United States. In the early years, half of the
samples were selected using full probability techniques and
half via block quota sampling. Samples since 1975 have been
selected with a full probability sampling design. The GSS uses
identical questions each year. Exact question wordings are
available from the General Social Surveys, 1972-1993: Cumulative
Codebook (Davis & Smith, 1993).
Among the hundreds of separate but continuous questions in
this GSS survey, the respondents were asked how much confidence
they had in the press, television, and radio respectively. The
possible answers were "a great deal," "only some" and "hardly
any." Although such questions do not specify particular aspects
of the media, the answers serve as instruments to measure the
overall confidence of respondents in the three types of media.
They were used as dependent variables. Based on findings of
earlier students, various demographic variables that may affect
people's confidence in media were selected as independent
variables. They include the respondent's age, sex, race,
education, income, mobility, occupation, job prestige, work
status, income, party affiliation, television viewing time,
frequency of reading a newspaper and radio listening habit. To
facilitate the statistical test, some of the variables were
recoded.(Note 1) To monitor longitudinal changes, the time in
which the surveys were conduced was recoded into three groups
-- the 1970s, 1980s and 1990s -- to reduce year-by-year sampling
errors and mark the three decades.
The dependent and independent variables were crosstabulated,
with possible contaminating variables under control, to test
their relationships. In testing the strength of association,
different statistics were used to measure the direction and
strength of association: chi-square (Mantel-Haenszel) for
significance; Phi for variables on dichotomous, nominal scales;
coefficient of contingency for variables on non-dichotomous
nominal scales; and Gamma for variables on ordinal scales.
Findings
Contrary to the findings of many previous studies, the
results of a crosstabulation between confidence in media and
the time of interview show that people do not have greater
confidence in television than in the press.
The results indicate an overall decline in people's
confidence in both television and the press. The proportion
of respondents who have a great deal of confidence has
decreased over the years while more and more people said
they had hardly any confidence in either television or the
press.
Although there were ups and downs in the percentage of
people who said they had a great deal of confidence in both
types of media, people were more likely to say they had
a great deal of confidence in the press than in television.
At the same time, they were more likely to say that they had
hardly any confidence in television than in the press. The
results show that people's confidence in television has
fallen behind that in the print media since the 1970s. The
only exception was found in the year of 1993 when television
fared a little better than the press in getting trusted, but that
was mainly the result of a decline in people's confidence in the
press rather than a gain of television in obtaining people's
trust. See Table 1 for details.
Table 1
Confidence in the Press and TV (1973 to 1993)
__________________________________________________________
Year Media Great Deal(%) Only Some(%) Hardly Any(%)
__________________________________________________________
93 press 11.0 49.6 39.3
TV 11.7 51.1 37.2
91 press 16.7 54.9 28.4
TV 14.5 55.0 30.4
90 Press 15.2 59.5 25.3
TV 13.9 58.9 27.2
89 Press 17.1 55.5 27.4
TV 14.3 56.3 29.5
88 Press 18.9 55.0 26.0
TV 14.4 58.9 26.7
87 Press 19.3 57.4 23.2
TV 12.7 59.7 27.5
86 Press 18.6 55.4 25.9
TV 15.2 56.6 28.2
84 Press 17.3 59.9 22.8
TV 13.4 57.8 28.8
83 press 13.7 62.3 24.0
TV 12.7 58.6 28.7
82 press 18.3 61.6 20.2
TV 15.5 59.3 25.2
80 Press 22.6 59.6 17.8
TV 16.3 55.5 28.2
78 Press 20.5 59.5 20.1
TV 14.0 54.4 31.6
77 Press 25.7 58.5 15.8
TV 17.7 56.8 25.5
76 Press 29.0 53.0 18.0
TV 19.1 53.2 27.7
75 Press 24.5 57.1 18.4
TV 18.3 58.8 22.9
74 Press 26.2 56.1 17.7
TV 23.7 58.8 17.5
73 Press 23.4 61.7 14.9
TV 18.8 59.1 22.1
____________________________________________________________
The results of a crosstabulation of time and confidence
in the media give a clearer picture of the decline of people's
confidence in both the press and television. As can be seen from
Table 2, people who had a great deal of confidence in the press
decreased from 24.6 percent in 1970s to 18.3 percent in the
1980s and then to 14.2 percent in the 1990s. At the same time,
people who had a great deal of confidence in television decreased
from 18.6 percent to 14.3 percent and 13.3 percent. For both the
press and television, people who had hardly any confidence in
them increased during the three periods.
