Gratification Discrepancies and News Program Choice

Communication
Research
http://crx.sagepub.com/

Gratification Discrepancies and News Program Choice
Philip Palmgreen, Lawrence A. Wenner and J. D. Rayburn II
Communication Research 1981 8: 451
DOI: 10.1177/009365028100800404
The online version of this article can be found at:
http://crx.sagepub.com/content/8/4/451
Published by:
http://www.sagepublications.com

Additional services and information for Communication Research can be found at:
Email Alerts: http://crx.sagepub.com/cgi/alerts
Subscriptions: http://crx.sagepub.com/subscriptions
Reprints: http://www.sagepub.com/journalsReprints.nav
Permissions: http://www.sagepub.com/journalsPermissions.nav
Citations: http://crx.sagepub.com/content/8/4/451.refs.html

>> Version of Record - Oct 1, 1981

What is This?

Downloaded from crx.sagepub.com by guest on April 2, 2012

This study examined gratifications sought and discrepancies in gratifications obtained from network news programs in an attempt to predict news
program choice. The results of both 1-test and discriminant analyses indicate that the decision to view a particular television news program is
strongly related to the perception of gratifications obtained (or potentially
available from) the various programs. A comparison of means indicates that
such choice behavior is not a function of overallperceptions of gratifications
obtained from oneS favorite program alone. nor is it usually a function of
gratifications sought from television news in general. It is only when an
audience member makes a comparison between the gratifications perceived to be obtained from his favorite program and competing programs
that a functional relationship with viewing choice emerges. While in certain
cases more specific perceptions of anchorpersons, program format. and
news quality may be more important, the discriminant analyses indicate
that the perception of differentialgratifications is at least as strongly related
to viewing behavior as the more traditionalmeasures of program attributes.

GRATIFICATION DISCREPANCIES
AND NEWS PROGRAM CHOICE

PHILIP PALMGREEN
University o f Kentucky

zy

zy

zyxw

LAWRENCE A. WENNER
Loyola Marymount University

J. D. RAYBURN II
University of Kentucky

In recent years t h e uses and gratifications approach to
t h e study of media has become one of the dominant theoretical frameworks for inquiry into audience orientations. As
the approach has gained a hearty acceptance by its adherents, it has at the s a m e time come under some serious
criticism for structural, theoretical, and methodological weaknesses. One of the most important criticisms of t h e uses
and gratifications approach has been the general failure by

researchers to consider or measure differences between
what audience members are seeking and what they have
received from any given media experience (Lometti et al.,

zyxwvu
zyxwv

COMMUNICATION RESEARCH, Vol. 8, No. 4, October 1981 451-478
Sage Publications, Inc.

0 1981

451

Downloaded from crx.sagepub.com by guest on April 2, 2012

452

zyxwvutsr
COMMUNICATION RESEARCH / OCTOBER 1981


1977). As a number of researchers have noted, it is becoming increasingly clear that there are both conceptual'and
empirical differences between gratifications sought and
gratifications obtained (Katz et al., 1974; Greenberg, 1974;
Lometti et al., 1977).
While some recent studies have begun to address this
issue more directly (Palmgreen and Rayburn, 1979; Palmgreen et al., 1980; McLeod et al., 1977), findings concerning the relationships among gratifications sought and gratifications obtained are both preliminary and tentative at this
point. Nonetheless, they are crucial to further theoretical
development and practical application of the uses and gratifications model. As McLeod et al. (1980) have recently suggested, a situation remains where "the t w o concepts are
confounded in many studies by the limitations of design and
ambiguous question wording. The separation w i t h appropriate research strategies is important if the uses and gratifications approach is to be used more widely in media evaluation research. Independent measurement of gratifications
sought and received is also crucial t o the development of an
adequate theoretical understanding of the role of motivation
in media effects" (p. 1).
Recently Palmgreen and Rayburn (1979) incorporated
measures of both sought and obtained gratifications in a
study of exposure to public television. Their model, which
took into account the discrepancy between gratifications
sought and obtained, successfully discriminated between
viewers and nonviewers of public television across a range

of gratifications. In addition, among those w h o ordinarily
made their o w n decisions concerning which programs to
watch, the discrepancy measure emerged as the second
strongest predictor of public television viewing-stronger,
in fact, than a number of traditional demographic correlates
of such viewing.
The Palmgreen and Rayburn model, however, deals specifically with gross media exposure and not program choice
(although its logic could be extended to choice behavior). A
problem also may arise in implementing the model in pro-

Downloaded from crx.sagepub.com by guest on April 2, 2012

z

zyxw
zyxw
zy

Palmgreen et al. / NEWS PROGRAMS


453

gram choice situations if t h e levels of abstraction chosen
result in generally higher mean levels of gratifications
obtained than gratifications sought. As Palmgreen and Rayburn point out, it is difficult to separate empirically gratifications sought from a particular medium (or content type,
program, and the like) from gratifications obtained a t the
s a m e level of abstraction. To use their example, if we
attempt to obtain measures of gratifications sought and
obtained from "television" (at t h e s a m e level of abstraction),
it may be impossible for respondents to make a meaningful
distinction. But when t h e level of abstraction of t h e obtained
measures is shifted to a component of t h e medium (e.g.
content type) under consideration at the sought level, the
problem is greatly reduced. W e applied this principle in t h e
present study by obtaining measures of gratifications sought
from "television news" in general, and measures of gratifications obtained from specific network news programs.'
In the Palmgreen and Rayburn study of public television,
mean levels of gratification sought from television in general were generally higher than t h e mean gratifications
obtained from public television by both viewers and nonviewers of P N . This is probably true due to stereotypically
negative perceptions of public television shared even by

many viewers of the medium. Viewers, however, were generally more satisfied with P N (had higher mean levels of
gratifications obtained) than. nonviewers, and thus had a
lower mean GS-GOdiscrepancy level. This is in accordance
with t h e assumptions underlying t h e model.
In the study reported here, on t h e other hand, levels of
gratifications obtained from respondents' favorite network
TV news programs were almost uniformly higher across
gratifications than the gratification levels sought from television news in general. One explanation for this finding is
that' the measures of gratifications sought had as their
referent television news in general, including both national
and local news programs, documentaries, news magazines,
and so on. Viewers' general expectations concerning this
aggregate of news program types might well be expected to

