University of Southern California Dissertations and Theses

An experimental analysis of source credibility and message discrepancy under differential levels of ego-involvement in political television commercials

(USC Thesis Other)

An experimental analysis of source credibility and message discrepancy under differential levels of ego-involvement in political television commercials

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content AN EXPERIMENTAL ANALYSIS OF SOURCE CREDIBILITY AND MESSAGE
DISCREPANCY UNDER DIFFERENTIAL LEVELS OF EGO-
INVOLVEMENT IN POLITICAL
TELEVISION COMMERCIALS
by
John M. Boyd
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
in Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Telecommunication)
August 5 1977
UMI Number: DP71316
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
in the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
UMI
Dissertation R jblishing
UMI DP71316
Published by ProQuest LLC (2015). Copyright in the Dissertation held by the Author.
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code
uesf
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106- 1346
UNIVERSITY OF SOUTHERN CALIFORNIA
T H E G R A D U A T E S C H O O L
U N IV E R S IT Y P A R K
LO S A N G E L E S , C A L IF O R N IA 9 0 0 0 7
T
'78
8 7 8 9
c^s/je
This dissertation, written by
John M . Boyd
under the direction of . Dissertation Com
mittee, and approved by all its members, has
been presented to and accepted by The Graduate
School, in partial fulfillm ent of requirements of
the degree of
D J D C T O R O F P H I L O S O P H Y
Dean
Date...
DISSERTATION COMM]
airman
.....
TABLE. OF CONTENTS
Page
LIST OF TABLES........... , . . . . ..... iv
LIST OF FIGURES. . . . . . . . . ............. v
Chapter
•
1. THE PROBLEM............................. 1
2. DESCRIPTION OF PERTINENT LITERATURE. .
Source Credibility
Message Discrepancy
Ego-Involvement
Integrative Summary of Pertinent
Literature
12
3. OPERATIONAL PLAN ......................
Operational Hypotheses
Procedure for Testing the
Hypotheses
Limitations
47
4. PRESENTATION AND ANALYSIS OF DATA . .
Questions and Analysis of Data
Hypothesis One
Hypothesis Two
Relevance of Findings to Pertinent
Theoretical Literature
Additional Findings
64
5, SUMMARY^ CONCLUSIONS, AND IMPLICATIONS
Summary
Conclusions
Implications
7 7
BIBLIOGRAPHY • ........... ....................
84
APPENDIXES
A.
B.
C.
Issues.. . . . . .. .. » . . . . . .
Newspaper Advertisement and Question
naires . . . . . ' . . .... . . . .
Ballot . . . . . . . . . . . . . . . .
9 0
114
12 4
ii
LIST OF TABLES
Table Page
1. Means and Standard Deviations of the
Dependent Variable (jVote)................ 6 6
2, Voting Scores.as: Adjusted for Differences
in the Groups Who Viewed the Coinmercials • 6 7
3. Analysis of Variance: Simple Main Effects
for Hypotheses: One and Two................ 69
4, Analysis of Variance: Control-Vote-
Adjusted Simple Main Effects for
Hypotheses One and Two........... .. 70
111
LIST OF FIGURES
Figure Page
1. Opinion Change as a Function of Credi
bility and Extent of Discrepancy. . . . . 21
2. Change in Attitude Against Change in
Discrepancy . . . . . . . . . . .......... 41
IV
CHAPTER 1
THE PROBLEM
The problem of this study focused on whether or
not the source credibility or physical attractiveness
or respectability of a television candidate is more
important than the message he is presenting to a pros
pective viewer. This media problem becomes more salient
when the candidate presents a message that is contrary
to the political stand of most of the viewers of the tele
vision message. The problem, then, was to investigate
whether or not source credibility is really as important
as, or more important than, is now believed in television
political campaigns. A typical euphemism for source
credibility in this context is the so-called "pretty
boy" phenomenon, which is a term that might apply to such
political candidates as John Tunney or Edward Kennedy,
whereas the term "ugly boy" might be aptly applied to a
candidate such as Jess Unruh or Richard M. Nixon. An
area equally related to the problem and interacting with
it was the investigation of whether or not the high
source credibility candidate would have the greatest
effect on persuading subjects to vote for that candidate,
even though the viewers were not greatly involved with
the candidate or the issues he represented.
Television has long been an important factor in
the democratic process. Attempts to exploit its
persuasive assets were made as early as 19 2 8 when Governor
Alfred Smith faced the television cameras on the steps of
the state capitol in Albany, New York.^ One of the first
assessments of the effectiveness of a television
presentation by a political candidate was that of Edwin H.
James, who judged President Harry Truman as "pretty good"
because of his sincerity and determination. Senator
Robert Taft was found to be just the opposite since he
2
gave a "shaky" presentation. Later in political history,
during the Stevenson-Eisenhower race, Adlai Stevenson was
seen by political experts as appearing "snobbish and
domineering" on television even though his oratorical
capacities were well known. In recent years the view has
been often expressed that the image presented by Kennedy
in the 19 60 debates may well have been crucial in
determining the election results. The factors that
caused these differences in image are still speculative,
as is the importance of the image itself. This disserta
tion studied variables which were determinants in
effective political use of the television medium.
^Edward W. Chester, Radio, Television and
American Politics (New York: Sheed and Ward, 1969), p. 70
^Ibid., p. 73.
Campaign spending for elective offices has
sharply increased in the last several years, A report
issued by the Twentieth Century Fund Commission on
Campaign Costs in the Electronic Era concluded that
expenditures for political advertising on television grew
from ten million dollars in 19 60 to over twenty-six
million dollars in 1968, an increase of over 150 percent
in only eight years. Additional increases also came in
19 72. Typically, a great deal of this money came from
donors who expected a political favor in return for their
support if their candidate was elected. In light of
these observations, there is great pressure on all
candidates to make the most effective use possible of the
large sums available to them. For instance, the mammoth
increase in expenditures for televised political
advertising has contributed to the growth of campaign
consultants and consulting firms, which profess to deal
specifically with grooming potential candidates so that
they can effectively use the mass media, specifically
radio and television. Dan Nimmo, in the book. The
Pblitical Persuaders, describes the emergence of this
new industry:
A century ago candidates relied on their wits ,
their friends, and a few trusted allies to mount
3 . . .
This rise is described in the Report of the
Twentieth Century Fund Commission on Campaign Costs in
the Electronic Era, Voter's Time (New York: The
Twentieth Century Fund, 1969), p. 10,
3
a campaign for office .... Today, however,
as aspiring office-seekers approach the
elections of the 19 70's, they turn less to
party leaders than to professional campaign
managers for political expertise.^
The scholarly purpose of this study was to
investigate the persuasive effects of two different types
of televised political commercials. One type of com
mercial was designated highly credible while the other had
low credibility. In the highly credible commercials a
favorable image of the candidate was portrayed. He had
certain definitive qualifications consisting of:
trustworthiness, competence, pleasantness, attention,
and tension, and, in general, an actor's type of tele
vision personality. On the other hand, the low credit'
bility commercials presented an unfavorable image of the
candidate, having him appear in poor light, being
hesitant, and generally not having a smooth presentation
in his delivery.
If the two variables of source credibility and
message were considered alone, it would seem obvious
that the candidate with the unfavorable image would be
less likely to be persuasive. However, an additional
consideration was that the high credibility candidate
presented a message which was discrepant with the beliefs
of the viewers while the low credibility candidate
^Dan Nimmo, The Political Persuaders (Englewood
Cliffs: Prentice-Hall, Inc., 1970), p. 34.
presented a low discrepancy message. An integral part
of this was to examine the interacting effect of these two
variables in respect to their effects upon people who have
different levels of ego-involvement with the issue being
discussed by the candidates.
Current research indicates that there are two
trends in thinking relevant to the utility of political
campaign consulting. One trend is negative and is
represented by the attitude that,
. . . the campaign slogan, theme and symbol has
replaced the discussion of issues by candidates,
because the advertising man has no confidence
in the public's ability to listen to campaign
discussion.^
This theme is often promulgated by the so-called
political bosses and other members of the old party
machinery. Essentially, they feel that the structure
of the new politics, as represented by professional
campaign managers with their vast arrays of spectacular
media displays have, for the most part, circumvented and
eroded the power of the traditional party machine. This
has been accomplished, they say, because those politicians
who practice the new politics appeal directly to the
voter through the mass media. A larger group of critics,
however, seems to take the view that the real question
revolves around the type of political presentation* A
5
Robert Agranoff, The New Style in Election
Campaigns (Boston: Holbrook Press, Inc., 1972), p. 42.
5
real concern is raised here about the polishing of
candidates. Dan Nimmo cites the concern of one political
science professor:
The mass media campaign is not directed to the
informed voter but to the eyes and ears of the
voter who does not care very much. . . . Many
of the viewers have more experience in evalu
ating television, film, and recording person
alities than in deciding public issues.G
What this seems to mean is that there is a fear that the
image rather than the issues has, because of electronic
campaigning, become of primary importance. Stated
another way, what has become more important in modern
campaigning, the issues discussed by a candidate or his
image?
The other trend in thinking relevant to campaign
consulting appears to be in a reaction of its leaders to
that of seeing this craft as a positive social influence.
It provides the expertise to design and produce a useful
political message using the highly technical electronic
media. It is believed that television plays an especially
unique role in this process of political education in
that, "... messages aimed at the poor and under-
7
educated are best conveyed through television. . . ."
People who would not otherwise receive any political
6
Ibid. 5 p. 118.
^Joseph Napolitan, The Election Game and How to
Win It (Garden City: Doubleday and Company, Inc., 1972),
p. 128.
information whatever because of either a lack of interest
or initiative are at least provided with a few essential
facts (typically in the form of television commercials)
as they watch their favorite programs each evening. How
ever, it seems that many questions remain unanswered as
to the most effective methods of presenting the political
television message. According to Nimmo, "... not all
8
persons or causes can be merchandised effectively."
Determining whether issue-oriented or image-oriented mes
sages are most effective, as was done in the present
study, seemed to be one way of improving the political
commercial on television.
As a refutation of those critics of the refined
political commercial, two points should be made clear.
First, it can be reliably demonstrated that not all
campaigns are completely image oriented, as is sometimes
asserted. For example, Robert Agranoff in the book.
The New Style in Election Campaigns, illustrates this
point by citing the re-election campaign of Hawaii
governor John Burns as being a campaign which deliberately
played down the governor's image:
None of the spots asked voters to vote for
Burns, nor did the governor appear in any
of the spots. They were strictly issues-
oriented,^ factual, and unemotional. All
carried the tag line 'Think about it.'^
Agranoff, op. cit., p. 43.
^Ibid., p. 117.
This campaign, says Agranoff, brought Burns from
behind with a sixteen point deficit to win re-election.
A second argument against the ei'itic who sees the
consultant as a manipulator and constructor of slick
political packages pertains to the increasing problem of
candidate credibility. A recent Harris poll indicated
that "More than 60 percent of those interviewed . . .
expressed the belief that politicians take graft and that
very few of them are dedicated public servants.Such
a view is potentially dangerous to the democratic system.
Seeing new ways whereby images of politicians and political
candidates could be effectively upgraded was a positive
social value and the social purpose of the present study.
A final factor which should be stressed is the
social influence on television and the impact it has on
the political process. With over 9 5 percent of all
homes equipped with televisions ets, and with the
average set on for over six hours a day, television is
the medium which has potentially greater impact than any
11
other medium in history. It seems only natural that
the present marriage between politicians and television is
occurring since politics operates under the premise of
swaying voters as effectively as possible in order to
10
Twentieth Century Fund Commission on Campaign
Costs in the Electronic Era, p. 12,
^^Robert T, Bower, Television and the Public
(New York: Holt, Rinehart and Winston, 1973), p. 3.
obtain power. The questions posed in this study were,
therefore, properly investigated by a researcher in tele
communication, using applications of principles taken from
the area of communication theory. Through this inter
disciplinary approach the discoveries concerning per
suasion were addressed to the questions of political mass
advertising. It is from communication theory that the
investigation of both the role of ego-involvement and the
role of message-discrepancy in the political communication
process came about, while the question of the effects of
source credibility and television commercials arose
actually from political exigencies. The present study was
different from past work in the area of communication
theory in two respects.f
1. It was conducted'using lifelike fictional
television commercials.