Table 2
Changes in Confidence in Press and TV over Three Periods
_____________________________________________________
Decades 70s(%) 80s(%) 90s(%)
_____________________________________________________
Great (Press) 24.6 18.3 14.2
deal (TV ) 18.6 14.3 13.3
Only (Press) 57.7 58.7 54.4
Some (TV) 56.8 58.0 54.8
Hardly (Press) 17.5 23.1 31.4
any (TV ) 24.6 27.7 31.9
Gamma=.11 p=.000
______________________________________________________
To re-examine which demographic variables are
associated with people's confidence in television, all the
demographic variables selected were crosstabulated with
people's confidence in television. The data gathered over twenty-
one years have given the researchers greater confidence in
identifying variables that might affect people's confidence in
television. See Table 3 for results.
Table 3
Relations between Confidence in Television and Various Variables
___________________________________________________________
Confidence Great Only Hardly Stat. sig.
in TV Deal(%) Some(%) Any(%)
___________________________________________________________
Time Spent
Watching TV
Excessive
Viewers 27.5 57.4 15.1
Heavy Viewers 17.9 59.7 22.4
Light
Viewers 11.2 55.7 33.1
Nonviewers 9.9 40.5 49.5 Gamma=.28 p=.000
Education
17-20 years 8.4 56.0 35.5
13-16 years 11.8 57.5 30.7
7-12 years 18.1 57.5 24.4
0-6 years 25.8 50.9 23.3 Gamma=-.19 p=.000
Income
Over $25K 8.2 58.3 33.5
Up to $25K 13.2 55.5 31.3
Up to $20K 13.7 57.9 28.4
Up to $15K 15.0 59.7 25.3
Under $10K 17.1 57.7 25.2 Gamma=-.17 p=.000
Job Prestige
70-89 11.8 53.1 35.1
50-69 10.6 59.5 29.9
30-49 15.3 58.0 26.8
10-29 21.3 55.7 23.0 Gamma=-.14 p=.000
Newspaper
Reading Habit
Very Often 13.7 58.8 27.5
Sometimes 16.6 55.1 28.3
Never 22.8 50.6 26.6 Gamma=.05 p=.000
Mobility (Note 2)
Same City 17.2 57.9 24.9
Same State 14.9 56.6 28.5
Different
State 14.5 56.6 28.9 Gamma=.07 p=.000
Race
White 14.9 56.9 28.3
Black 20.9 59.0 20.1 Contingency
Other 21.9 55.2 22.9 Coefficient p=.000
=.80
Work Status
Employed 14.2 57.8 28.0
Jobless 19.5 58.6 54.0
Retired 19.1 58.6 54.0 Contingency
At Home 17.4 57.0 25.6 Coefficient p=.000
=.06
Sex
Male 16.9 55.1 28.0
female 15.0 58.7 26.2 Phi=.04 p=.924
Age
18-19 25.8 54.7 19.6
20-39 15.4 57.6 27.0
40-59 14.5 57.6 28.0
60 and above 17.5 55.9 26.6 Gamma=.002 p=.931
_________________________________________________________
Of the various variables selected, four variables were
found to be worthy of consideration on the basis of
statistical values obtained. The amount of time
people normally spend watching television is positively
related to their confidence in television. People who spent
more time watching television tended to say that they had a
great deal of confidence in television than people who
watched less. At the same time, people who watched less
television were more likely to say they had hardly any
confidence in the medium.