zyxwv
Downloaded from crx.sagepub.com by guest on April 2, 2012

454

zyxwv

zyxw

COMMUNICATION RESEARCH / OCTOBER 1981

be lower, on t h e average, than their perceptions of gratifications to be obtained from their favorite network news program. If, then ( a s t h e assumptions underlying t h e uses and
gratifications approach would predict), respondents perceive they are obtaining higher levels of gratifications from
their favorite program than they a r e receiving (or would
receive) from competing programs, we are faced with a
situation where the GS-GO discrepancy would actually be
greater for respondents' favorite programs than for other
programs. This contratheoretical result is directly traceable
to the levels of abstraction chosen and to the particular
medium-component combination under consideration. Unfortunately, a s previously noted, the choice of differing levels of abstraction is forced upon u s if we are to succeed in
a n attempt to distinguish gratifications sought from those
obtained.
In the situation where we wish to compare gratifications
within program content types, the gratification-seeking behavior is more abstract than t h e gratification obtained from
t h e program chosen for viewing. The abstraction level we
use in constructing our gratifications sought is a general
one; it pertains to our seeking behaviors with regard to all

programs of that type. Our cognitive criteria for evaluating
gratifications sought for a certain program type remain constant across all programs of that type. Thus, in situations
where we are investigating programs of a similar content
type, and where mean levels of gratifications obtained are
likely to exceed mean levels of gratifications sought, it is
preferable to use a discrepancy model which more clearly
contrasts the gratifications which might be obtained from
the programs being evaluated. The general form of such a
model is:

zyxw
zy

n

c = (f) c

zyx
zyxwvuts
[ G S , - GO(I),l - [ G S , - G O ( l l ) i ]

n

i=l

GO(ll)i - GO(I)i
= (f)

2
i= 1

n

Downloaded from crx.sagepub.com by guest on April 2, 2012

1

zy
zy

Palmgreen et al. / NEWS PROGRAMS


where:

455

C = a choice between (or among) programs, newspapers, magazines, and so on of a similar content type;

GO(I), = a measure of the extent to which the ithgratification i s

zy
zy

perceived to be obtained from content type I, program I, channel
I, and so on;

GO(ll), = a measure'of the extent to which the ithgratification i s
perceived to be obtained from content type II, program II, channel
I I , and so on;
n = the number of gratifications under consideration; and

GS, = a measure of the extent to which the ithgratification i s

sought from the relevant content type, medium, channel, and so on.

The equation states that program choice (or similar choice)
is a function of t h e average discrepancy between t h e gratifications t h e audience member perceives h e would obtain in
watching a particular program, a n d s o m e similar program(s).
Since t h e gratifications sought from a given program type
are general o n e s which underlie t h e seeking behaviors for a
range of programs within that type, they are considered in
t h e preliminary conceptualization of t h e model; but because
they a r e t h e s a m e with regard t o similar programs, they act
a s constants which cancel e a c h other w h e n t h e model is
algebraically reduced. Thus, while t h e consideration of gratifications sought is crucial t o t h e audience members' evaluation of gratifications obtained, it serves a n orientational
but not computational role within t h e model.
The model is especially pertinent in any study which
seeks t o understand h o w differences in gratifications obtained from similar or competing programs influence t h e
decision t o watch o n e program over another. Following
McGuire's (1 974) reasoning based on learning theory, we
would expect that w h e n a positive discrepancy between
Program II and Program I increases, t h e r e is a corresponding increase in t h e likelihood of a n audience member deciding t o watch Program II.
While in this study we u s e t h e model to investigate h o w
t h e s e differences affect viewing of o n e of t h e three competDownloaded from crx.sagepub.com by guest on April 2, 2012

456

zyxwv
zyxw

COMMUNICATION RESEARCH / OCTOBER 1981

ing television network news programs, the model could be
used to gain a n understanding of audience choice in othei'
similar content-type decisions on television, such a s network morning shows ("Today" versus "Good Morning America"), network panel interview shows ("Meet the Press"
versus "Face the Nation" versus "Issues and Answers"), or
network magazine news shows ("60 Minutes" versus
"20/20"). In addition, the model could be valuable in describing the role of gratifications in the decision to read one
of two or more competing newspapers, or to read one news
magazine rather than another (Time versus Newsweek versus U.S. News and World Report).

z
zyxw
zyx

HYPOTHESES

W e hypothesized that there would be no differences in
the levels of gratifications sought by viewers of the three
network news programs. We assumed that the seeking of
gratifications from television news a s a conteht"type would
be influenced by rather general expectations abstracted
from experiences with network TV news, local TV news,
documentaries, news magazines, and the like. While these
expectations may be more strongly influenced in some cases
by the audience member's favorite network news program
(Palmgreen et al., 1980), we did not feel that perceptions of
the availability of certain gratifications from each respondent's favorite program would be influenced sufficiently by
program preference a s to cause noticeable discrepancies
in gratifications sought by viewers of ABC, NBC, and CBS.
Nor w a s it felt that demographic differences among the
three viewer types in our sample were great enough to
foster different need hierarchies.
W e also hypothesized that no differences would be found
by program preference in mean levels of gratifications
obtained from each respondent's favorite program. For
example, our hypothesis states that viewers of CBS would

Downloaded from crx.sagepub.com by guest on April 2, 2012

zyxw
zy
z

zyxwvu
Palmgreen et al. / NEWS PROGRAMS

457

perceive themselves to be as entertained by the CBS Evening News as ABC viewers felt themselves to be entertained
by ABC World News Tonight. Even though current critical
opinion holds that the ABC format is more "entertainment"
oriented, our assumption was that entertainment (or beauty)
is in the eye of the beholder. Thus different program attributes would be perceived as entertaining by viewers of ABC,
CBS, and NBC, leading to equal satisfaction on this gratification regardless of more objective definitions of entertainment. This reasoning was extended to the other gratifications studied. Another reason for predicting no differences
was Palmgreen et al.'s (1980) finding of a moderate t o
strong relationship between gratifications sought and obtained. To the extent that gratifications sought influence
those obtained, a lack of differences by program preferences in gratifications sought (as hypothesized) should mit'
igate similar differences in gratifications obtained.
Finally, w e hypothesized, in accordance w i t h the model
presented, that viewers of a particular program would perceive higher levels of gratifications obtained from their
favorite program than they perceived they obtained (or
would obtain) from competing programs. W e further postulated that such differential perceptions of need fulfillment
would be a major correlate of program choice.

z

METHODOLOGY
SAMPLING

Telephone interviews were obtained in November 1978,
from 327 male and female heads of households in Lexington, Kentucky. Respondents' phone numbers were selected
through systematic random sampling from the Lexington
telephone directory. To qualify as a respondent, an individual had to watch at least one network evening newscast per
week and have at least "fair" reception of all three network
affiliates in the Lexington area.