2, It investigated the use of these messages
regarding political attitude change and
change in voting behavior.
In general, attempts were made to make the commercial
announcements and their presentation to the subjects as
realistic as possible in order to maximize external
validity in the real world. These measures increased
the expenses involved in the research, but the increases
were felt to be justified by the additional value of the
information obtained.
Questions
Two questions were answered by this study. The
first of these questions concerned source credibility, or
the physical attractiveness or respectability of the
political candidate. It was based upon the belief that
image is an important factor in determining the impact of
a political message. The question ;was then:
1. Is the influence of source credibility
stronger than that of the message being presented? This
question was tested by having a highly credible source
present a message which was discrepant with the beliefs
of the viewers. For example, when it was determined that
the population being examined favored the issue of
campaign spending limitations, the candidate advocated a
stand against limitations. An opposite type of treatment
was also administered for comparison in which a low
credibility source, a candidate with an unfavorable
image, presented a message which supported the viewers'
beliefs at that time.
A second question emerged as to what types of
people were most influenced by the televised political
commercial. The research determined the characteristics
of this type of person. Since a considerable amount of
work has been done regarding the effects of persuasive
communications in the area of ego-involvement, the writer
believed that a useful hypothesis could be formulated
based upon highly- versus lowly-involved subjects. As
the review of literature will show, it has now been demon
strated that low-involvement subjects are usually more
easily persuaded. Accordingly, by applying this informa
tion to the realm of televised political commercials , it
was asked :
2. Will source credibility have its greatest
effects upon the voting behavior of lowly ihvolved
subjects?
11
CHAPTER 2
DESCRIPTION OF PERTINENT LITERATURE
The literature review includes both académie books
on political television in this area and research articles
in the field of communication research. A synthesis of the
two areas yields a potent approach for applying principles
from communication to the disciplines of telecommunication
and political science.
One of the more useful books, of a practical nature
which deals with television politics was Robert Agranoff s
The New Style in Election Campaigns.^ This book was partic
ularly useful in formulating a comprehensive idea of what
should be meant by source credibility as this term applies
to political media-campaigning. As Agranoff points out:
The personal dimension of a candidate is critical
in the minds of some voters. Aspects of this
dimension include the degree of warmth of the
candidate's personality, skill and competence for g
the office, personal appearance and issue domains.
All but the last of these aspects, issue domains, were
considered important in establishing source credibility in
the political commercials which were used in this study.
^Robert Agranoff, The New Style in Election Campaigns
(Boston: Holbrook Press, Inc., 1972).
^Ibid., p. 121.
12
Issue domains were treated separately under the area deal
ing with message discrepancy. Agranoff discussed the
contrasted personal images presented by the two major
candidates in the 19 64 presidential election (Johnson and
Goldwater) in terms of their differing images. These
noted differences also precisely exemplify high and low
source credibility as they exist in the present study.
In this election Johnson was shown in surveys to be con
sidered "warm and friendly," "showing good judgment," and
"speaking his own mind," as well as being "honest and
sincere." In short, Johnson had high credibility. On the
other hand, the personal image of Goldwater was often
described as "too conservative" in his physical appearance
although this may or may not have reflected his political
stand on various issues. Furthermore, many people said he
"acts without thinking," but also that he "has strong
convictions." Clearly, Johnson came across in the media
as more credible,
A second useful book was that of Mendelsohn and
Crespi entitled Polls, Television and the New Politics.^
They state that there is ". . . emergence of a new cri
terion for the public to use in judging the President of
the United States— the acceptability of his personality.^
3
Ibid, , p. 218.
4
Harold Mendelsohn and Irving Crespi, Polls, Tele-
vision and the New Politics (Scranton: Chandler Publish-
ing Co., 197 0).
_________^Ibid. ■ p. 2RS._____________________________________ 13
. . . Anyone who appears physically attractive,
sincere, diligent, toughminded, poised, bright,
assertive, and at the same time appears to be of
Presidential timber willgUltimately have a very
strong go at the office.
The experts agreed that source credibility is important ;
Considerable anxiety about the projection of the
'right' television image to the electorate is
exhibited by politicians in the knowledge that a
'correct' projection can make them successful,
while the 'wrong' projection can doom them to
failure.
Mendelsohn and Crespi quoted James Reston of the New York
Times who had some further thoughts on this issue:
Instead of the old-fashioned emphasis on what a
candidate thinks, or what he says, the emphasis
now seems to be on how he looks, especially on
television, and on what kind of personality he
has. 8
It was statements such as these that helped
determine that source credibility was sufficiently
important to justify the formulation of hypothesis con
cerning its effects.
Dan Nimmo's The Political Persuaders^ was an
excellent source which described some of the practical
aspects of creating the proper image. According to Nimmo
. . often the candidate must project an image of one
10
capable of acting in ambiguous situations." He candidly
admits to the existence of the problem area to be
examined in the proposed study by saying, "Because of the
importance of television in campaigning, managers are
Ibid., p. 218. ^Ibid., p. 267. ®Ibid., p. 272
Nimmo, op cit. Ibid. , p. 8. - j _
------------ -------------- --------------------- J - J - -------------------------
concerned about physical appearance. In an actual real-
life situation, the importance of a proper image was drama
tized graphically by the Nixon-Kennedy debates which demon
strated that ". . . viewers of the first of the Kennedy-
Nixon debates on television generally credited Kennedy with
'winning':; but persons who had listened only to radio
12
thought Nixon had done the superior job." Nimmo provided
this description of campaign media planners:
. . . television advisers are convinced that
personalities and not issue stands or political
parties win votes. The overall ploy is contrived
spontaneity, the effort to appear uninhibited,
candid, open, and credible without running the
risk of an unrehearsed performance. Contrived
spontaneity has made idols of ordinary performers--
Johnny Carson, Andy Williams, Jerry Lewis. .
These statements reinforced the first hypothesis
of the present study that source credibility was more
important than was the message being presented.
An example of the influence of this factor was
reflected by former President Nixon's efforts to improve
his appearance in the Campaign of 1968, as Nimmo described
him dieting and tanning his face in order to improve his
appearance for the television camera. Furthermore, Nimmo
stated: "To sharpen his performance at televised news
conferences, Nixon avoided notes and the rostrum to appear
cool, informed, and fearless in the face of the pressures
of the office."
^^Ibid. , p. 43,. ^^Ibid. ^^Ibid. , p. 141
^^Ibid., p. 143.
The practical side of political television
campaigning was important to the review of literature as
well as formulating the principles of this research
problem. However, it should be noted, that a strenuous
examination was■also made of scholarly books and articles
to discover what principles of persuasion might be useful
to formulate hypotheses.
Source Credibility
When recalling the first question of the study
which asked whether source credibility was more persuasive
than was the message being presented, a review of past
research done on each of these variables proved crucial.
Such articles are discussed in this and the following sub
section. Bowers and Phillips have posited some important
findings in the area of source credibility.^^ Two
dimensions in the perception of the credibility of a
source were discovered by Bowers and Phillips during the
course of their research. They labelled these dimensions
"trustworthiness4 or "character" and "competence" or
"authoritativeness." One finding especially relevant to
the present study was that:
. . . trustworthiness and competence are appro
priate perceptual dimensions for semantic
John Waite Bowers and William A. Phillips,
7A Note on the Generality of Source-Credibility Scales,7
Speech: Monographs, 34:185-186, June, 1967.
16
differential tests for source-concept construc
tions as well as for^gests of sources separated
from their concepts.
In the so-called real world of political campaign
ing, the concepts presented by a candidate are an integral
part of his image and any separation of image from concepts
creates an artificial situation with low external validity.
With the discovery of these dimensions it was decided that
the two could be examined together. The term "source credi
bility" was used as an umbrella term to denote the dimen
sions of trustworthiness, competence, pleasantness, atten
tion, and tension as they are combined to produce a single
dynamic communication. This consisted of the independent
variable. In one instance the programs consisted of an
image, while in the other case they consisted of an issues
appeal.
In 19.66 a L study was undertaken by James McCroskey
in which the authoritativeness and character dimensions
were identified using two different, but highly correlat-
17
ing approaches. It was found that both the Likert-type
scale and that of the semantic differential, as set forth
by Osgood, produced these similar dimensions. Through
factor analysis, six polar adjectives were suggested as
useful for obtaining measurements of each of the two
dimensions. These polarities were:
^®Ibid., p. 186.
17
James C. McCroskey, "Scales for the Measurement
of Ethos," Speech Monographs, 33:65-7 2, March, 19 66,
70._____________________________________________________________: _______________________________________: __________________________________________________________12
Authori ta t i ven é ss Charac t er
Reliable-Unreliable Hones t-Dishonest
In forme d-Un in forme d Friendly-Unfri endly
Qualified-Unqualifiéd Pleasant-Unpleasant
Intelligent-Unintelligent Unselfish-Selfish
Valuable-Worthless Nice-Awful
Expert-Inexpert Virtuous-Sinful
Each of these sets of adjectives was adopted for the
purposes of measuring source credibility, or ethos, as
it is referred to by McCroskey in his study.
In order to make predictions regarding the above-
mentioned dimensions it was necessary to decide upon the
selection of an appropriate theory as a basis for explain
ing results. Several such theories were examined.
Congruity theory, as demonstrated by Osgood and Tannenbaum
in 19 57, can be used to predict the manner in which
evaluations of a communication are affected by its ’
source.According to congruity theory, when two
objects are positively associated, the evaluation of one
will be displaced toward the evaluation of the second.
Therefore, a communication from a positively regarded
source would be evaluated more positively than the same
communication if attributed to a negatively regarded
source. This seemingly straightforward approach, however,
becomes complex upon consideration of actual findings.
18 .
Ibid., p. 72.
^^'Charles E. Osgood, George J. Suci, and Percy H.
Tannenbaum, The Measurement of Meaning (Urbana: Université
of Illinois Press, 195 7).
18
For instance, Wyer and Schwartz found some
interesting results based on congruity theory:
. . . source effects on communication evaluations
are predictable from congruity theory only to
the extent that the communication is contra
dictory to previous communications from the
source that form the basis for the evaluation
of that source.20
They state that these findings can be better explained
through exchange theory, as put forth by Thibaut and
Kelly in 1959, than by Osgood's congruity theory. That
is, ”. . . that a communication from a positively regarded
source would be evaluated less positively than the same
communication if attributed to a negatively regarded
,,21
source."
Another complication was inserted when a now
classic 1951 study by Hovland and Weiss was also brought
2 2
into the evaluation of various theories. Contrary to
the Wyer and Schwartz analysis, these investigators
found that:
. . • significant differences were obtained in
the extent to which opinion on an issue was
changed by the attribution of the material to
different sources . . . . Subjects changed
their opinion in the direction advocated by
2 0
Robert S. Wyer, Jr., and Sandra Schwartz, "Some
Contingencies in the Effects of the Source of a Communica
tion on the Evaluation of that Communication," Journal of
Personality and Social Psychology, 11:1-9, 1969, p. 1.
^hbid.
Carl I. Hovland and Walter Weiss, "The Influence
of Source Credibility on Communication Effectiveness," Pub
lic Opinion Quarterly, 32:635-650, Winter 19 51-52.______
the communicator in a significantly greater
number of cases when the material was attributed
to a ^high credibility* source than when go
attributed to a * low credibility * source.
Message Discrepancy
A refinement of the question of whether source
credibility was crucial in sw^ing opinions was introduced
by Aronson, Turner, and Carlsmith. These investigators
felt that message discrepancy would play an interacting
24
role with source credibility in the influencing process.
They discovered that when subjects were exposed to a high
credibility source, the amount of influence increased as
the discrepancy of the message presented increased. That
is, the greater the discrepancy, the greater the amount
of influence. However, this trend was not completely true
for a mildly credible communicator. In this instance, the
amount of attitude change increased concurrently with
discrepancy only to a point, and then began to decline
with, further increases in discrepancy. Placed in graphic
form, the relationship of discrepancy, opinion change,
and source credibility appears as follows:
Ibid. , p. 642.