Education is negatively related to confidence in
television. The more education a person has received, the
more likely for that person to say that he has little
confidence in television. Only 8.4 percent of the people
who have received 17 to 20 years of education said they
had a great deal of confidence in television while 35.5
percent of them said they had hardly any confidence in
television. As the educational level increased, the chance
for people to say they had a great deal of confidence
decreased.
The same kind of relationships were found between the
respondents' income and confidence in television as well as
between respondents' job prestige and confidence in
television. People with higher income and job prestige
tended to show less confidence in television while people
with low income and job prestige were more likely to show
stronger confidence in television.
To monitor changes significantly related to people's
confidence in television, these four variables were
crosstabulated with the three time periods. Results are
presented in Table 4.
Table 4
Changes of TV Viewing Habits, Education, Income
and Job Prestige in Three Decades
___________________________________________________________
Variables 70s(%) 80s(%) 90s(%) Stat. Significance
___________________________________________________________
Income
Under $10K 60.7 35.2 25.0
Up to $15K 20.4 17.8 12.9
Up to $20K 9.8 13.2 13.4
Up to $25K 4.4 11.7 12.2
Over $25K 4.6 22.0 36.5 Gamma=.43 p=.000
Education
0-6 6.1 4.2 2.6
7-12 63.0 56.8 49.3
13-16 25.1 30.8 37.8
17-20 5.8 8.1 10.3 Gamma=.19 p=.000
Television
Viewing Habits
Nonviewers 4.6 4.9 3.6
Light
Viewers 46.1 43.9 47.4
Heavy
Viewers 44.6 45.2 43.9
Excessive
viewers 4.8 5.9 5.1 Gamma=.01 p=.379
Prestige
10-29 26.4 24.0 21.1
30-49 52.5 50.3 50.8
50-69 19.3 23.3 24.2
70-89 1.8 2.5 3.9 Gamma=.09 p=.000
____________________________________________________________
Although the average hours people spent watching
television over the years had increased slightly (Hao,
1994), no significant changes took place in the categories
of television viewing habits listed here. Therefore, the
possibility that changes in people's confidence in
television over the years could be accounted by changes in
people's television viewing habits can be dismissed.
Because job prestige and income can be highly
associated with people's education, their relationships with
education were first examined. The results show that when
the educational factor is under control, people with higher
income tend to trust television less, excluding the
influence of education. People who have received the same
level of education differ in their attitudes towards
television if they differ in income. People with higher
income tend to have less trust in television. The only
exception was found among people who have received
only primary school education or less. See Table 5 for details.
Table 5
Crosstabulation of
Confidence and Income with Education under Control
____________________________________________________________
Confidence in TV
Income Great Only Hardly Stat. Sig.
Deal(%) Some(%) Any(%)
____________________________________________________________
(Education=0-6 years)
Over 25K 22.0 53.7 24.4
Up to 25K 25.9 59.3 14.8
Up to 20K 14.0 58.0 28.0
Up to 15K 26.3 51.8 21.9
Under 10K 27.0 50.5 22.5 Gamma=.06 p=.379
(Education=7-12 years)
Over 25K 11.5 60.1 28.5
Up to 25K 15.5 59.0 25.5
Up to 20K 15.5 60.0 24.4
Up to 15K 18.7 57.7 23.5
Under 10K 23.3 55.2 21.5 Gamma=.14 p=.000
(Education=13-16 years)
Over 25K 8.4 57.9 33.6
Up to 25K 11.1 57.9 31.0
Up to 20K 13.5 57.2 29.3
Up to 15K 15.1 57.1 27.8
Under 10K 16.1 55.1 28.8 Gamma=.11 p=.000
(Education=17-20 years)
Over 25K 6.4 55.5 38.2
Up to 25K 10.6 53.7 35.6
Up to 20K 8.9 60.7 30.4
Up to 15K 6.1 70.6 23.3
Under 10K 13.7 50.7 35.6 Gamma=.10 p=.001
____________________________________________________________
The crosstabulation of people's confidence in television and
job prestige with education under control produced inconsistent
results. While the overall pattern shows that people who have
the same educational level but hold jobs of higher prestige tend
to have less confidence in television, it is not true for all
categories of people in terms of education. Such an inconsistent
pattern makes job prestige a questionable factor in explaining
changes in people's confidence in television. Please see Table 6
for details.