Downloaded from crx.sagepub.com by guest on April 2, 2012

458

zyxwv
zyxwvu

zyxwv

COMMUNICATION RESEARCH / OCTOBER 1981

MEASUREMENT

Use and Gratifications. Gratifications sought (GS) and
gratifications obtained (GO) were each measured by 15
statements encompassing 5 categories of gratifications
(with 3 statements in each category): (1) General Information Seeking; (2) Decisional Utility; (3) Entertainment; (4)
Interpersonal Utility; and (5) Parasocial Interaction. A more
complete discussion of the rationale for these statements
and the results stemming from a series of factor analyses
using the statements appears in Palmgreen et al. (1980).
Gratifications sought were measured in the following
manner: "We are (also) interested in w h y people watch TV
news. Here are 15 reasons other people have given. As I
read each reason, please tell me h o w much that reason
applies to you. If the reason very definitely applies, give it a
5; if it does not apply at all, give it a 1; if it applies somewhere in between give it a 2, 3, or 4, depending o n h o w
much it applies." The respondent was then read the list of
15 GS items (randomly ordered).
Gratifications obtained (GO) were measured immediately
after the gratifications sought. Respondents were instructed:
"Now we'd like t o know to what extent the network evening
news programs provide you with some of the things we've
just been talking about, w h e n you get a chence to watch
them. . . . First, I want you to tell me h o w much each statement applies to the news programs you ordinarily watch the
most (the "most-watched" program was determined earlier
in the interview).2Then I want you t o tell m e h o w much you
think that statement would apply to the other t w o news
programs, if you had a chance to watch them more often."
Respondents then replied to the same 15 items (slightly
reworded) used to measure GS. For example, for gratification 1, a CBS news viewer was read the statement: "CBS
News helps m e to keep up w i t h current issues and events."
The respondent replied using the same 5-point scale employed to measure GS. For the other t w o news programs,
the respondent was then read the following statement
twice in succession, w i t h the name of the appropriate network inserted in the blank: "
news would help m e to

zyxwvu
Downloaded from crx.sagepub.com by guest on April 2, 2012

zyxw
zyxw
zy

Palmgreen et al. / NEWS PROGRAMS

459

zyxw

keep up with current issues and events.” If t h e respondent
had difficulty in answering this item, he/she w a s asked to
give a n estimate based on “what you think YOU know about
t h e program.” If t h e respondent still could not respond concerning a particular program, h e w a s then asked only about
the remaining program^.^

Program Perceptions. Ratings were obtained mncerning
the anchorpersons on t h e three programs, the news correspondents, program style or format, and quality of news COVerage. Respondents indicated their liking for each anchorperson on a 5-point scale for ABC (Frank Reynolds, Peter
Jennings, Max Robinson, Barbara Walters), NBC (John
Chancellor, David Brinkley), and CBS (Walter Cronkite). For
ABC and NBC, responses were averaged across anchorpers o n s to arrive at a single anchorperson rating for t h e
network.
Ratings of t h e n e w s correspondents on each network program were obtained on a 5-point scale with questions:
“How about the other news correspondents or reporters on
the network programs besides t h e anchormen? How well
do you like these people on
News?” Five-point
scales were also used to measure liking of program style or
format and to rate the quality of news coverage of each
program. In general, those respondents w h o could not
respond to t h e GO measures for least-watched programs
also could not provide program perception measures for
these s a m e programs.

Other Measures. Data were also gathered on total network news viewing, political interests, political discussion,
various media exposure measures, and demographic items.

RESULTS
GRATIFICATIONS SOUGHT COMPARISON

The first general hypothesis of t h e study stated that there
should be n o differences in gratifications sought by program
Downloaded from crx.sagepub.com by guest on April 2, 2012

460

zyxwvu
zyxwvuts
COMMUNICATION RESEARCH / OCTOBER 1981

z

preference. T h e s e results appear in Table 1. For t h e most
part, this hypothesis is supported. T h e analysis of'variance
among network programs for each gratification-sought item
yielded only 4 of 15 comparisons significant beyond t h e .05
level. These results would indicate that viewers, regardless of which network program they viewed, d o not differ
generally in gratifications they a r e seeking.
A possible explanation can b e given for t h e four instances
in which statistically significant differences were found.
Two of t h e indices, "surprised by higher prices" a n d "TV
n e w s is often dramatic," fall o n t h e s a m e factor dimension
which w a s labeled "entertainment" by Palmgreen et al.
(1980). In recent years, ABC n e w s h a s tended to format its
newscasts in a w a y that many consider to b e lighter, or
"entertaining," a n d t h u s it may be that ABC viewers have
come to expect this more often a n d consequently seek what
they have become accustomed to obtaining.
Similarly, t h e indices of "can trust information they give
you," and "find i s s u e s affecting people like myself" also
clustered-together o n a single factor in t h e s a m e analysis.
This factor Palmgreen e t al. labeled interpersonal utilitysurveillance. Examination of t h e m e a n s for e a c h program
indicates t h e personal, humanistic, informal touch of ABC
may attract viewers w h o have more personal orientations in
their information seeking.
GRATIFICATIONS OBTAINED COMPARISON

The second general hypothesis stated that t h e r e should
be no differences in gratifications perceived to b e obtained
across network programs by viewers of t h e s e s a m e programs. A one-way analysis of variance yielded n o differences significant beyond t h e .05 level. These results also
appear in Table 1.
T h e results for hypothesis 1 might indicate that we
should expect significant differences in gratifications obtained on t h e four items w h e r e differences w e r e found in
t h e G S analysis. An explanation for this inconsistency is

Downloaded from crx.sagepub.com by guest on April 2, 2012

zyxw
zyxw
zy
zyxwvuts
Palmgreen et al. / NEWS PROGRAMS

461

TABLE 1
One-way Analysis of Variance of Gratifications Sought and
Gratifications Obtained by Network News Program Viewers

zyxwv

Gratifications
Sought

Cratification
LET-

.i

n
61
79
144

4.46
4.16
4.43

ABC
SBC
CBS

61
79
144

2.85
2.73
2.35

.03

Can trust infornation
they give you

ABC
KBC
CBS

61
80
143

3.30

.01

3.77
3.74
3.54

.26

3.01
2.66

Find out about
goverment officials

ABC
LBC
CBS

61
79
144

3.39
3.10
3.06

.39

3.85
3.78
3.60

.30

?lake up mind about
important issues

ASC
SBC
CBS

61
80
144

3.39
3.10
3.11

.39

3.68
3.68
3.59

.80

KORK

. .