24
Eliot Aranson, Judith A. Turner, and J. Merrill
Carlsmith., "Communicator Credibility and Communicator
Discrepancy," Journal of Abnormal and ■Social Psychology,
67: 31-36 . July. 196 3. 2h
o
g-
p.
0
p
1
Discrepancy
Figure 1
High Credibility
Mild Credibility
Opinion Change as a Function of Credibility and
Extent of Discrepancy's
Later, it will be shown that a similar relation
ship between discrepancy, opinion change, and ego-
involvement had been uncovered by Jonathan Freedman,
though Freedman spoke in terms of attitude change rather
than opinion change as did Aranson, Turner, and
Carlsmith.
Since this study also took into account the
effects of message discrepancy, a review of research on
this variable was. in order. An important finding by
Bochner and Insko explained why this additional variable
was examined in this study. They discovered that:
Opinion change was found to be linearly related i
to commun ica tor-c ommuni cate e discrepancy for a
high credibility source, and curvilinearly
related to communicator-communicatee discrepancy
2 5
Ibid., p. 34.
21
for a medium, credibility source. Furthermore,
the curve trends for the 2 sources interacted
significantly.2 6
The fact that there was interaction, that is,
the combination of these two variables to form an effect
which was greater than the total effects of either
variable alone, suggested that a more meaningful approach
would take into account both variables simultaneously.
Of their finding of a significant source credibility-
message discrepancy interaction F, Bochner and Insko
stated :
. . . this interaction between communictor-
communicatee discrepancy and source credibility
warns against any further study of source credi
bility that does not incorporate a manipulation
of a discrepancy.2 7
The present study manipulated discrepancy in that
the low credibility source communicated a low discrepancy
message, and a neutral source communicated a neutral
message, while a high credibility source communicated a
high discrepancy message. The extent of the discrepancy
was determined by the difference between the attitude of
the subject and the views stated by the message.
In addition to the findings of Bochner and Ihàko
and those mentioned earlier of Aranson, Turner and
Carlsmith, many other scholars have sought to ascertain
Stephen Bochner and Chester A. Insko, VCommuni
cator Discrepancy, Source Credibility, and Opinion Change,'
Journal: of Personality and Social Psychology, 4:614-621,
19 66, p. 614.
_________ ^"^Ibid. ^ p. 619.____________________________________^
the effects of message discrepancy, A doctoral disserta
tion written by Scott provided a comprehensive summary of
much of the past work in this area, as well as making the
point that there was still a great deal of conflicting
information about the question of whether greater or lesse^
degrees of message discrepancy were most effective in
2 8
altering attitudes. Basically Scott felt there were
two schools of thought, one advocated by Festinger, and
consistent with balance theory, which predicted greatest
change from high discrepancy messages, and one advocated
by Sherif, Sherif, and Nebergall, which predicted that
". . .in the absence of high source credibility increases
in the level of message discrepancy will lead to less.
2 9
attitude change," Supporting the Festinger school of
thought were Goldberg, Hovland and Pritzksr, Zimbardo,
Bergin, Aronson, Turner, and Carlsmith, and Bochner and
Insko, while those holding the opposite position were
Whittakër, Freedman, Atkins and Beri, and Mortensen and
Sereno, all of whose works were reviewed in the present
^ . 30
s tudy.
kichael D. Scott, "Attitude Change as a Function
of Ego Involvement, Message Discrepancy, and Source Credi
bility: An Experimental Test of Opposite Theoretic State
ments," (unpublished Doctor's dissertation. University of
Southern California, 1973).
2 9
Ibid., p. 21.
30
Ibid., pp. 20-21. 23
Turning chronologically to this question of
which theory provided better predictions, in 1964 Whittaker
discovered empirical support for the hypothesis that there
was an optimal point of message discrepancy which resulted
in maximum shifts in opinion and that smaller amounts of
discrepancy yielded negligible opinion shifts, while
larger amounts of discrepancy yielded the same amount of
change as did the optimal level. An important finding
was that for extremely large discrepancies in message,
there was a boomerang effect, which meant that the amount
of persuasion dropped in this area. The significance of
this discovery became clear much later when other -
researchers such as Freedman established the nonmonotonic
nature of the discrepancy/attitude-change relationship.
In addition, it cast some doubt on the validity of
Festinger*s cognitive dissonance theory. Whittaker
concluded from his findings :
It is difficult to explain these results on the
basis of dissonance theory. Although greater
dissonance may be produced through large than
small discrepancies, it is apparent that such
dissonance is not reduced through attitude
change.
Three years later, in a study supported by the
National Science Foundation, again Festinger's theory was
31
James 0. Whittaker, "Cognitive Dissonance and
the Effectiveness of Persuasive Communications," Public
Opinion Quarterly, 28:547-555, Winter, 1964, p. 55#.
rejected in favor of that proposed by Sherif et al. Spe
cifically, using two levels of credibility, two levels of a
variable labelled attitude stability, and four levels of
communication discrepancy, it was found that subjects who
were unstable in their attitudes were likely to be inf].u-
enced by source credibility and, most importantly, that
". . . opinion change was a curvilinear function of dis-
3 2
crepancy." Because of this, it was concluded that the
curvilinear change results are in accord with social judg
ment theory (Sherif and Hovland, 19 61; Sherif et al., u
1965).
The next important study came in 1968, a year
after the findings of Koslin, Sfoops, and Loh. In this
case Sereno had a high credibility source present a belief-
discrepant message to subjects with high or low levels of
ego-involvement with the following results : ". . . lowly
involved subjects had significantly greater amounts of
attitude change toward the topic than "did highly involved
34
subjects." Also, though to a lesser extent, the highly
involved subjects lowered their evaluation of the source
more than did lowly involved subjects. Though message
3 2
Bertram L. Koslin, James W. Stoops, and Wallace
D. Loh, "Source Characteristics and Communication Discrep
ancy as Determinants of Attitude Change and Conformity,"
Journal of Expérimenta1 and Social Psychology, 3:23-242,
July, 1967) , p. 230
^^Ibid., p. 239.
34
Kenneth K. Sereno, "Ego-Involvement, High Source
Credibility, and Response to a Belief-Discrepant Communica
tion," Speech Monographs, 35:477-481, November, 1968.
25
26
discrepancy was not an independent variable in this study,
the results reaffirmed the efficacy of Sherifs* theory
over that of Festinger in predicting what behavior would
result during the resolution of cognitive inconsistencies.
The relationship of cognitive dissonance theory
and ego-involvement theory was concisely summarized in a
later article by Sereno. He pointed out that the problem
was that :
. . . cognitive dissonance theory postulates that
all forms of dissonance-reduction, including
attitude change, resistance to change, and dero
gation of the communicator are isolated and
operational alternatives., Unlike cognitive
dissonance theory, however, the ego-involvement
approach postulates that all forms of cognitive
balance— whether they be change of attitude on
the issue, derogation of the communicator, or
assimilation and contrast of the message— are
interrelated functions of an individual's
involvement in the issue, and are hot merely
optional, isolated choices,36
This characteristic of ego-involvement theory greatly
heightened its utility for making practical predictions.
Also worth noting was a study by Hovland and
Pritzker which was done under a grant from the Rocke
feller Foundation in 19 57. As will be seen later when
research on the variable of ego-involvement is discussed,
Hovland was, at this time, heavily concerned with the
phenomenon of assimilation contrast of attitudes and the
so-called anchoring effect.
3 5
Kenneth K. Serend, "Ego-involvement:: A
Neglected Variable in Speech-Communication Research," The
Quarterly Journal of Speech, 55:69-7 7, February, 196 9, p. 74
: ' 2&_
A historical note of interest was that although
Hovland was, at this point, studying communication
discrepancy and its effects on attitudes, he had not yet
discovered that attitudes may be nonlinear. He stated:
Opinion on each of the twelve topics was obtained
on a seven-point scale. The Ss were asked to check
the one category which most nearly expressed their
opinion toward the statement.3 6
This linear approach to attitude research, typical during
the 19 50's and used even by such prominent researchers
as Osgood, made results obtained from studies of this
period of only limited value. Later Hovland began his
landmark work with Sherif in the discovery of assimila-
tion-contrast and the accompanying attitudinal latitudes
which refuted the linear interpretation of the construct
attitude. This later research was reviewed under the ego-
involvement section below.
Ego-Involvement
The second question to be answered in the study
was whether source credibility had its greatest effects on
lowly-involved subjects. While the variable of source
credibility was previously discussed, that of ego-
involvement needed to be examined if it were to be
included in this study, rShefif^and^Sherif in the book
Social Psychology stated that as a person's thinking
Carl I. Hovland and Henry A. Pritzker, "Extent oi'
Opinion Change as a Function of Amount of Change Advocated,
Journal of Abnormal and Social Psychology, 54:251-257,
1957, p. 257.
2 7
was stabilized as an attitude, his ego-involvement
delimits and regulates his characteristic modes of
behavior to the stimulus that triggers that behavior.
Such a construct is defined as ego attitudes;
Ego-involvement refers to the arousal of a sit-
uationally relevant ego-attitude, whose
participation in ongoing psychological activity
generates modes of behavior that are more
consistent, more selective, and more charac
teristic of the person in that respect.37
This definition permits the researcher to
distinguish degrees of consistency as well as degrees of
selectivity of the behaviors chosen. Therefore, the
provided definition could be useful in specifying the
needed rationale for determining the degree of involve
ment of the subjects being examined under certain
specific conditions.
Sherif and Sherif reported that the higher the
priority or rank of an ego-attitude of the self system
that is triggered by the given situation, the more
consistent and selective the subject is. Operationally,
the researcher could then infer the degree of involve
ment from degrees of selectivity and consistency of
behavior of different individuals in different situations.
In order to be more precise and to express the
degree of ego-involvement of a subject on a specific
^^Muzafer and Carolyn W. Sherif, Social 'Psychology
(.New York: Harper and Row, 1969), p. 38 7.
issue, the Shérifs expressed this relationship in terms
of a pattern of acceptance, rejection and noncommitment.
This categorization, they believed, was appropriate when
analyzing the dimensions of attitude:
Operationally, the degree of a person's ego-
involvement in ongoing psychological activity
is inferred from the relative magnitudes of
his latitudes of acceptance, rejection, and
non-commitment in categorizing the relevant
stimulus domain: The greater the size of the
latitude of rejection relative to the latitude
of acceptance and the latitude of non-commitment,
the greater is the person's ego-involvement in
that stimulus domain.38
This procedure is known as the "own categories"
procedure in attitude research. Using this method, atti
tude is conceived as a set of evaluative categories
which the individual had either formed or learned during
his interaction with persons and objects in the social
world. This approach began with a search by Sherif and
Sherif for specific reference scales which would yield
information that was psychologically valid for the study
of attitudes and attitude change. A wide range of
studies using this method have been completed since its
development in 1948. They have indicated that this
measurement was valid for what it was intended to do.
That is, the ranges of positions that an individual
advocated, rejected, and toward which,he remained
, p. 3
29
noncommital (when not required to evaluate every position)
would vary systematically according to his involvement
with the issue.
More specifically5 Sherif and Sherif found that
individuals who were highly involved with a certain stand
on a social issue would tend to reject many more positions
than they accept, and would quickly weigh every position
as either acceptable or objectionable. Also, pro
portionally to his lack of involvement with an issue, the
number of positions the individual would accept and
reject would become approximately equal as his latitude
of noncommitment increases. Quite clearly this could
be interpreted as meaning that persons who were highly
involved would tend to have a much broader latitude of
rejection than persons less concerned, and, in turn,
they would remain noncommital toward fewer positions on a
specific issue, even when they were not required to
evaluate all of them, A division of subjects into high,
and low involvement categories must account for these
latitudinal characteristics, A useful set of distinctions
for accomplishing this has been established by Sereno and
Bodaken, For scores ranging between 4 and 28:
Highly involved subjects were operationally
defined as those negative, extreme subjects
having a minimum latitude of rejection oi'
nonacceptance of 16— at least four of the
30
seven possible positions on each of the four
pairs of scales to be unacceptable— and a
maximum latitude of acceptance of 4.^^
Proportionally similar criteria were used to
break subjects into three groups in this study so that a
3x3x3 factorial design could be effectuated. Since
three, instead of four, sets of adjectives (as listed on
page 61), were used in the research, high involvement was
defined as being a minimum latitude of rejection of 12
and a maximum latitude of acceptance of 3.