Table 6
Crosstabulation of Confidence
in TV and Income with Education under Control
____________________________________________________________
Confidence in TV
Job Prestige Great Only Hardly Stat. Sig.
Deal(%) Some(%) Any(%)
____________________________________________________________
(Education=0-6 years)
70-89 00.0 00.0 00.0
50-69 39.3 50.0 10.7
30-49 22.4 53.2 24.4
10-29 27.8 49.7 22.5 Gamma=.04 p=.825
(Education=7-12 years)
70-89 5.4 67.9 26.8
50-69 12.3 62.6 25.1
30-49 16.8 58.1 25.1
10-29 21.1 56.4 22.5 Gamma=.10 p=.000
(Education=13-16 years)
70-89 10.6 55.8 33.6
50-69 10.8 58.0 31.2
30-49 11.8 58.0 31.2
10-29 17.3 56.9 25.9 Gamma=.07 p=.0003
(Education=17-20 years)
70-89 13.8 48.6 37.7
50-69 7.0 59.7 33.4
30-49 7.5 59.3 33.2
10-29 23.9 47.8 28.3 Gamma=.03 p=.362
____________________________________________________________
Because both people's income and education increased
over the years, their influence must be excluded in order to
find out if the overall decrease in people's confidence in
television could be solely explained by changes in their
education and income. To accomplish this, people's
confidence in television was again crosstabulated with the
three periods covered, with educational and income
differences under control.
With personal income under control, not much significant
change was found in people's confidence in television over the
three time periods. Among those earning $25,000 a year or less,
there were no significant changes. Among those earning more than
$25,000, a significant relationship was found but such a
relationship was neither strong nor consistent. Although there
was a consistent increase in the percentage of people who had
hardly any confidence in television, the proportion of people who
had a great deal of confidence dropped from 12.5 percent in the
1970s to 8.9 percent in the 1980s and then increased slightly to
9.3 percent in the 1990s. See Table 7 for details.
Table 7
Changes of TV Confidence
during Three Periods with Income under Control
___________________________________________________________
Confidence 70s(%) 80s(%) 90s(%) Stat. Sig.
in TV
___________________________________________________________
(Income=Under 10K)
Great Deal 23.1 21.1 20.1
Only Some 54.3 55.2 53.4
Hardly Any 22.6 23.7 26.5 Gamma=.05 p=.017
(Income=Up to 15K)
Great Deal 17.5 17.7 16.6
Only Some 58.8 57.3 55.4
Hardly Any 23.7 25.0 28.0 Gamma=.03 p=.272
(Income=Up to 20K)
Great Deal 14.1 14.4 16.3
Only Some 61.5 57.6 55.2
Hardly Any 24.4 28.0 28.5 Gamma=.04 p=.175
(Income=Up to 25K)
Great Deal 12.8 13.7 16.7
Only Some 55.1 60.9 55.1
Hardly Any 32.1 25.4 28.2 Gamma=-.09 p=.015
(Income=Over 25K)
Great Deal 12.5 8.9 9.3
Only Some 58.9 59.6 54.9
Hardly Any 28.7 31.5 35.8 Gamma=.09 p=.000
____________________________________________________________
Next, the relationship between people's confidence and
the passage of time was examined with education under
control. No significant changes were found among people
with only primary school education or less, but people with
higher education showed changes in their attitude towards
television. Even when the education factor was held
constant, a decline was still found in people's confidence in
television. See Table 8 for details.