KeeD UD with
current issuesfevents

A0C
LBC
CBS

Surprised by higher
prices

P
.10

Cratifications
Obtained
E
4.46
4.43
4.48

P
.92

3.21

.53

3.03

2.99

zyxwvutsrq

Find issues affectins
people like myself

ABC
KBC
CBS

61
80
144

4.13
3.74
3.66

.02

3.97
3.81
3.70

.26

news is often
entertaining

ABC
LBC
CBS

61
80
163

2.52
2.56
2.36

.51

3.18
3.24
2.94

.21

N news is often
dramatic

ABC
XBC
CBS

61
78
144

2.42
2.32
1.91

.01

3.01
2.93
2.68

.13

N news is often

ABC
LBC
CBS

61
80
144

2.67
2.54
2.28

.09

3.10

.57

exciting
Support viewpoints
to other people

ABC
NBC
CBS

61
80
,143

3.20
3.05
2.91

.55

3.64
3.50
3.32

.ll

Pass information to
other people

ABC
LBC
CBS

61
80
144

2.85
2.65
2.74

.65

3.68
3.50
3.36

.21

Gives ne things to
talk about

ABC
KBC
CBS

61

2.98
2.79
2.96

.60

3.42
3.46
3.45

.98

Kewscasters give
human quality to news

ABC
SBC
CBS

2.88
‘3.00
3.10

.5R

3.63
3.58
3.69

.82

Ty

Compare own ideas
to commentators
Reporters are like
people I know

3.03

2.90

zyxwvuts
79
143
61
80

143

ABC
NBC
CBS

3.03
2.82
2.96

.62

143

3.69
3.50
3.48

ABC
NBC
CBS

61
79
144

2.11
2.16
1.96

.51

2.54
2.89
2.74

61

80

Downloaded from crx.sagepub.com by guest on April 2, 2012

*

.53

.37

462

zyxwv
zyxwvutsr
COMMUNICATION RESEARCH / OCTOBER 1981

that while ABC viewers may "value" these gratifications,
and therefore more actively seek them, it may well be (as
reasoned earlier) that subjective perceptions by different
viewer-types of t h e relation between program attributes
and gratifications obtained lead .to perceptions of equal
satisfaction.
PERCEIVED GRATIFICATIONS OBTAINED FROM
OTHER NETWORKS

The third general hypothesis stated that viewers of a particular program would report higher,levels of gratifications
received from the program they watched than from the
other two programs. t-tests were used to compare mean
differences. These results appear in Table 2.
In testing this hypothesis, each program which an individual said h e viewed w a s compared in turn with t h e other two
network programs. For ABC viewers, 26 of t h e 30 tests
were significant a t .05 in t h e predicted direction. Three
were nonsignificant, and o n e w a s significant in the opposite
direction.
For NBC and CBS viewers, in all 60 instances t h e mean
for each gratification-obtained item for t h e most-watched
program w a s significantly greater than t h e mean for the
other two network programs. These results, then, indicate
substantial support for t h e hypothesis that individuals perceive higher gratifications obtained from the program they
watch than they perceive they obtain (or would obtain) from
t h e competing programs4

zyxwvu

D IS CR IM INANT A N ALY S IS

The general hypothesis that discrepancies in gratifications
obtained from programs within a content type would be
important factors in influencing program-viewing decisions
w a s also tested using multiple discriminant analysis.5 Viewers who could make judgments about all three network news
programs on all 15 items (N = 133) were assigned to groups
according to which program they reported watching most

Downloaded from crx.sagepub.com by guest on April 2, 2012

zyxwv
zyxwvutsrqpon
zyxwvutsrqpo
TABLE 2

t-test (Correlated Groups) Comparison of Gratifications Obtained from Networks

News Programs by Viewers of Different Programs

zyxw
zyxwvutsrqponmlkj
zyxwvuts
zyxwvutsrqponml
ABC VIEhTRS

Network

comparison*

Keep u p w i t h
c u r r e n t issues/
events

S u r p r i s e d by
higher p r i c e s

NBC VIEWERS

i

ARC
NBC
ADC
CBS
ADC
NDC

4.bO
1.98

ABC

cns

p

(1-tail)
,003

Network

comparison**

CBS VIEWERS

Z

1.66
1.16

p
Yetwork
( l - t n i l ) comparison***
.003 CBS

NBC
ARC
NBC
CBS
NBC
ABC

1.60
1.10

.001

4.42
1.70

.On1

4.46
4.05

.007

3.26
1.22

.337

1.28
3.17

.091

NBC
CDS

4.46
3.97

.001

NBC
ARC
NBC
CBS

3.12
2.71

.004

2.97
2.6R

.009

NBC

nnc

1.59
3.19

NBC

3.51

Can t r u s t i n f o r m a t i o n they
g i v e you

ABC
N BC
ARC
CBS

3.74
3.46

.006

3.83
1.56

.004

Find out about
government
officials

Anc
NBC

3.11
1.55

,011

ABC
CBS

3.80
3.41

.001

cns

1.06

ABC
CBS
NBC
CBS
ABC

-

x

4.49

P
(1-toil)
.001

3.55

4.48
3.71

.001

3.05
2.62

.001

CBS
NBC

3.07
2.78

,001

CBS
ABC
CRS
NBC

3.68
3.07

.001

3.63
1.19

.001

.001

CBS
AOC

3.66
3.24

,001

.001

CBS
NRC

i:!:

Downloaded from crx.sagepub.com by guest on April 2, 2012

Graeificatlon

.no1

_________________________I______________-_______-------__--------------------------------

nakc u p mind
a b o u t Importnnt i s s u e s

Find issues
affectinp, peoplc
l i k e me

ABC
NBC

1.69
3.51

.019

NBC
ARC

1.64
1.12

.001

CBS
{BC

3.71
1.06

.OO:

ABC
CBS

1.66
3.47

.029

NRC

3.54

.OOL

cns

1.25

NBC

1.70
3.10

,001

cns

ABC
NBC

3.96
1.71

.019

NBC
ABC

1.36
1.10

.013

CRS
ABC

3.77
3.41

.001

ARC

3.98
1.79

,029

NBC

3.17
3.01

.015

CBS
NBC

3.76

.001

3.4h

cos
w

cos

(Continuedl

P

zyxwvut
zyxwvuts

zyxwvutsrqpo
TABLE 2 (Continued)

ABC VIEWERS

Network

TV news is aften
entertalnlng

cornpsr iwn*
ARC
NBC

N news is

?V news is

often
exciting

CBS VIEWERS

x

p
Network
p
(l-tail) compnrison** X (l-tail)
.on2
NBC
3.27
.OOl
ARC
2.57
.001
NBC
3.15
.001
CBS
2.60

Network
compnrison***
X
CRS
2.96
ABC
2.71

P
(I-tail)
.002,

CBS
N RC

2.97
2.79

.006

.

cns

1.17
2.81
1.18
2.66

ABC
N RC

3.02
2.68

.OOh

NBC
ARC

3.89
1.42

,001

CRS
ARC

2.74
2.56

.019

ARC
CBS

2.98
2.75

.019

ARC
CRS

3.93
3.50

.on1

CRS
N RC

2.77
2.50

.001

ABC
NBC

3.04
2.06

.075

NBC

1.54
3.19

.007

CBS
ABC

3.02
2.hl

.001

ARC

ABC

3.12
2.83

.012

NRC
CBS

1.60
1.10

.023

CBS
NBC

1.50
2.59

.001

ABC

often,
dramatic

NBC VIEUERS

zyxwvutsrqponm
zyxwvutsrqponmlkji
zyxwvutsrqponmlkj
zyxwvutsrqpon

Gratification

cns

--------------------___I________________-------------------------------------------------------

Support viewpoints L O other
people

Pass information
to other
people

G i v e s me things

to talk about

ARC
N BC

1.67
1.19

.001

A BC

CBS

1.70
1.10

ABC
NBC

3.42

,001

CRS
NRC

2.83
1.41
2.91

.no1

,001

CBS
ABC

3.47
3.08

.001

,014

CBS
NBC

3.45
3.12

.001

cns

1.47
3.10

.001

3.50
3.13

.001

NBC
ARC

3.04
2.53

.001

.OOl

NBC
CBS

1.04
2.52

.001

1.67
3.45

.nio

NBC
ABC

3.86

ARC
CDS

3.70
3.49

.013

NRC
CBS

3.81

ABC
NBC

3.47
3.24

.016

NRC
ABC

3.39
2.98

.OOI

ABC
CBS

3.45
1.28

.041

NBC
CBS

1.11
2.83

.ooi

3L38

1.54

CBS
ABC

ABC

cns
NBC

Downloaded from crx.sagepub.com by guest on April 2, 2012

0,

zyxwvut
zyxwvutsrqp
zyxwvutsrqp
TABLE 2 (Continued)

hnc VIEWERS

Network

Newscasters give
human quality to
t h e news

Network

E

,,!tail)

ABC

N nc

3.55
3.20

.ozi

xnc

ARC

3.63

.005

NBC

Anc

3.79
3.35

.003

NBC
ARC

2.57
2.31

.048

NDC
CBS

comparison*

CBS VlEh'ERS

NDC V I C K R S

comparison**

Anc

-

x

3.84
3.23

p
Network
(l-tail) c a p a r i s o n * * * 7
.001
CBS
3.71
AIlC
2.92
.001
CBS
3.69
NBC
3.02

(lrtail)
.001

zyxwvutsrqpo
3.79

.001

cns
3.21
CBS
3.22
...........................................................................................
Compare o m
i d e a s to

know

NRC

ABC
CBS

,

* n varies f r o m 4 7 t o 5 6 f o r each comparison.
* * n varles f r o m 5 5 t o 7 1 for each comparison.
***n varles f r o m 1 0 6 to 1 2 2 f o r each comparlson.

Downloaded from crx.sagepub.com by guest on April 2, 2012

Grot i f ication

3.80
3.47

.003

3.64
3.11

.ooi

CDS

3.18
2.99

.oc1

cns

2.80
2.29

.OC1

nnc

NnC

466

zyxwvutsr
zyxwvu
COMMUNICATION RESEARCH / OCTOBER 1981

zyxw

often. A variety of independent variables thought to be relevant t o the decision process were used t o discriminate
among the three types of viewers. These variables fell into
five basic types: (1) Demographic variables (age, family
income, education); (2)News Exposure (total network news
viewing, radio news exposure, newspaper readership); (3)
Political Interest and Discussion; (4) New Program/Anchorperson(s) Discrepancies, and; (5) Gratifications Obtained
Discrepancies.
The demographics, exposure, and political interest variables have traditionally been used to explain information
source preferences and exposure, and have been shown to
be stronger influences than any gratification perceptions
(see McLeod and Becker, 1974; Becker et al., 1979). Taking
into account the importance the professional press has
placed on the role of anchorperson in the popularity of news
programs, w e assumed that differences in perceptions
among anchorpersons on the three shows would influence
program choice (Powers, 1977). Related t o this, w e speculated that differences in other program attributes could help
explain viewer program preferences. Our discrepancy measure here was based o n a summative total of viewer perceptions of the correspondents (other than the anchorpersons),
the overall program style, and the overall news quality of
each program.
The gratification-obtained discrepancy measure took into
account the average discrepancy in all gratifications obtained (or perceived to be obtainable) from each pair of programs (ABC-NBC, ABC-CBS, CBS-NBC) from the perspective of the viewers of each of the programs. This measure
was based on the model presented earlier (equation 1). For
example, the average discrepancy of gratifications obtained
from ABC and CBS viewing by viewers of the three networks would be computed by summing responses for all 15
gratification-obtained items separately for both ABC and
CBS viewing, subtracting the CBS score from the ABC
score, and dividing this total by 15. Thus a positive discrepancy score would indicate greater gratification obtained