A number of studies completed over.’ the past
several years have made a marked contribution to the
literature of ego-involvement.^^ These studies have
contributed to the development of a conceptual approach
that does account for the distinctive patterning of
latitudes of acceptance and rejection of this method as
well asTsubstantiating the theoretical base of the "own
categories procedure."
One method of working the own categories
procedure, which incorporated the use of specific
39
Kenneth K. Sereno and Edward M. Bodaken, ''Ego-
involvement and Attitude Change: Toward a Reconceptualiza
tion of Persuasive Effect," Speech Monographs,39:151-158,
August, 1972.
^*^Muzafer Sherif and Carolyn Wood Sherif, "The Own
Categories Procedure in Attitude Research," Readings in
Attitude Theory and Measurement, ed. Martin Fisbhein
(New York: John Wiley and Sons, Inc., 1967), p. 192.
31
differential scale was developed by Diab.*^^ This method
was particularly useful because of: (1) the sharp reduc
tion in the time required to develop the instrument, and,
(2) the reduction in the time required to administer it.
The utilization of a semantic differential scale,
which incorporates interval scales permits the application
of powerful statistical techniques.
When a researcher utilizes the Diab method, sub
jects are asked to rate a particular concept on a number
of semantic differential scales that represent the main
factors of meaning (evaluative, potency, and activity)
to the particular concept being questioned. The difference
however, is that the subject is asked to indicate the
position most acceptable to him as well as other positions
that are also acceptable. Each subject's score is
the mean value of the latitude of acceptance of the
particular issue in question. In addition, the same
subject is also asked to indicate on each scale the
positions that are objectionable. Coded symbols are used
for identification by each subject to express his
preferences. By utilizing this method, each subject has
then provided ratings of a particular concept in terms of
Lufty N. Diab, '.'Measurement of Social Attitudes,
Problems and Prospects," Attitude, Ego-involvement and
Change, eds. Carolyn W. Sherif and Muzafer Sherif (New
York: John Wiley and Sons, Inc., 1967), pp. 140-158.
32
its acceptability, rejection, and noncommitment that are
easily indicated by an average noncommitment score.
Accordingly, the researcher would then evaluate the score
of each subject on a particular issue in terms of dif
ferences between subjects with, for example, identical
extreme positive or negative stands respective to their
latitudes of acceptance, rejection and noncommitment on
that particular issue.
The theory of ego-involvement has been utilized
as a means of contributing to attitude-change research
both by itself and in conjunction with a number of other
communication variables. For this reason, it is often
difficult to classify research in this area as strictly
ego-involvement research. An attempt was made to provide
a reasonably comprehensive review of articles which were
important in the development of this research while
realizing that additional variables relevant to this
study would also be simultaneously covered, especially
concerning interaction effects,
In addition to his pioneering work on source
credibility and message discrepancy which were discussed
earlier, Hovland, in conjunction with Harvey and Sherif,
also produced one of the earliest experiments in the
area of ego-involvement in which, the phenomenon of
33
assimilation-contrast was explored. The study used a
controversial issue and presented varying stands or posi
tions on the issue, with each subject identifying the
stand which, came closest to his own thinking, then with
stands with which he could agree, and finally, those
which he found objectionable. Several important
principles; emerged from the findings:
1. When the distance between S's own stand and
the position advocated in communication is small,
the communication is judged favorably, as fair
and factual. With increasing distance, the
favorable reaction is sharply reduced and the
communication perceived as propagandistic and
unfair.
2. The S's whose own stands diverge widely
from the position advocated perceive the com
munication as. further removed from their own
stand than it is (contrast effect). The
present results indicate, though less clearly,
that S's whose own stands are close to the
position advocated perceive the communication
as closer to their own stand than is the case
(assimilation effect).
3. The most frequent result for Ss whose
own stand diverges widely from that advocated
in communication is to remain unchanged in
their initial attitudes. More Ss with moderate
positions closer to the stand in communication
changed in the direction advocated.^2
These three principles, then, laid the groundwork
for understanding what cognitive reactions are to be
expected when messages of varying levels of discrepancy
ho
Carl I. Hovland, 0, J. Harvey, and Muzafer
Sherif, "Assimilation and Contrast Effects in Reactions
to Communication and Attitude Change," Journai of Abnormal
and Social Psychology, 55:242-2 52 , September 19.5 7 , p. 2617
34
are presented. This was why the variable of ego-
involvement, which clearly laid out the nonlinear nature
of attitudes, was examined in relation to the other
variables in this study.
Zimbardo relied heavily on the work of Hovland
in his 19 6 0 experiment. He was one of the first
investigators to combine the study of message discrepancy
with the timely work being done by Hovland, Sherif, and
others on the assimilation-contrast phenomenon, or as it
has since been labelled, level of involvement. Zimbardo,
perhaps incorrectly, attempted to reconcile the
conflicting theories of Festinger and of Hovland,
Sherif, et al. by making a prediction that he felt would
be consistent with both. In his words:
This research studied the relationship between
conformity and (a) the extent of the discrepancy
between the opinions of a communicator and a
recipient and (b) the degree of involvement of
the recipient. These variables are central to
a dissonance theory analysis of the social
influence process, as well as to many previous
investigations of attitude change. It can be
derived from the theory that opinion change
increases with increases in both involvement
and discrepancy between communicator and
recipient.
It was not found to be true that the variable of
involvement was central to dissonance theory analysis.
4 3
Philip G, Zimbardo, "Involvement and Communica
tion Discrepancy as Determinants of Opinion Conformity,"
Journal of Abnormal and Sociai Psychblbgy, 60:86-94,
January, 1960, p. 93.
35
Further, it was never a tenet of involvement theory that
. . opinion change increases with increases in . . .
44-
involvement In fact, just the opposite, was
true. Zimbardo's findings supported his interpretations
of dissonance and involvement theory. It was found that
highly involved subjects changed significantly more than
did "Ss not involved," Also, the larger the message
discrepancy, the greater the change in stand incurred,
and "the maximum dissonance group changed significantly
more than the other groups, while the minimum dissonance
45
group consistently changed least,"
Turning now to more recent studies on ego-
involvement, Sereno and Mortensen attempted to determine
whether dyads consisting of subjects who were slightly
involved would reach agreement more often than would
highly involved subjects. This was in fact the case. In
addition, a second hypothesis that the low-involvement
dyads would more frequently abandon their initial stands
was also supported. Both findings confirmed the
efficacy of ". . . the distinction between extremity of
attitude position and the intensity of attitudinal
commitment.
^Zimbardo, p. 9.3.
^^Ibid.
^^Kenneth K. Sereno and C, David Mortensen,"The
Effects of Ego—Involved Attitudes on Conflict Negotiation
in Dyads," Speech Mohographs, 36:8-12, March, 1969.
______________________________________' _________________________36
Atkins and Rieri discovered that by heightening
the awareness of the attitude of the sub]ects toward a
particular topic, the effects of the assimilation-contrast
could be enhanced. For example, subjects who were highly
involved already perceived discrepant communications as
more distant from their positions when they had, prior
to the measurements, engaged in a heated discussion about
the involvement topic than did subjects of similar involve
ment who had not engaged in such discussions,. The topic
was organized religion. The findings were not completely
in agreement, however, because for the subjects who had
anti-religious sentiments, the contrast effects were not
nearly so definitive. Of this discrepancy, Atkins and
Bieri state:
Although there is no simple means of recon
ciling these differences, the data do suggest
that pro-ness and anti-ness of own attitudes
may have different properties, despite the
fact that both represent extreme positions
along the same continuum.
This observation was important because it brought
up the consideration that ego-invovlement theory could
still have some weaknesses which should be explored in
order to reconcile otherwise unexplainable findings.
In another investigation of the value of ego-
involvement theory in making predictions about certain
perceptions of subjects, Mortensen and Sereno predicted:
h n
Alvin L. Atkins and James Bieri, "Effects of
Involvement Level and Contextual Stimuli, on Social Judg-^
ment," Journal of Personality ahd Social Psychology, 9:
197-204. June. TW6W7 -p. 2U2. ^
1. Highly involved subjects will perceive a
person advocating the discrepant stand as less
credible than will subjects who are slightly involved.
2. Highly involved subjects will differ
from slightly involved subjects in the degree
to which they are predisposed toward communi
cating with a person who advocates a discrepant
stand.
3. Highly involved subjects will have lower
expectations of reaching consensus with someone
advocating a discrepant stand than will slightly
involved subjects.
4. As the magnitude of discrepancy becomes
larger, there will be greater differences between
highly and slightly involved subjects' percep
tions of credibility, prédisposions toward g
communication, and expectations of outcome.
The first three hypotheses were not supported, while
the fourth was mildly supported. Thus, as with the
Atkins and Bieri study, the findings exposed possible
weaknesses which could arise while using ego-involvement
theory. Mortensen and Sereno offered the explanation that
possibly the operationalism of discrepancy on three
levels, which necessitated using only subjects who
endorsed extreme attitudinal positions could have
obliterated the distinction between highly and slightly
involved subjects because . . most people endorsing
4 9
extreme stands are highly involved." An additional
Kenneth K. Sereno and C. David Mortensen, "The
Influence of Ego-involvement and Discrepancy on Perceptions
of Communication," Speech Monographs, 37:127-134, June,
1970.
49
Ibid. , p. 134.
38
possibility for error was the chance that subjects were
not perceptive enough to distinguish: various levels of
discrepancies 3 which would mean that it would be unlikely
that level of involvement would make any difference on
their judgments. These are problems which were avoided
in the present study by using different measurement
techniques for the variables considered.
A study which exposed the still relatively unknown
nature of the changeability of latitudes of acceptance,
rejection, and noncommitment, was that of Sereno and
Bodaken.Exploring the effects of the high discrepancy
message, they hypothesized that such a message would
decrease latitudes of rejection, increase latitudes of
acceptance, and increase latitudes of noncommitment, while
at the same time having no effect on the most acceptable
position of the subjects. It was found that the subjects,
all of whom were highly involved, exhibited a decrease
in their latitudes of rejection and an increase in their
latitudes of noncommitment. Consequently no increase
took place in the latitudes of acceptance, and the most
acceptable positions, significantly changed. Neither of
these latter two findings was expected. Additional
research was needed to substantiate the exact nature
158.
^^"Ego-involvement and Attitude Change, , .7 pp. 151-
39
of changes in latitudes when discrepant messages were
presented. It is hoped that some light was cast on
this problem by the present study.
As support for the second hypothesis, that
source credibility had its greatest effects on voting
behavior 5 of lowly-involved subjects, Eagly and Manis
found that "involved subjects were more negative in
their evaluations of messages and communicators than
were non-involved Ss."^^
In another investigation, Freedman examined the
involvement-discrepancy relationship, and found
that :
Under low involvement, there was more
change with greater discrepancy; but under
high involvement, the relationship was non
monotonic 3 with maximum change occurring
at moderate discrepancy.^2
This relationship between ego-involvement, discrepancy,
and amount of attitude change was presented in graph
form :
Alice H. Eagly and Melvin Manis, "Evaluation of
Message and Communicator as a Function of Involvement,"
Journal of Personalitv and Social Psvchology, 3:483-485,
April, 1966, p. 483.
52
Jonathan L. Freedman, "Involvement, Discrepancy,
and Change," Journal of Abnormal and Social Psychology,
69:290-295, September, 1964, p. 290.
40
Amount of
Attitude
Change
3 .
2 .
1 .
Low : -
Involvement
High
Involvement
High Moderate Low
Amount of Discrepancy Change
Figure 2
Change in Attitude Against Change in Discrepancy
For low involvement subjects, attitude change
continually increased with increases in message
discrepancy. For high involvement subjects, however, a
critical point was reached at moderate message
discrepancy where the message rejected and the attitude
change became less likely.