Table 8
Changes of Confidence in TV over Three Periods
with Education under Control
___________________________________________________________
Confidence 70s(%) 80s(%) 90s(%) Stat. Sig.
in TV
___________________________________________________________
(Education=0-6 years)
Great Deal 30.0 22.9 14.3
Only Some 47.8 52.8 61.4
Hardly Any 22.3 24.3 24.3 Gamma=.12 p=.012
(Education=7-12 years)
Great Deal 20.3 16.3 17.5
Only Some 57.1 58.7 54.1
Hardly Any 22.6 25.0 28.4 Gamma=.08 p=.000
(Education=13-16 years)
Great Deal 13.4 11.4 9.8
Only Some 57.4 57.9 56.5
Hardly Any 29.2 15.4 5.6 Gamma=.07 p=.000
(Education=17-20 years)
Great Deal 11.6 7.2 6.3
Only Some 60.0 55.5 50.3
Hardly Any 28.3 37.3 43.4 Gamma=.19 p=.000
____________________________________________________________
Discussion
The results of this study challenge the popular belief that
television is a more credible medium than the print, which was
supported by many past studies. However, this does not mean that
past researchers were all wrong in their conclusions. The
explanation for the difference is likely to lie in the criteria
used for judgment rather than fundamental differences in the
samples used.
As mentioned earlier, different researchers examining
media credibility issue tend to use different criteria for
judgment, which are implied in the way the questions are phrased
or the specified conditions under which such comparison is made.
For example, Abel and Wirth (1977) compared the two media
in terms of local news coverage, Lee (1978) focused on
conflicting or
different reports of the same story, and Gaziano and McGrath
(1985b) focused on news reports of conflict, the sole source
and reliability for news. In the General Social Surveys,
however, the questions regarding confidence in the media were
asked in a general manner and without specific conditions. The
respondents were left free to use their own criteria to
rate different media separately. While past researchers
were probably right in concluding that people sometimes trust TV
news more than newspaper reporting, that could not be taken as an
answer to the question which medium is more trustworthy when
their overall performance is compared.
If we ignore the argument about how to define the term
"credibility" and under what conditions the two media should be
compared, the fact that more and more people are reluctant to say
that they have a great deal of confidence in television deserves
attention from both the television industry and communication
scholars. The decline of people's confidence in television as a
mass communication channel both over the years and the three
time periods has made it hard to deny that television faces a
credibility problem as serious as, if not more than, what the
print media have to deal with.
Although Americans' confidence in television and the
press has experienced ups and downs over the years, the
overall comparison shows that Americans trust television
less than the press. The gap between television and the
press in terms of people's confidence has certainly
been reduced in the 1990s compared with the 1970s and 1980s,
but that was a result of the more drastic decline of press
credibility rather than a gain by television in its
believability. That television has been quickly catching up with
the press in winning people's trust cannot hide the fact the
proportion of people who have had a great deal of confidence in
television has dropped from 18.6 percent in the 1970s to
13.3 percent in the 1990s while the percentage of people
who have had hardly any confidence in television has
increased from 24.6 percent to 31.9 percent during the same
period.
As mentioned earlier, the issue of media credibility
has been tackled with two different approaches -- focusing
on the performance of the media themselves and focusing on
the changing audience. There is no denying that the decline of
media credibility can be a result of the changing media
themselves. An irresponsible press and unethical practices
by journalists can contribute to people's distrust of the
media. A report commissioned by the American Society of
Newspaper Editors (Chicago Tribune, 1985), for instance,
listed some key aspects of the credibility problem. These
include failure to show enough concern for how ordinary
people might be hurt by news coverage, "personal biases of
reporters shown up in their news stories; reporters being rude
and having patronizing attitudes," and "the news media giving
more coverage to stories that support their point of view."