Downloaded from crx.sagepub.com by guest on April 2, 2012

zyxw
zy

zyx

Palmgreen et al. / NEWS PROGRAMS
TABLE 3

467

Multiple Discriminant Analysis of Viewers of ABC, NBC,
and CBS News Programs
Standzrdlrtd Dlscrinlnant Function C o e f f i c i e n t s
V a r i a b l e nane

Function 1

Fmctlon 2

.A1

.18

Average discrepancy of GO between ABC
and S B C vievlng (GO for ABC-SBC)

-.20

.A3

Average discrepancy betveen perceptions of
ABC acchor pcrsons and S B C anchorpersons
(ABC-SBC anchorpersons)

-.01

A v e r a g e discrepancy of Rratification
obtalned (CO bet=.sen ABC and C B S
v i e u l n a ) (GO for ABC-CBS)

zyxwvutsr
.21

A v e r a g e d l s c r e p a n c y between perceptiens
of ABC anchorpersons and C B S a n c h o r m n
(ABC-CBS anchorpersons)
A v e r a g e discrepancy betveen pcrccptions o f
correspondents, s t y l e and overall n e w s
quality (ABC-SBC over311 perception)
A v e r a g e discrepancy between perceptions o f
correspordents. style and overall ncus qualitv
for ABC and CBS (ABC-CBS overall perception)

T o t a l ncws virving
I n t e r e s t in p o l i t i c s
D l s c u s s i o n of politics

zyxwvu

-.24

.33

-.01

.LZ

.73

-.21

.25

-.I4

-.20
.02

.21
-.A2

Radio news exposure

-.01

.n9

Sewspaper resdershlp

-.I3

-.09

Fanily 1ncm.e

-.IS

.28

Age

-.I1

.12

.15

-.06

Ed"Cati0"

Group Centrolds

ABC VIEKERS
SBC VIEKERS

C B S VlEKERS

zyxwvutsrqp
1.58
.R2

-1.08

1.06

-1.24

.18

from ABC viewing, and a negative discrepancy score would
indicate greater gratification obtained from CBS viewing.
Parallel measures were used t o evaluate gratification discrepancies in the t w o other possible network news contrasts.

Three-Group Analyses. The results from the three-group
multiple discriminant analysis show that both possible discriminant functions are significant (see Table 3), and each

Downloaded from crx.sagepub.com by guest on April 2, 2012

468

zyxwvu
zyx
zyxwvuts
COMMUNICATION RESEARCH / OCTOBER 1981

serves to distinguish ABC n e w s viewers from t h e two other
types of viewers. That this analysis serves t o isolate the
ABC viewers from the other types can also be s e e n from an
examination of those variables which failed t h e minimum
tolerance test to enter t h e equation.6 These variables, the
average discrepancy of GO between NBC and CBS viewing,
t h e average discrepancy between perceptions of NBC and
CBS anchorpersons, and t h e average discrepancy between
perceptions of correspondents, style, and overall n e w s quality of t h e NBC and CBS programs, all concern themselves
with comparative judgements between NBC and CBS. While
individually each of the three variables is significant (p <
.001) in a simple one-way analysis of variance,’ t h e between group differences fail to contain enough n e w or nonredundant information (thus, failing t h e tolerance test) to be
entered accurately into t h e discriminant equation.
The first discriminant function derived w a s significant ( x 2
= 170.13, df = 28, p < .001) and serves primarily-to distinguish t h e ABC viewers from t h e CBS viewers. There is a
strong relationship between t h e independent and dependent s e t s (Rc= .76), and t h e function s h o w s that t h e group
centroids are most heavily influenced by t h e average discrepancy between perceptions of correspondents, style, and
overall n e w s quality for ABC and CBS. Not surprisingly, a
high positive discrepancy score w a s associated with CBS
viewing. In line with this result, t h e discriminant function is
next most influenced by t h e average GO discrepancy between ABC and CBS viewers-the ABC viewers receiving
higher levels of gratification from their program than the
CBS program, and t h e CBS viewers being more gratified by
their own program than t h e ABC program. The amount of
newspaper reading also distinguished t h e ABC viewers
from t h e CBS viewers. While t h e ABC viewers tended to
read newspapers infrequently, t h e CBS viewers were most
likely to read a newspaper every day.
The second discriminant function derived from this analysis w a s also significant (xz = 64.20, df = 13, p < .001) and
described a strong relationship between t h e sets of varia-

zyxwv
z
zyxwv

zyxw

Downloaded from crx.sagepub.com by guest on April 2, 2012

zyxw
zyxw

Palmgreen et at. / NEWS PROGRAMS

469

zyxwvutsr
TABLE 4
Classification Results of Discriminant Analysis

Retr.ork Group

so. o f C a s e s

Predicted Group ?lembership

zyxwvu
zyx
zyx

ABC
ABC

29

29 (100%)

SBC

CBS

0

0

SBC

3s

2 (6%)

30 (86%)

CBS

69

6 (9%)

9 (13%)

3

(a)

54 (789,)

NOTE: Percentage of cases c o r r e c t l y classified: 85%.

bles (Rc= .64). The function mostly serves t o contrast the
NBC viewers with the ABC viewers. The average GO discrepancy score between ABC and NBC viewers is the most
influential in maximizing the difference between these t w o
groups. Similar t o the relationship in the first function,
viewers of the NBC and ABC programs reported higher gratification levels from the programs they watched most frequently. Very close to the GO discrepancy in discriminating
power were the discrepancy scores between both the overall perceptions of ABC and NBC and the perceptions of their
anchorpersons. Once again, viewers of both ABC and NBC
programs rated attributes of their program and the anchorpersons o n them as better than those on the other program.
The frequency of discussing politics w i t h friends or family
also distinguished the NBC viewers from the ABC viewers.
NBC viewers tended t o engage in political discussion frequently, while the ABC viewers were much less likely to talk
a bout polit ics.
To test the predictive power of the discriminant functions
derived from this analysis, w e ran a classification analysis
(see Table 4) which reassigned viewers t o a news program
viewing category according t o the group centroids to which
they were nearest o n the basis of the linear combination of
discriminant function coefficients. Our results were surprisingly accurate, correctly reclassifying 85% of our respondents into their most-watched news viewing categories. Of
the ABC viewers, 100%were correctly reclassified, as were
86% of the NBC viewers, and 78% of the CBS viewers.