The treatment conditions for the present study
were those indicated by the letters A and B inserted in
Figure 2. Therefore, the hypothesis presented was that
source credibility would have had its greatest^effects on
voting of lowly-involved subjects, since over this entire
range the lowly-involved subjects experience greater
attitude change. Furthermore, the difference in attitude
change between lowly and highly involved subjects was
greatest at the low and high discrepancy points A and B.
41
The last work to be discussed in this review was a
study by Johnson and Scileppi which manipulated all of
the variables with which the present investigation was
concerned. In a 2x2x2 factorial design which varied
ego-involvement5 source credibility, and message plausi
bility, Johnson and Scileppi found that :
. . . results indicated greater attitude '
change in low-ego-involvement-high-source-
credibility conditions than in the other
three combinations of source credibility
and ego-involvement.53
This led them to conclude that source credibility
operating primarily under low ego-involvement
conditions. It will be recalled that the second hypothesis
of the present study stated essentially this same idea,
but transformed it to the realm of political
advertising.
Integrative Summary of Pertinent Literature
The literature relevant to the area discussed in
this proposal fell into two categories: practical informa
tion and scholarly information. The practical informa
tion consisted of the opinions of people who have had
everyday experience in attempting to persuade voters that
they should vote for a particular candidate. These
53
Homer H. Johnson and John A. Scileppi, "Effects o
Ego-involvement Conditions on Attitude Change to High and
Low Credibility Communicators," Journal of Personality and
Social Psychology, 13:31-36, September, 1969, p. 31.
42
individuals generally agreed that it was more important
how a candidate appeared in the media than it was to
present a certain point of view relevant to a specific
issue. Simply stated, the image was more important than
the issues. The overwhelming opinion exuded by these
practitioners was that a candidate must demonstrate a
personality which was warm, feeling, decisive and
seemingly honest. Although a pleasant physical appearance
was also helpful, it was not mandatory.
The scholarly information reviewed for the. purpose
of this study did not have the same widespread external
validity achieved through practical applications of the
ideas presented that the practical information did.
Instead, it attempted to offer considerable refinement
of what were only crude, gut reaction feelings of the
campaign experts. It also provided a basis for making
predictions or formulating specific hypotheses regarding
the image-issue question.
It was demonstrated that trustworthiness and
competence were two viable cognitive dimensions, rein
forcing what the media experts had said. These
dimensions were refined in a study by McCroskey who added
a set of factor-analyzed bi-polar adjectives which were
adopted for the study. Because most of the scholarly
information has been evaluated only in an experimental
environment, it might be conjectured that some results
_________________ 4 3
have been conflicting. Wyer and Schwartz, for instance,
found that source credibility had a negative effect on
persuading the audience. The converse was found by
Hovland and Weiss:. That is, the more positive the source,
the greater the persuasive impact of the communication.
A refinement of the question and resolution to the
difficulty of contradictory results was offered by Bochner
and Insko who found that source credibility was only
meaningfully examined when the discrepancy level of the
experimental communication was also considered. This
approach., along with the incorporation of the additional
variable of ego-involvement, was taken by Freedman. He
was able to plot the relationship of two of these
variables to graphically and visually show how they co
varied. This graph revealed that the real persuasion
nearly always took place in low involvement subjects.
Furthermore, high-discrepancy was most effective in
initiating persuasion.
One emerging principle was that there was still a
great deal of confusion and disorganization in the various
approaches to answering questions about the variables
which were examined in this study. An area where this
was especially true was that regarding the conflict
between dissonance theory and ego-involvement theory as
outlined by Scott, Because of the extensive research
supporting ego-involvement theory and because dissonance
theory was often impossible to make predictions from,
________________________________ 44
ego-involvement was used as a basis for thxs study.
This decision, in turn, led to an examination of the
measurement techniques validated by Sherif and Sherif for
measuring ego-involvement including the discovery of
latitudes of acceptance, rejection, and non-commitment
in addition to the most favorable position traditionally
used by other attitude researchers and the modification
of this technique.by Diab.
Because of the dynamic nature of communications
research., it was generally more productive to analyze
several interacting variables together rather than to
just look- at any one variable. This approach was taken
by Zimbardo when he jointly investigated the effects
of ego-involvement and message discrepancy. In a
different approach Sereno and Mortensen, in two studies,
focused primarily upon the characteristics of ego-
involvement itself, though treating it as an assigned
variable. Atkins and Bieri successfully manipulated
the variable by having a heated discussion about the
involvement topic with subjects before they were tested.
Finally, in a recent study by Sereno and Bodaken a
successful attempt was made to manipulate the latitudes
of rejection and noncommitment through presentation of
belief diecrepant communice.tions.
Freedman moved toward a more comprehensive; examina
tion of the dynamic communication process with his study
______________________________________ 45
of message discrepancy, ego-involvement, and attitude
change all of which were relevant to the present study
by providing a useful plotting of the three variables in
increments. All of the variables relevant to the present
study, source credibility, message discrepancy, and ego-
involvement, were brought together and examined in the
final study reviewed. In the research of Johnson and
Scileppi, it was found that low involvement-high-
credibility was the most effective means of persuasion.
When this was combined with the low-involvement-high-
discrepancy findings of Freedman, the following synthetical,
inference was possible; the most persuasive message was
one which combined high source credibility, low involve
ment, and high message discrepancy. All of these
variables were incorporated into the persuasive treatment
used.
46
Chapter 3
OPERATIONAL PLAN
Operational De fin ition s
Four key terms must be defined for the purposes of
this study. The first term is ■■source credibility.” In
this research, source credibility meant the degree to
which a "source” or in this case, a candidate, is believed
respected, and able to motivate his audience to respond
to such requests as voting for him. This credibility
can vary from high to low, depending on a number of
factors or characteristics of the candidate himself and
his presentation. A highly credible source is one which
creates a favorable image. The important aspects of
presenting a favorable image include the physical
appearance of the candidate. Is he, for example, over
weight, haggard, unkempt, or in any other way visually
unattractive? Another aspect of this dimension is his
personality or ability to perform before the camera.
Is the candidate, for example, indecisive, unsure of him
self, hesitant, too slick, or stilted? In short, any
one or a combination of the aforementioned characteristics
can cause a decline in credibility. Therefore, to present
any kind of a favorable image, the candidate must present
both a positive physical appearance and a dynamic
47
personal performance. The actual people: used as
candidates were judged by a panel of subjects that
represented the larger population which was later tested
to determine if the abovementioned qualities were, in
fact, identified.
The second key term to be defined is "ego-involve
ment.” For this study, ego-involvement meant the
"relevance, significance, or meaningfulness that an issue
or topic has for an individual.”^ The greater an indi
vidual * s involvement, the less he is able to tolerate
opposing points of view. The level of ego-involvement
was obtained by measuring this variable using a Diab
2
scale. High involvement meant that the respondent
had a large latitude of rejection and a small latitude
of acceptance on a scale of one to seven. That is, there
were many attitude positions unacceptable to him, and
^C. David Mortensen, Communication ; The Study of
Human Interaction (New York: McGraw-Hill Book Company,
1172), p. 164.
2
Diab provides a brief explanation for his scales
in Lutfy N. Diab, "Some Limitations of Existing Scales in
the Measurement of Social Attitudes,” Psychological
Reports, 17:42 7-430, October, 1965. Further developments
and validation are presented in Lutfy N. Diab, "Studies in
Social Attitudes: I. Variations in Latitudes of Acceptance
and Rejection as a Function of Varying Positions on a
Controversial Social Issue,” Journal of Social Psychology,
67: 283-29-5, October 19.65, and in Lutfy N. Diab, "Studies
in Social Attitudes: III, Attitude AssessmejiL Through the
Semantic Differential Technique,” Journal of Socia1
Psychology, 67:303^314, 0ctober, 1965.
48
only a few acceptable to him, when he rated bi-polar adjec
tives as being descriptive of a particular concept or
3 .
attitude object. It is through this method that subjects
were divided into high, medium, and low involvement groups.
The third key term to be operationally defined is
"message discrepancy." For this study, message discrepancy
was the difference between the message presented (i.e.,
positive or negative viewpoint) and the attitude of the
subject on the topic discussed in the commercial. This was
determined by having a panel of judges who represented the
population to be tested later to rate the television com
mercials. Then, the greater the difference between the
rating scores by the judges and the subsequent ratings of
the commercials by the subjects, the greater the degree of
discrepancy for each subject.
A final important term to be defined is that of
a "political issue." For this study, it was defined as a
major topic of discussion, as ascertained from conversa
tions with political experts, which was expected to be of
significant public interest for a reasonable period of
time. Ideally, the topic was felt to play a key role in
future local, state, and/or national elections. Four
such issues were used in this study. (See Appendix A).
3 . . .
Exact numerical limits for these latitudes is dis
cussed later under Ego-involvement section of Review of
Literature, as are sources from which they are obtained.
49
Nine groups were used, each of which.received
a different treatment. The experimental method was
selectëd becauee it answered tbe specific questions posed
regarding the persuasive effects of a communication and
because it, at the same time, provided strong external
validity.
Op erational Hypothèses
For convenience, a 3x3x3 factorial design was used
with the three frichotomized variables being ego-
involvement, source credibility, and message discrepancy.
The two hypotheses tested with this design were :
1. A high source credibility candidate who
presents a discrepant message will receive significantly
more votes than will a low source credibility candidate
who presents a consonant message.
2. The high source credibility candidate who pre
sents the highly discrepant message will receive the
greatest number of votes from lowly involved subjects.
Each of these hypotheses were tested using the
F test at the .05 level of signifiance. Omega squared
tests were performed on the main effects and interaction
effects which proved to be significant at the .05 level
in order to determine the amount of variance due to
these effects.
As a further explanation of the assigned vari
ables of this study, the following list was utilized:
50
VARIABLE DESCRIPTION VARIABLE RANGE
1. Discrepancy for the evalu
ative dimens io n 0-18
2. Discrepancy for the activity
dimension 0-18
3. Discrepancy for the potency
dimension 0-18
4. McCroskey*s source credibility
rating on the authority dimension 6-42
5. McCros:key-s source credibility
the character dimension 6-42
6. Ego-involvement 0-1
Each of these variables was not manipulated, but
was instead, used for correlational purposes for
dividing subj ects into the groups in the factorial blocks
based on variable scores. Also provided in the above
table are the numeric ranges along which the variables
could fluctuate. It was noted that the discrepancy
variable was actually three distinct variables, one for
each dimension. Two dimensions existed for the
McCroskey scale. The ego-involvement variable is
dichotomous with either high or low involvement, the
criteria for which were outlined previously. The
dependent variable which was manipulated and which was
not in the above list was the mock ballot or voting
behavior.
Procedure for Testing the Hypotheses
The following list of procedures were used:
Three sets of television commercials were written and
51
produced at the one-inch video studio facilities of
California State University, Fullerton. Actors were
obtained from the Theatre Arts Department, one for each set
of commercials. In one version the candidate presented a
low discrepancy'message, but had low credibil ity. In a
second version, the candidate presented a neutral message,
and had medium credibility. In the third version, a high
credibility candidate presented a very discrepant message.
The messages and the candidates for each of the three sets
of commercials were rated by a panel of ten student judges,
who comprised the population which was later tested.^. Each
message and each candidate was rated by the judges using a
one-to-three scale. Instead of being asked to rate the
candidates and issues on how they personally felt, as was
the case for the subjects, the judges were asked to provide
an objective evaluation based on their beliefs of the
attitudes of their fellow students, who would later be
exposed to the televised messages as subjects. This evalu
ation consisted of ranking the commercials on a scale of
one to three according to the panel?s judgment as to the
popularity of the issue stands.
This insured that the commercials reliably pre
sented issues on which there were divided levels of ego-
involvement and candidates who actually differed in their
image appeal.
^Results of this pilot examination may be found in
Appendix A. -
A 3x3x3 factorial design was utilized in this,
research-. Two of the, varia hi es involved were independent
variables, wliiie a third variable was an assigned variable
The two independent variables were; source credibility
and message discrepancy, while the assigned variable was
that of ego involvement.
The method of manipulating the independent
variables was through the medium of a television com
mercial. Source credibility was manipulated by thetype
of "image" of each candidate who made a presentation.