On the other hand, the demographic changes on the part of
the audience may also affect their perception of media
credibility even though the quality of mass media remains more or
less unchanged. By analyzing the GSS data, this study confirms
what has been suggested in earlier studies, i.e., people of
varying characteristics tended to place different amounts of
trust in the media. Of the various demographic variables covered
by the GSS, the amount of television viewing was found positively
related to viewers' confidence in television while education,
income and job prestige were found to be negatively related to
their confidence in television. The longitudinal nature of the
GSS data lends more support to the social categories perspective
in explaining public reaction to the media.
Because it was found that the amount of time people
spent watching television has not changed significantly, the
possibility that it may have a significant impact on the
changes in television credibility over the years can be
excluded. In addition, if people's television viewing time
had increased, it should have boosted rather than reduced
people's confidence in television. Therefore, changes in
people's television viewing habits should not be used to
account for the decline of people's trust in television during
the three time periods.
The remaining three variables, education, income and
job prestige are all negatively related to people's
confidence in television. People with better education,
greater income and higher job prestige tend to be more critical
and place less trust in the media. However, the three variables
are not totally independent of each other. People with higher
education tend to have a greater chance to earn more and
hold more prestigious jobs. Job prestige appears to have little
impact on people's confidence in television with education under
control. This indicates that the direct relationship found
between confidence in television and job prestige might have been
contaminated by education and other variables which better
indicate the critical ability of the audience.
Income as a variable has shown a strong relationship with
people's confidence in television even with educational levels
under control, indicating that people's income has a direct or
indirect relationship with their confidence in television. While
a causal relationship between income and confidence in television
cannot be inferred from the results, income serves well as an
indicator of the critical ability of the audience. It is
possible that people with higher income tend to enjoy a wider
range of information sources to double check what they learn from
the media, or they can be better informed or entertained through
other means not available to people with lower income. Such
hypotheses, however, could not be tested with the data used for
this study.
That education influences a person's perception of media
credibility has become more or less an established fact.
Because this analysis of GSS data has found that Americans'
education and income have significantly increased in the past
two decades, it suggests that television's decrease in its
credibility rating is partially due to the changes of these
two factors.
To see if the decline in people's trust in television could
also be attributed to television's own performance, demographic
changes in the audience that may lead to their distrust of
television must be excluded. Although no strong significant
changes were found in the respondents' confidence in television
over the years when the factor of income was controlled, the
decline of television credibility could still be demonstrated
when the effects of education were controlled. That shows the
decline of confidence in television cannot be attributed to the
increase in Americans' critical ability alone. Internal factors,
i.e., changes in television's performance, as well as changes in
the audience, should be explored to account for the decline in
television's credibility ratings.
In conclusion, this analysis of representative longitudinal
national survey data posed a serious challenge to the common
belief that television is more trusted by Americans than
the print media. Although television may enjoy some advantage in
reporting news over the newspaper, it actually does not enjoy as
much trust as the newspaper when the overall performance is
evaluated. The credibility of television is also declining and
such a decline cannot be simply explained by the increasingly
more discerning American public alone.
Notes
(1) Television viewing time, ranging from 0 to 24 hours,
was recoded into four categories: non-viewers, those who
claimed they did not watch TV at all; light viewers, those
who watched television between one and two hours a day;
heavy viewers, those who watched TV between three and
six hours a day; excessive viewers, those who watched TV
more than seven hours a day. Job prestige was recoded into
four categories: 10 through 29, 30 through 49, 50 through
69, and 70 through 89. Education, ranging from 0 to 20
years, was recoded into four categories: 0-6 years, 7-12
years, 13-16 years and 17 to 20 years. Personal income,
originally classified into 12 categories, was recoded into
five: under $10,000, $10,001 to $15,000, $15,001 to
$20,000; $20,001 to $25,000, and over $25,000. The
frequency of reading a newspaper was recoded
into three categories: very often (every day or a few times
a week), sometimes (once or less than once a week), and
never. Work status was recoded into four categories: employed,
jobless, retired and homemaker.