Downloaded from crx.sagepub.com by guest on April 2, 2012

470

zyxwv
zyx
zyxw

COMMUNICATION RESEARCH / OCTOBER 1981

Two-Group Analyses. Although t h e three-group analysis

w a s extremely accurate in reclassifying n e w s viewers into
their correct viewing group, t h e analysis failed t o make any
finer distinctions between the CBS and NBC n e w s viewers.
In order to see more clearly differences between viewers of
these two programs, we ran a two-group multiple discriminant analysis o n only the NBC and CBS viewers. Also, in an
attempt to see other differences between t h e ABC and CBS
viewers, and t h e ABC and NBC viewers, analyses w e r e run
between t h e s e two pairs of viewer types a s well. In each of
t h e two-group analyses, only those discrepancies contrasting t h e two programs w h o s e viewer types w e r e involved
were used (see Table 5).
T h e function derived from t h e NBC-CBS viewer analysis
w a s signficant ( x 2 = 84.37, df = 1 1 , p < .OOl). The relationship between t h e sets of variables (Rc= .73) w a s most influenced by the overall program perception discrepancy and the
GO discrepancy. In each case, viewers favored t h e program
they watched t h e most. Viewers also favored their anchorperson(s) over t h e competing program's, but t o a lesser
degree. In addition, there w a s a tendency for NBC viewers
to discuss politics more often and to watch television news
more frequently than t h e CBS viewers. The reclassification
results from this analysis a r e slightly more accurate than
those from t h e three-group analysis. Overall, 90% of t h e
NBC and CBS viewers were correctly reclassified. The function w a s more accurate, however, in correctly reclassifying
CBS viewers (97%)than in reclassifying t h e NBC viewers
(81%) to their original viewing group.
T h e ABC-NBC viewer analysis yielded a slightly stronger
(Rc = .78)relationship between sets of variables and the
discriminant function w a s significant (x2 = 54.69, df = 11,
p < .001).The findings largely parallel t h e three-group findings. The variables with t h e most discriminating power, the
discrepancy between overall program perceptions and the
discrepancy between GO, once again favor t h e mostwatched programs. In addition, t h e results again suggest
that NBC viewers tend t o discuss politics more frequently

zyxwv
zyx
zyxwv
zyx
Downloaded from crx.sagepub.com by guest on April 2, 2012

zyxwv
zyxw
zyxwv
zyxwv

Palrngreen et al. / NEWS PROGRAMS

471

TABLE 5

Multiple Discriminant Analyses for ABC versus NBC Viewers,
ABC versus CBS Viewers, and NBC versus CBS Viewers
ABC-SEC Canparison

variable

ABC-CBS Conparisan
Discrtainant
Function
Coefficients

rime

Varlable

SBC-CBS Comparison

Discrirnlnm
Function
Coefficient

rime

Variable n m e

Dlscrlmindnt
Func t i o n
Caefflclents

zyxwvutsrq
zyxwvu

GO f a r ABC-SBC

-.so

CO f o r ABC-CES

.so

GO f o r SBC-CBS

.I9

ABC-SEC anchorpersons

-.26

ABC-CBS anehorpersons

.23

SBC-C8S anchorpersons

.25

ABC-CBS avera l l perception

.S2

ssc-cns

.so

-.s i

ABC-SBC overa l l perception

.OO

T o t a l news viewing

-.I1

I n t e r e s t in
p o l i t 1c 5

.31

D i s c u s s i o n of
politics

Radlo new?.
exposure

Sewspaper
readershlp
F'anily I n t o n e
Age

Letvork

ABC
SBC

.09

-.I6

T o t a l news vlewinz

I n t e r e s t ln
politics

-.oi

I n t e r e s t in
palftics

Dlscusslon of

-.21

D1scuc;sion of
p"lIt1CS

-.04

Radlo news
exposure

,oliticn

Radlo news
cx PO sur e

.26

-.21
.21

.02

zyxwvu
zyxwvutsrqp
.20

-.14

Scwspoprr
readershlp

Xewspaper
readership

-.21

-.40

Fanfly i n c n e

-.06

F m l l y incme

-.la

-.I3

Ane

-.13

~ ee :

-.17

.I8

Education

-

T o t a l new5 viewing

over-

a l l perception

Group
Centroid

-1.30
1.12

Percent of c a s e s c o r r e c t l y
c l a s s i f led: 91%

Educat i o n

.I1

Education

croup
Ketvork
ntrold
_
_ _ _C e _

tvork
__

.I0

croup
Cen f ro i d

ARC

1.74

LBC

1.45

CBS

-.a4

CES

-.7R

P e r c e n r o f cases c o r r e c t l y
c l a s s i f i e d : 92X

P e r c e n t of c a s c s c o r r e c t l y
classlfled:
90%

when compared with t h e ABC viewers. One relationship
which w a s suggested in t h e three-group analysis became
stronger in the two-group analysis. Here t h e loading for
family income level became greater (-.a),
linking higher
income with ABC viewers.and a lower income level with
NBC viewers. Considering 'the similarities with t h e threegroup analysis, it is not surprising that t h e reclassification
results are accurate here a s well. Overall, 915% of t h e viewers w e r e reassigned to their correct viewing group .Some
94% of t h e ABC viewers were correctly reclassified, as were
89% of t h e NBC viewers.

Downloaded from crx.sagepub.com by guest on April 2, 2012

472

zyxwv

COMMUNICATION RESEARCH / OCTOBER 1901

zyxw
zyxwv
zyxwvu

The final two-group analysis makes distinctions between
the ABC and CBS viewers similar t o those found in the three
group analysis. Again, the relationship between sets of variables is strong (Rc = .78), and the discriminant function is
significant (x2 = 88.20, df = 11, p < .OOl). Pa;alleling the
three group findings, both ABC and CBS viewers rate the
overall qualities and gratifications they obtain from their
respective programs the highest, and ABC viewers tend to
read newspapers infrequently when compared t o the CBS
viewers. The reclassification analysis here shows that 92%
of the t w o types of viewers were reassigned t o their correct
viewing groups. This breaks down to the correct reclassification of 92% of the ABC viewers and 93% of the CBS
viewers.
DISCUSSION