Those candidates were designated as having images that
were a "pretty boy," "middle boyor "ugly boy."
In terms of creating images, the "pretty boy"
candidate was a candidate who had a "Paul Newman" type of
demeanor, who was. handsome and spoke positively.and was
well-dressed. The "middle boy" candidate had no out
standing physical attributes either positive or negative.
He spoke in an unobtrusive manner with a low key
presentation. His manner of dress was average. The
"ugly boy" candidate had a very slovenly physical
appearance. Furthermore, he spoke in a halting fashion
with great hesitation. His voice was squeaky, and his
manner of dress was shabby and unattractive.
In order to establish that the Candida Les did,
in fact, appear to be the subjects as was expected,
namely either pretty, neutral or ugly, they were rated
___________________________________________ 5 3
on a scale of one to three with one being extremely
attractive and three being unattractive. The actual
ratings by the panel proved to confirm the selection of
actors, The pretty boy candidate had a mean rating of
1,2 with little variance due to the type of stand he was
presenting. The middle boy candidate had a mean rating
of 2.0, and the ugly boy candidate had a mean rating of
2.5.
The second independent variable, that of message
discrepancy, was manipulated according to the type of
message which was presented. There were three types of
messages, each coinciding with a particular stand on a
particular issue, either popular, neutral, or unpopular.
The issues from which the various stands were deduced,
were determined from various major news issues which were
confronting voters in California at the time the study
was conducted.
The third variable used in this study, that of
ego involvement, was an assigned variable. Subjects who
were part of nonrandomized intact class groups were
assigned to the various cells of the factorial design,
based on their level of ego involvement. The various
levels of ego involvement were assigned according to
strict numeric criteria. High ego involvement consisted
of a latitude of rejection, greater than or equal to
twelve, and a latitude of acceptance less than or equal
........................................... hlL
to three. Medium ego involvement consisted of a latitude
of rejection less than twelve, and greater than or equal
to six, plus: a latitude of acceptance less than or equal
to six and greater than three. Low ego involvement was
defined as a latitude of rejection less than six and
latitude of acceptance greater than six. These scores
were derived through the use of the Diab scales in the
following manner: for each issue, there were three
scales of seven positions each. In each position of the
scale the subject had the option of marking as either the
most favorable, acceptable, or unacceptable. For each
of the three scales the number of acceptable and unac
ceptable positions were tallied in order to determine the
size of the latitudes of acceptance and rejection.and,
thus, the ego involvement of each subject.
2. A total of 94 8 subjects were used for purposes
of this research. These students were in classes at
California State University, Fullerton.
The distribution by class level of the students
participating in this experiment was relatively even.
There were 14 5 freshmen, 175 sophomores, 16 7 juniors,
28 5 seniors, and 175 graduate students. The distribution
was also quite even for the variables of sex with 4 87
males and >170 females . The respondents V*ages were skewed
towards a higher age category with nearly forty-five
percent of all subjects being twenty-four years of age
------------------------------------------------------------------- S l E-
or older. Only twenty—five percent of those tested were
age twenty or under. A surprising number of the students
were registered voters. The frequency was; 742, or
76 percent of 9 74 total respondents. When asked what was
their political party affiliation, 268 respondents replied
they were Republicans, 402 were Democrats, and 9 8 were
registered in other categories. A rather large unexplaine
nonresponse was evident with 20 6 declining to state their
party affiliation.
The'mean age of the respondents was 22.7 years.
Almost twenty-five percent stated that they had made a
donation during the last Presidential election. These
last four variables of age, registration, party affili
ation, and campaign donating all tended to lend credi
bility to the study in that they, at least, modestly
reflected a general political concern and a similarity
with demographics found in the general population.
In addition, 75 percent of the respondents stated
that they voted in the 19 72 Presidential election. Nixon
was the choice for 332 of the respondents, 229 voted for
McGovern, and 40 for other candidates. Nearly every
respondent had a television set in working order. When
asked their family income, : ten percent stated that it was
under five thousand dollars per year, fourteen percent
said five thousand to ten thousand, forty-one percent
stated that their income was between ten and nineteeh
56
thousand per year, twenty percent stated it was between
twenty and twenty-nine thousand dollars per year, and
ten percent said their income was thirty thousand dollars
a year or over.
The racial composition of the subjects tended to
be skewed towards the White race. There were only forty-
one Blacks, fifty—one Mexican Americans, and twenty-four
Asians in the sample, due likely to the fact that the
research was conducted in Orange County, a predominantly
Caucasian inhabited county.
3 . ' . ' . Procedures" for producing television com
mercials: All commercials for purposes of this study
were produced at the one-inch video facilities of the
Theatre Arts Department at California State University,
Fullerton. Three cameras were utilized in the production
of the commercials. One camera was set on a tight
close-up shot (TCU), while a second was set on a medium
shot CMS). A third camera was set on a wide angle shot
(..cover shot) . No background music was utilized in the
production of these commercials. An interviewer, whose
face was never seen, was utilized to ask the questions.
A prerecorded announcement, asking voters to vote for
each candidate, was utilized at the end of each com
mercial. The. set used for the production of these
commercials was very simple. A small background flat,
painted in a neutral color was used. Each, candidate sat
__________________ 57
in a comfortable-appearing' chair, facing the interviewer.
There were no variations in production for purposes of
this study. That is, the same interviewer was used for
each treatment, as was:the same voice-over announcer.
Camera shots were the same for all treatments.
4. Subjects also read a simulated newspaper
article about another fictitious candidate which was the
same for all cells. This simulated newspaper article
may be found in Appendix B.
The newspaper advertisement was formulated in a
Style which was adopted from an advertisement by a then
prominent California gubernatorial candidate. It
advertised the candidate ^s qualifications, but purposely
made no mention of issues so as to provide a neutral
control.
The simulated newspaper article was utilized to
adjust for possible cell differences which could have
resulted by using intact class groups. While an ideal
solution to counteract the nonrandomization of subjects
would have been to randomize treatments, this was not
possible due to the type of treatments that were used in
the study. Specifically, it would have been difficult
to ha.ve a television monitor for each subject, since over
9-0-0 subjects were used in the research. Therefore, a
control was deemed necessary to demonstrate the compara
bility of the cells; and to adjust for any differences
which have occurred. A computation was made to isolate
58
the actual differences in voting due to the treatments by
subtracting the score or vote for the candidate shown in
the newspaper advertisement.
This control candidate was subjected to the same
type of testing as was the television candidate. He was
rated on the McCroskey scales. Subjects then voted for
him just as they did for the ^television candidates. The
scores could range from one to four with one meaning the
subject definitely would vote for him and four meaning
the subject definitely would not vote for him. For each
subject, this score was subtracted from the identical
type of rating for the television candidate. Since this
could possibly result in a possible negative number, a
constant of three was added to this difference in order to
make all scores positive so as to facilitate the data
analysis.
5. A pilot test was conducted to determine
if the measuring techniques were adequate for the study.
Twenty-five subjects were shown commercials depicting
the "pretty boy" presenting discrepant messages on all
of the pertinent issues to be tested. These commercials
were inserted in a 30-minute television program called
"That Girl," featuring Mario Thomas. The program "That
Girl7 was used in this study in order to simulate reality
as closely as possible. It was determined that there
were no substantive changes, necessary in the operational
_______________ 5 9
plan. That is to say , the questionnaires: were, not
vague, and the format of presentation was clear. There
fore, the pilot results were incorporated into the actual
study for economy.
6. A video tape recorder and monitor were brought
to the intact classes so the subjects could view one of
the nine sets of commercials.
7. After the commercials were viewed and the
article about the fictitious candidate was read, subjects
were asked to cast a mock ballot expressing their opinion
about the two candidates to which they had been exposed.
Finally, they filled out the questionnaires measuring
their ego-involvement, discrepancy, and credibility levels
Sheriffs Own-Sort Categories method, which was
described earlier, as adapted by Diab was selected to
determine the levels of ego-involvement as well as
5
attitude scores. Following the administration of the
treatment, each subject was instructed on the proper
technique to complete the Diab scales. These scales
contained nine pairs of randomized bi-polar adjectives.
These adjectives and loadings, as established by
Osgood, were as follows:
^A copy of the test questionnaire is located in
Appendix B,
’ 6 0
Adjectives Dimension Factor Loading^
good-bad Evaluative 0.8 8
beautiful-ugly Evaluative .86
clean-dirty Evaluative .82
sharp-dull Activity .52
active-passive Activity .59
fast-slow Activity .70
large-small Potency .62
strong-weak Potency . 62
heavy-light Potency .62
The Diab scales needed no review after the pilot
test.
7
The mock ballot represented voting behavior.
It was felt that the use of a mock ballot was reasonably
representative of actual voting in an election. The
same technique of using a mock ballot had been success
fully used by professional research firms hired by the
Republican Party during the 1972 Presidential campaign.
The technique is described in an article in Advertising
Age as follows:
. . . instead of asking a respondent how he
intends to vote, he is given a ballot to fill
out in privacy and then deposit in a sealed
container. This procedure, proponents argue,
leads to a more honest, uninhibited reply.
Therefore, the mock ballot was preferred over strict
reliance only on traditional attitude scales.
^Osgood and others. The Measurement of Meaning, p. 37.
7
A sample of the mock ballot may be found in
Appendix C.
^Jack J. Honomichl, "Nixon^s * 72 Election
Researchers Stressed Interviews, Polls Phone," Advertising
Age, June 4, 1973, p. 95.
61
The mock ballot (’ dependent variable) was scored in
the following manner: A value of 1 was assigned to those
subjects who would definitely vote for the candidate, 2 to
those who would probably vote for him, 3 to those who
would probably not vote for him, and 4 to those who iwould
definitely not vote for him. This range of 1 to 4 ; ■ i :
permitted interval scale analysis. This scale was applied
to the control candidate as well as to the television
candidates.
7. After testing had taken place, the question
naires were gathered. Data were transferred to computer
tape. The tapes were then analyzed at the University
of Southern California Computer Center using programs
written in SPSS and BMD.
Limita tion s
One limitation was due to the necessity of using
the classroom, television., equipment and questionnaires.
Each of these three factors likely produced some
distortion in results, as opposed to the real-life situ
ation in which these factors are non-existent. Theo
retically, the experimental design used controls for
such distortions. Unfortunately there was no means to
control for interaction effects between thase factors and
the dependent variable. Despite these limitations it
62
was believed that the study could be of great use
since it was considerably more realistic than any other
study to date.
63
CHAPTER 4
PRESENTATION AND ANALYSIS OF DATA
Questions and Hypotheses
This research conducted during May and June of
19 74 evolved from two principal questions, from which two
operational hypotheses were deduced. The questions posed
were ; (1) Is the influence of source credibility, or
image, stronger than that of the message being presented;
and (.2) will source credibility have its greatest effects
upon the voting behavior of lowly involved subjects? From
these two basic questions, two research hypotheses were
postulated. The first hypothesis stated: A high source
credibility candidate who presents a discrepant message
will receive significantly more votes than will a low
source credibility candidate who presents a consonant
message. The second hypothesis, relevant to the area of
ego involvement, was stated as follows : The high source
credibility candidate who presents the highly discrepant
message will receive the greatest number of votes from
lowly involved subjects. This chapter deals with.the
findings of the study relevant to these two hypotheses
and their implications to past theoretical literature
that made the original hypotheses possible.
64
Hypothesis One
It can be stated without equivocation that, as
a result of the evidence of this study, hypothesis one
was confirmed. That is, the image of a political
candidate is much more important than what he verbalizes.
Even at maximum levels of message discrepancy, utilizing
unpopular messages that were basically totally discrepant
with the views of the majority of the population that
viewed the political television commercials, the message
was not as important in influencing votes as was the image
of the politician who delivered the message.
Table 1 presents the means and standard deviations
of the dependent variable (the vote). Table 2 presents
the same type of information for the voting scores as
Table 1, except that, in this instance, the data are
adjusted by the control vote in order to compensate for
inter-cell discrepancies. It was found that when the
scores were compared across the nine cells there were no
differences larger than one standard deviation apart.