(2) For mobility, respondents were asked if, by the time they
were 16 years old, they had lived in the same city, different
city in the same state, or different state in the same country.
REFERENCES
Abel, John D. & Wirth, Michael O.(1977). Newspaper vs. TV
Credibility for Local News. Journalism Quarterly, 54(2), 371-75.
Anderson, James A. (1971, April) An analysis of the
methodologies used in media credibility study or: The medium
they believe depends on the question you ask. Paper
presented at the annual conference of International
Communication Association, Phoenix, Arizona.
Atkins, Paul A. & Elwood, Harry. (1978). TV News Is First Choice
in Survey of High Schools. Journalism Quarterly, 55 (3), 596-99.
Chang, Lawrence K.& Lemert, James B. (1968). The invisible
newsman and other factors in media competition. Journalism
Quarterly, 45(3), 436-444.
Chicago Tribune, April 13 (1985), News media rate low in poll.
C. 4.
Culbertson, Hugh M. & Somerick, Nancy (1976). Cloaked
attribution--what does it mean to news readers? ANPA News
Research Bulletin (No.1). Washington D.C.: American
Newspaper Publishers Association.
David, James Allan & Smith, Tom W. (1993). General Social
Surveys, 1972-1993: Cumulative Codebook. The Roper Center for
Public Opinion Research, University of Connecticut, distributor.
Einsiedel, E. F.& Casey, William (1979, August). Some
correlates of media political advertising credibility.
Paper presented at the annual meeting of the Association
for Education in Journalism, Houston, Texas.
Gaziano, Cecilie & McGrath, Kristin (1985a, November). Media
Publics and Media Trust. Paper presented at the 11th
Conference of the Midwest Association for Public Opinion
Research, Chicago, IL.
Gaziano, Cecilie & McGrath, Kristin (1985b, May). The media
credibility problem: Putting the research into perspective.
Paper presented at the 40th Conference of the American
Association for Public Opinion Research, McAfee, NJ.
Hao, Xiaoming (1994). Television viewing among American
adults in the 1990s. Journal of Broadcasting & Electronic
Media, 38(3), 353-360.
Lee, Raymond S. H. (1978) Credibility of newspaper and TV
news, Journalism Quarterly, 38(3): 282-87.
Los Angeles Times, April 13 (1985). 75% in editors' study doubt
credibility of newspapers. 14.
Los Angeles Times, January 16, (1986) Serious reservations on
fairness are cited; poll finds most American believe Press. 1.
Los Angeles Times, March 31, (1993) Poll delivers bad news to the
media. 16.
Miami Herald, April 14 (1985) Report Defines Newspapers' task in
Earning Trust, 4E.
Morgan, Harmon Loyd (1975). The medium and mass
media credibility: A study of the relationships among
apparent credibility, the medium as an approximation of
reality, and the level of priority of the credibility
dimension. Ph.D. Dissertation, Southern Illinois University,
Ann Arbor, Michigan: University Microfilms.
Newhagen, John & Nass, Clifford. (1989). Differential
criteria for evaluating credibility of newspapers and TV
news. Journalism Quarterly, 66(2): 277-84.
Pratt, Cornelius (1981) Mass media credibility: A study of
African and U.S. Students. Ph.D. Dissertation, University of
Minnesota. Dissertation Abstracts International.
Reagan, Joey & Zenaty, Jayne (1979) Local news credibility:
newspapers vs. TV revisited. Journalism Quarterly, 56(1),
168-72.
Scheibe, Karl E. (1970) Beliefs and Values. New York: Holt,
Rinhart and Winston, Inc.
Stamm, Keith R. (1987, August). Characteristics of loyal and
non-loyal television newscast viewers. Paper presented at
the annual meeting of the Association for Education in
Journalism and Mass Communication, San Antonio, TX.