The results of both t-test and discriminant analyses indicate, then, that the decision t o view a particular television
news program is strongly related t o the perception of gratifications obtained (or potentially available from) the various
programs. A comparison of means indicates that such choice
behavior is not a function of overall perceptions of gratifications obtained from one's favorite program alone, nor is
it usually a function of gratifications sought from television
news in general. It is only when an audience member
makes a comparison between the gratifications perceived to
be obtained from his favorite program and those offered by
competing programs that a functional relationship with
viewing choice emerges. While in certain cases more specific perceptions of anchorpersons, program format, and
news quality may be more important, the discriminant analyses indicate that the perception of differential gratifications are at least as strongly related to viewing behavior as
the more traditional measures of program attributes.
It may be argued, however, that this strong relationship
does not indicate that differential perceptions of nratifica-

Downloaded from crx.sagepub.com by guest on April 2, 2012

Palmgreen

zyxw
zyxw
zy
81

al. / NEWS PROGRAMS

473

tions obtained are causes of viewing choice, but rather that
viewing choice determines the perceptions of gratifications
obtained. Such an interpretation is consistent with dissonance theory studies of the reevaluation of alternatives following choice (see Brehm and Cohen, 1959; Festinger,
1964). It also seems consistent with Bem's (1968) selfperception theory of attitudes, which holds that w e infer our
attitudes from our behaviors, or from external cues.
Gratifications, however, are not the same as attitudes.
Gratifications, in fact, are more similar operationally t o
Fishbein's (Fishbein and Ajzen, 1975) operationalization of
belief as the perceived probability that an object possesses a
certain attribute, or that a particular behavior will have a
certain outcome. For example, the GS statement "I watch
TV news to be entertained" indicates an expectancy of a
certain outcome (being entertained) as a function of a certain behavior (watching TV news). Similarly, the GO statement "CBS News entertains me" indicates the perception
that either CBS News possesses the attribute "entertainment," or that the.behavior "viewing CBS News" has the
outcome "being entertained." A gratification sought may be
viewed, then, as the seeking of a valued outcome mediated
by the expectancy of obtaining that outcome, while a gratification obtained may be viewed as the perceived outcome of
engaging in a particular behavior (in this case, viewing of a
specific network news program).
Still, even though Fishbein treats behavior as a function
of such perceived outcomes and the expectancies of such
outcomes (as well as attitudes), it is reasonable to postulate
along consistency theory lines that behavior may also affect
"beliefs" in a manner similar to its effects on attitudes. Or,
t o extend Bem's reasoning, w e may infer that because w e
primarily watch NBC News, it must therefore provide us
with valued outcomes more efficiently than competing programs. Ray (1 973) has posited a "Dissonance-Attribution"
hierarchy of effects in which behavior occurs first, then attitude change, followed by cognitive change or learning. This

zyxwvu

z

Downloaded from crx.sagepub.com by guest on April 2, 2012

474

zyxwv
zyxw

zyx
zyxw

COMMUNICATION RESEARCH / OCTOBER 1981

is t h e reverse of the typical "learning hierarchy," and
according to Ray typically involves
situations in which t h e audience has been involved but t h e
alternatives have been almost indistinguishable. The consumer or pseudoconsumer is forced to make a choice of behavior on t h e basis of s o m e non-media a r non-marketing
communication source. Then h e or s h e changes attitude in
order to bolster that choice-often on t h e basis of experience
with t h e chosen alternative. Finally, learning itself occurs on
a selective basis [p. 1521.

This may describe the behavior of many consumers of
television news. These viewers initially may see little to
distinguish among the three networks, but "select" one
program over the others because someone else in the
household (who controls the s e t at that time) generally
watches that program. Or, a person may always watch a
preceding local news program, and conveniently leave the
set tuned to that channel. In such cases, program "choice"
may occur first, followed by attitude change to justify this
choice, and finally cognitive change. The latter may involve
the construction of a schema of gratifications obtained from
the three networks which is favorable to the most-watched
program.
It is not necessary, though, t o postulate such cognitive
self-deception to account for such a behavior-gratifications
obtained sequence. It is just a s likely that the perceived
attributes of NBC News ( v i s - h i s ABC News and CBS News)
are such that we actually are differentially rewarded by
viewing the various programs. In other words, frequently
engaging in a behavior itself is one obvious (though not the
only) way to obtain a n accurate perception of the outcomes
of that behavior.
. W e contend that both the gratifications-behavior, and behavior-gratifications causal sequences are at work in varying degrees in different audience members, and that program choice, gratifications obtained, and gratifications

Downloaded from crx.sagepub.com by guest on April 2, 2012

zyxwv
zyxw
zy
zyxwv
zyxwv

Palmgreen et

I

Sought

al. / NEWS PROGRAMS

475

3

COiN)

A

Figure 1: Model of GS, GO, and Exposure Patterns

sought are

Dokumen yang terkait

Analisis Komparasi Internet Financial Local Government Reporting Pada Website Resmi Kabupaten dan Kota di Jawa Timur The Comparison Analysis of Internet Financial Local Government Reporting on Official Website of Regency and City in East Java

19 819 7

MANAJEMEN PEMROGRAMAN PADA STASIUN RADIO SWASTA (Studi Deskriptif Program Acara Garus di Radio VIS FM Banyuwangi)

29 282 2

Improving the Eighth Year Students' Tense Achievement and Active Participation by Giving Positive Reinforcement at SMPN 1 Silo in the 2013/2014 Academic Year

7 202 3

The Correlation between students vocabulary master and reading comprehension

16 145 49

Improping student's reading comprehension of descriptive text through textual teaching and learning (CTL)

8 140 133

The correlation between listening skill and pronunciation accuracy : a case study in the firt year of smk vocation higt school pupita bangsa ciputat school year 2005-2006

9 128 37

Perilaku Kesehatan pada Mahasiswa Program Studi Pendidikan Dokter UIN Syarif Hidayatullah Jakrta Angkatan 2012 pada tahun2015

8 93 81

Implementasi Program Dinamika Kelompok Terhada Lanjut Usia Di Panti Sosial Tresna Werdha (Pstw) Budi Mulia 1 Cipayung Jakarta Timur

10 166 162

Analisis Prioritas Program Pengembangan Kawasan "Pulau Penawar Rindu" (Kecamatan Belakang Padang) Sebagai Kecamatan Terdepan di Kota Batam Dengan Menggunakan Metode AHP

10 65 6

Sistem Informasi Pendaftaran Mahasiswa Baru Program Beasiswa Unggulan Berbasis Web Pada Universitas Komputer Indonesia

7 101 1