That is to say, extraneous variables due to such factors
as nonrandomization of subjects did not seem to interfere
with the results or effects of the experimental manipu
lations. As Table 2 indicates, the mean for each cell
for the rating adjusted by the control vote, along with
the standard deviation of the rating for that cell, shows
the means ranging from a low of 3.1 to a high of 4.2,
65
Table 1
Means and Standard Deviations of
the Dependent Variable (.Vote)
(3e 11 Mean
Standard
Deviation
Pretty boy giving the most
favorable message 2.8 .78
Middle boy giving the most
favorable message 2.6 .62
Ugly boy giving the most
favorable message 3.4 . 71 .
Pretty boy giving the
mildly discrepant message 3,0 3.02
Middle boy giving the
mildly discrepant message 3.0 ,72
Ugly boy giving the
mildly discrepant message 3,6 , 53
Pretty boy giving the
most discrepant message 3.5 .69
Middle boy giving the
most discrepant message 3,6 2.72
Ugly boy giving the most
discrepant message
3.5 . 56
Scores ranged from 1 to 4 with 1 indicating that
subject definitely would vote for the candidate,
and 4 indicating that the suhject definitely would
not vote for the candidate.
66
Table 2
Voting Scores as Adjusted for Uifferences
In the Groups Who Viewed
The Commercials
Cell Mean
Standard
Deviation
Pretty boy giving the most
favorable message 3.2 1.0
Middle boy giving the most
favorable message 3,1 1.0
Ugly boy giving the most
favorable message 3.8 1. 7
Pretty boy giving the
mildly discrepant message 3,1 1.1
Middle boy giving the
mildly discrepant message 3,4 1.0
Ugly boy giving the
mildly discrepant message 4,1 0,9
Pretty boy giving the
most discrepant message 4.1 1.1
Middle boy giving the
most discrepant message 3.6 1.2
Ugly boy giving the most
discrepant message 4.2 0.8
Scores ranged from 1 to 4 with 1 indicating that
subject definitely would vote for the candidate,
and 4 indicating that the subject definitely would
not vote for the candidate.
67
while standard deviations average around 1.1. Therefore,
the range of scores are within a single standard deviation
and are thus acceptable in their level of fluctuation.
As Table 3 indicates, several significant dif
ferences were noted; Specifically, the message variable
F ratio was recorded at 16.9, while the candidate variable
F ratio was recorded at 38.1. Both of these F ratios are
significant. This finding verified hypothesis one that the
image of the candidate was indeed more important than was
the issue involved.
It should be also noted that the interaction effect
of the message variable with the candidate variable was
also significant at the .05 level with an F of 8.1 for
variable 1, although certainly not as high as for either
the candidate or message influence when studied alone.
Table 4 represents the control-vote-adjust F tests
and accompanying statistics for hypothesis one and two.
This information was derived through computer computation
and was defined as the difference between the candidate
vote and the control vote. Ego-involvement proved to be an
insignificant factor in considering voting behavior, as
both Tables 3 and 4 indicate nonsignificant F’s for this /■
variable of 0.9 and 3.7 (with a 4.7 necessary for signifi
cance at the .05 level). Neither was ego-involvement
significant as interaction variable with the message. The
omega squared test^ was used to test for the amount of
Iwilliam L. Hays, Statistics (New York: Holt,
Rineholt and_JLLnsJ:_on, J._9_6.37~7" pp . d 2 V - 3 2 9.___________________6J
: Table 3
Analysis of Variance;
Effects for Hypotheses
Simple Main
One and Two
.... Source. V.... . d F . ' . . . . . . S . S : . . . / V MS:. . . . . ...... F . . . .
Message 2 16.1 8.1 16.9*
Candidate 2 36.3 18.1 38.1*
Ego Involvement 2 0.8 3 0.42 0.9
Interaction of
Message. wxth_
Candidate 4 15.3 3,8 8.1*
Interaction of
Message with.
Ego Involvement 4 1.4 ,36 0.7
Interaction of
Candidate with
Ego 4
2.1 .55 1.2
3-way Interac-
tion of Message,
Candidate and
Ego Involvement 8 7.0 .88 1.8
Error 9 . 0 . 5 . 431.4 .48
.T.O.TALS...........9:31 . . . . . . 5 . 1 0 . 4 . 3 .
*P < . 0.5. Critical F =
Freedom. Critical F
Freedom, and Critical
3.02 with 2 and 91.5 degrees of ^
= 2. 39: with 4 and 905 degrees of
F - 1.9.6 with 8 and 905 degrees of
69
Analysis: of
Simple Main
Tahle 4
Variance: Control-
E f fects for Hypothe
Vote-Ad]usted
ses One and Two
Source dF SS MS F
Message 2 23.0 11.5 9.7*
Candidate 2 48.4 24.2 20.5*
Ego Involvement 2 8.7 4.3 3.7
Interaction of
Message with
Candidate
4 27.7 6.9 5.9 *
Interaction of
Message With
Ego Involvement 4 2.1 0.5 0.4
Interaction of
Candidate with.
Ego 4 10.1 2.5 211
3-Way Interact
tion of Message,
Candidate and Ego
Involvement 8 18.3 2.3 1.9.
Error 9,05 1069.5 1.2
TOTALS 931 1207.8
* P < .05. Critical F = 3.0.2 with 2 and 90.5 degrees of .
Freedotij Critical F = 2.39: with 4 and 9.05 degrees of
Freedomj and Critical F • = 1.9.6 with 8 and 9.0 5 degrees
of Freedom.
70
dependent variable variance produced by the main and inter
action effects which yielded significant F!s, The dependen
variable of candidate vote which was significantly influ
enced by the message presented showed a 2.96 percent vari
ance level when tested with omega squared. The influence
of the candidate upon the variance was better than double
that of the message with an omega square of 6.91 percent.
The interaction effect of message with candidate accounted
for 2.6 3 percent of the variance.
A similar trend was noticed when examining the
dependent variable of candidate vote-control vote
difference. A variance level of 1.71 percent was calcu
lated for the message variable, 3.81 percent for the
candidate variable, and 1.90 percent for the message-
candidate interaction effect.
Hypothesis Two
As a result of the analysis in this study, there
was no evidence to support hypothesis two. As noted in
Table 3, the F ratios for the variable of ego-involvement
are not significant at the .05 level. Ego-involvement
was not a significant factor in voting behavior, for high
ego involvement subjects were as likely to vote for a
candidate as were middle and lowly involved subjects;
therefore, the F ratio meant that subjects with low ego-
involvement were no more likely to be persuaded by either
the television message or the control message than were
71
the more highly involved suhj ectsv Thiis, source
credihility did not have its strongest effects on lowly
involved suhjects, as was postulated in the second
hypothesise.
The newspaper control candidate appeared to he
more effective in persuading the subjects than were the
carefully devised commercials which dealt at great length
with specific issues. This unexpected finding related to
the question of whether it was good campaign strategy to
avoid issues 5 regardless of how popular the candidate!,s
stand was. The answer seemed to be positive; avoid
specific is sue s. A politician should speak in generali
ties. Again, this reinforced the results of testing
hypothesis one which stated that the image was more
important than specific Issues. The newspaper candidate,
in only presenting a short resume of his background and
expertise, focused only on image, was more successful than
the television candidates who presented a variety of
treatments on specific political issues.
It should be noted that these findings correlate
with the current beliefs of prominent professional
campaign consultants whose viewe were discussed earlier
in this dissertation in the Review of Literature section.
72
Relevance of Findings to Pertinent Theoretical Literature
In general, past literature dealing with source
credibility and message discrepancy indicated that these
variables were important in the process of persuasion.
Evidence from such studies as those of Wyer and Schwartz,
Hovland and Weiss, Aronson, Turner and Carlsmith, Bochner
and Insko were cited in the literature review of this
report to support this statement. However, none of these
scholarly works dealt in the area of television and
persuasion, but were in fact, generally either in the form
of print or'personal communication as opposed to tele
vision. However, there is still considerable speculation
as to whether the variable of the message discrepancy or
source credibility is more important under certain circum
stances. As the result of the evidence of the first
hypothesis, it can be stated that when these variables
were utilized experimentally in political television
commercials, the source credibility variable seemed to be
considerably more important than that of message
discrepancy.
As the review of literature of this study indicated
there has been considerable research in the area of
ego-involvement. Relevant constructs dealing with this
variable have been identified and expanded over a
prolonged period of time. Basically, the development
of this construct has been logical and forthright. It has
73
been found that high, ego-invo.lvejnent generally, causes
smaller shifts in attitude or persuasion. However, in
the presrent study',, this, did not seem to be. the case.
Th-e results of this research, relevant to hypothesis two,
are not consistent with earlier findings of Sereno and
Bodaken, Sherif and Sherif, Sereno and Mortensen, Eagly
and Man i s . ' and Atkinson and Bieri. However, the findings
of the present research do support Zimbardo, who found
that highly involved subjects changed significantly more
than did subjects who were not involved.
It can be recalled from the review of literature
that Freedman discovered that ego-involvement significantly
interacted with message discrepancy using change in
attitude as the dependent variable. The present research
found the converse to be true. Specifically, hypothesis
two indicated that ego-involvement did not conclusively
interact.with message discrepancy as indicated by Table 3
which presented an interaction F ratio of 0.7 for the
variables of ego-involvement and message discrepancy when
considering candidate vote. An interaction F ratio of
0.4 was found when considering the variable of vote
difference or adjustment for the control vote.
The research of Scileppi and Johnson was similar
to that of the present study in that they manipula Led
each of the variables considered in this study utilizing
a three-way: factorial design. Their findings, however,
74
were, different; from those: of the. present study,
irrespective of the variable of .television. That is to
say, they found that the greatest .atti tude change took,
place with low-involvement subj ects with the existence of
a high source credibility communication, while the
present study discovered that only the credibility of the
communication was important and that ego-involvement made
no difference in the outcome.
Additional Findings
It was: believed 'that, although correlational
findings could not specifically support tentative causal
relationships in substantive research, nevertheless, they
would be useful in providing additional information rele
vant to why a specific hypothesis had either been accepted
or rejected. For this reason, the present research
utilized some correlations, to supplement the analysis of
variance.
The first set of correlations to be discussed are
those between the variable of group treatment and those of
votes; and attitudes on issues. It was found that there
were many moderate-to-strong correlations that were
significant at the .0-0-1 level that helped in interpreting
sis one.
The first significant correlation was the way a
subj ect voted for each, television candidate and the set
75
of. commercials that he viewed• It was found that there
was a positive correlation of .51 hetween the way .subjects
voted for the televisiori candidates and the specific
treatment they were given. This correlation supported
the significant 'F for hypothesis one., The next, correla
tion of interest was that between the way the subjects
voted for the newspaper or alternate candidate, and the
television treatment he received. That is to say that,
although in each cell the newspaper treatment was exactly
the same., there was a variance in how the subjects voted,
depending on which television candidate and message they
were exposed to. It was, therefore, apparent that there
was a significant correlation between the way a subject
voted for the television candidate and the way he voted
for the control or newspaper candidate. A moderate
correlation of .37 significant at the .01 level, was
established between these two variables, showing that the
way a subject voted for the television candidate was
related to the way he voted for the control or newspaper
candidate.
Another interesting finding was that the variable
of sex seemed to have no overall relationship to the
way s.iibjects: voted for the ”pretty boy^V or the ^middle
boy/’ That is, the female subjects: were not swayed by
a "comely appearance” as might be expected.
76
chapter: 5
SUMMARY^ CONCLUSIONS, AND IMPLICATIONS
Summary
This study examined the role of three variables
which.have been shown by research to be relevant in the
communication situation. These were message discrepancy,
source credibility, and ego-involvement. In reviewing the
literature, it was discovered that in most instances, the
greater the message discrepancy and the lower the
credibility, the less effective the communication was in
its ability to persuade an audience. It was also noted
that people who had a low involvement with a particular
issue would be more easily persuaded.
Applying this information to the area of political
television commercials, two hypotheses were formulated:
1. Source, credibility is more important than is
the message being presented.
2. High source credibility has its greatest effects:
on the voting behavior of lowly-involved subjects.
To test these hypotheses , nine sets, of television
commercials were produced in the. Theatre-Television Studio
at California State University, Fullerton. These
commercials; were Interspersed, at typical intervals, into
a program entitled "That Girl." The variations in the
77
connner.ciala were such that each of three candidates ; a
’ ’pretty hoy," ’ ’middle hoy,” and ’ ’ ugly hoy” gave each of
three messages;, of low",: medium, and high discrepancy.
Nine groups of suhjects,: totalling 9:48 people, were each
exposed to one version^.of the commercials. They were
as;ked to vote, for the candidate they saw, arid, in addition,
to vote for a control candidate they had read about in a
mock:new^aper advertisement. Using Diah scales, their
attitudes toward the issues discussed in the commercials
were examined. The Diab scale also measured the subjects’
levels of ego-involvement, McCroskey scales were employed
to measure the subjects’ reactions to the candidates.
Finally, a number of demographic questions were asked to
gain background information about subjects.
All of the information gathered was computer
analyzed. The findings supported the first, but not the
second hypothesis. That is, source credibility was more
important than message discrepancy, but ego-involvement
did not seem to play a role in the process. It was.
observed that nonrandomization of subjects did not appear
to distort results so that using a control candidate was
a reasonable approach to the problem of adjusting for
int er-cel1 differences.
CQnclusiohs
It was concluded frpm the findings that; ClI
during the course of political campaigns, persuasive
communijcations .were moat effective if they concentrated
more on image, than on iasue,' although issues did have some
importance; (.2) the theory of ego-involvement was not
a successful predictor of persuasive impact, and C3). that
the voting situation simulated in this study was one in
which a number of interactions were taking place, implying
attitudinal fluctuations based on the impact of the
commun i cations.
Therefore, we can conclude from this study, first
that it does not matter what a political candidate
communicates, but, rather, his source credibility, which
includes his. apparent believability and sincerity, as well
as physical appearance and style of delivery. In the arena
of political campaigning, candidates are justified in
giving a great deal of consideration to the areas of:
manner of speaking, ease of appearance, and personality.
It would likely be beneficial to these candidates to
emphasize personal facets of their campaigns while
deemphasizing time spent on the issues. Obviously, such,
an approach could, admittedly, be detrimental to voters’
understanding of what position a candidate stands for on a
variety of issues; however, the fact remains that a
candidate: muat seriousTy .address himself to those facets
of a campaign which will contribute the most to a vicluri-
ous campaign.
79
The second conclusion, from this study is that the
. theory of ego-involvement. does not seem to be useful in
the situation of assessing the impact of political tele
vision commercials. The finding of an insignificant F
ratio for the second hypothesis of the study was due either
to the Diab scales, to the testing situation in which
those scales were used, a combination of both previously
mentioned factors, or to a weakness in ego-involvement
theory which permitted the"formulation of the hypothesis
that low-involvement subjects were more easily swayed by
a persuasive message than were those subjects who were
highly involved.
It can be observed from the results, of the omega
srquared tests that both main and interaction effects
accounted for relatively small amounts of the total
variance encountered in the study. This indicates that a
great number of influences other than those of the
manipulative variables of this study were responsible
in influencing the results. Therefore, it is possible to
conclude that the test instruments used may not have been
sufficiently sensitive to detect the influence of ego
involvement in relation to all other types of study which
occurred in this ôtudy.
A final conclusion of the study was that the experi
ment had many dynamic and interactive aspects, some of
which emerged from the. data. In addition to the
80
significant interaction ,F’ s discussed in the findings:, it
was; als;o discovered that there, was a relationship hetween
what commercial a sahj'ect'?viewed and how he subséquently
felt about an issue. The. implication of this finding was
that the sub.] ects I attitudes were not def initeily
established, but rather th:at the messages he received in
this fictitious campaign interacted with his initial
attitudes on a particular issue in such a way that it
changed them. The conclusion reached was that political
campaigns likely form, change, and define the issues in
addition to merely addressing them.
Implications
It should be noted that many politicians in recent
history have had extremely popular public images, while,
privately, their deeds and actions, and motivations might
have been contrary to their public image. That is not to
say that they lacked in public charisma; quite the
contrary. Some public leaders have excelled in this
quality while seemingly being deficient in actual deeds and
action. Such, might have.been the case with the late
President John F. Kennedy. Bernard Kalb and .Marvin Kalb,
in their recent book, Kissihghr, noted that Henry
Kissinger had an opportunity.to serve in the Kennedy
admihistr'ation during its early years but that, for a
variety of ..reasons, he left. One of Kissinger’s primary
complaints:, according, to the. Kalb brothers; was that though
81
KennedY had a favorahle; image with, the American people.,
he was basically lacking in experience and substantive
.decision-making abilities.Kissinger, felt that Kennedy
held a rather naive approach to the administration of
foreign policy, and that he, Kennedy, would probably
intervene in the foreign affairs of other countries
because it was the thing to do; after all, we were
Americans and that was American’s duty. The Kalbs point
out that Kissinger became very disaffected with the key
people around Kennedy who helped make important policy
decisions affecting America.
The revelations relating to the confidential side
of the administration of John F. Kennedy are just begin
ning to be shown, and perhaps more information of a similar
nature will be forthcoming in later years. But, never
theless, here is what appears to be a direct contra
diction between the image of a much-beloved and well-
remembered President and his actual deeds. This excerpt
from an event in Kissinger’ ,s association with Kennedy
has some parallels to the present study inasmuch as the
evidence of hypothesis one was so overwhelmingly
postulated that the image of a politician was so much
more influential than the issues.
It is; not an accident of- history that John F.
Kennedy became the President of the: United States at
^Bernard Kalb, 'Marvin Kalb, Kissinger (boston;
Little, Brown and Co., 1974), p. 62. 82
the same time- that television was. emerging as the: most
pervasive form of .communication in the history of the
world.
While it may he true that history may remeiriber
John: F. Kennedy as being, at best, a mediocre president
in terms o-f his;: actions and accomplishments, most people
could not deny the fact that his positive image served as
a rallying point for more than one dark crisis in the
history of America during his presidendy.
Irrespective of the fact that both image and
a command of specific issues are important in the conduct
of foreign policy of the United States or the administra
tion of the smallest municipality, it can be clearly
inferred from this study that a clear, positive image is
certainly more important than the widest possible command
of any issue.
83
BIBLIOGRAPHY
BIBLIOGRAPHY
Agranoff, Robert. The New Style in Election Campaigns.
Boston: Holbrook Press, Inc., 1972.
Arkin, Herbert and Raymond R. Colton. Tables for
Statisticians. New York: Barnes and Noble, Inc.,
1941.
Aronson, Eliot, Judith A. Turner, and J. Merrill
Carlsmith. "Communicator Credibility and Communi
cator Discrepancy,” Journal of Abnormal and
Socia1 Psychology, 67:31-36, July 1963.
Atkins, Alvin L. and James Bieri. "Effects of Involvement
Level and Contextual Stimuli on Social Judgment,”
Journal of Personality and Social Psychology,
9:197-294, June, 1969.
Bochner, Stephen and Chester Insko. "Communicator
Discrepancy, Source Credibility and Opinion
Change,” Journal of Personality and Social
Psychology, 4:614-621, 1966.
Bower, Robert T. Television and the.Public. New York:
Holt, Rinehart and Winston, 1,97 3 .
Bowers, John Waite and William Phillips. "A Note on the
Generality of Source-Credibility Scales,” Speech
Monographs, 34:18 5-18 6, June, 19 67.
Chester, Edward W. Radio Television and American Politics
New York: Sheed and Ward, 196 9.
Diab, Lutfy N. "Some Limitations of Existing Scales in
the Measurement of Social Attitudes,” Psycho-
logicdl Reports, 17:427-430, October, 1965.
"Studies in Social Attitudes: I. Varia
tions in Latitudes of Acceptance and Rejection
as a Function of Varying Positions on a Contro
versial Social Issue,” Journal of Social
Psychology, 67:283-2 95, October, 1965.
. "Studies in Social Attitudes: III. Atti
tude Assessment Through the Semantic Differential
Technique,” Journal of Social Psychology, 67:3 03-
314, October, 19 65.
85
Eagley, Alice H., and Melvin Manis. "Evaluation of
Message and Communicator as a Function of Involve
ment," Journal of Personality and Social Psychology
4:483-485, April, 1966.
Emmert, Philip and William D. Brooks. Methods of
Research in Communieation. Boston: Houghton
Mifflin Company, 1970.
Fishbein, Martin (ed.). Readings in Attitude Theory and
Measurement. New York: John Wiley and Sons,
Inc. , 1967 .
Freedman, Jonathan L. "Involvement, Discrepancy, and
Change," Journal of Abnormal and Social Psychology
69:290-295, September, 1964.
Hays, William L. Statistics. New York: Holt, Rinehart
and Winston, 1963.
Honomichl, Jack J. "Nixon’s ’73 Election Researchers
Stressed Interyiews, Polls Phone," Advertising
Age, June 4, 19 73, p. 95.
Hovland, Carl I. and Walter Weiss. "The Influence of
Source Credibility on Communication Effective
ness," Public Opinion Quarterly , 32.: 6 3 5-6 50 , 19 51.
______________ and Henry A. Pritzker, "Extent of Opinion
Change as a Function of Amount of Change Advocated
Journal of Abnormal and Social Psychology,
54:257-261, 1957.
_________ , 0. J. Harvey, and Muzafer Sherif, "Assimila-
' tion and Contrast Effects in Reactions to Com
munication and Attitude Change," Journal of
Abnormal and Social Psychology, 55:242-252,
September, 1957.
Johnson, Homer H. and John A. Scileppi. "Effects of Ego-
Involvement Conditions on Attitude Change to High
and Low Credibility Communicators," Journal of
PersQhality and Social Pscyhology, 13:31-36,
September, 196 9.
Kalb, Bernard and Marvin Kalb. Kissinger. Boston: Little
Brown 8 Company, 1974.
Koslin, Bertram L., James W. Stoops, and Wallace D. Loh.
"Source Characteristics and Communication
Discrepancy as Determinants of Attitude Change
and Conformity," Journal of Expérimenta1 and
Sbcial Psychology, 3:230-242, July, 1967 . g g
McCroskey, James C. "Scales for the Measurement of Ethos,"
Speech Monographs, .33:65-72, March, 1966.
Mendelsohn, Harold and Irving Crespi. Polls, Television
and the New Politics. Scranton: Chandler Publish
ing Co., 197 0.
Mortensen, C. David. . . ' . Communication: The Study of Human
Interaction. New York: McGraw-Hill, Inc., 1972.
and Kenneth K. Sereno. "The Influence of
Ego-involvement and Discrepancy on Perceptions
of Communication," Speech Monographs, 37 :127-134,
June, 19 70.
Napolitan, Joseph. The Election Game and How to Win It.
Garden City: Doubleday and Company, Inc., 1972.
Nimmo, Dan. The Political Persuaders. Eaglewood Cliffs:
Prentice-Hall, Inc., 1970.
Osgood, Charles E., George J. Suci, and Percy H.
Tannenbaum. The Measurement of Meaning. Urbana:
University of Illinois Press, 1957.
Scott, Michael D. "Attitude Change as a Function of Ego
Involvement, Message Discrepancy, and Source Credi
bility: An Experimental Test of Opposite Theoretic
Statements.V Unpublished dissertation. University
of Southern California, March, 197 3.
Sereno, Kenneth K. "Ego-involvement, High Source Credi
bility, and Response to a Belief-Discrepant
Communication," Speech Monographs, 35:477-481,
November, 1968.
"Ego-involvement: A Neglected Variable in
Speech-Communication Research," The Quarterly
Journal of Speech, 55:69-77, February, 1969.
and C. David Mortensen. "The Effects of Ego-
Involved Attitudes on Conflict Negotiation in
Dyads,7 Speech Monographs, 36 : 8-12 , March, 196 9.
and Edward M. Bodaken. "Ego-involvement and
Attitude Change : Toward a Reconceptualization of
Persuasive Effect," Speech Monographs, 39 :151-158 ,
August, 1972.
Sherif, Muzafer and Carolyn W. Sherif. Social Psychology.
New York: Harper and Row, 19 69.
87
Twentieth Century Fund Commission on Campaign Costs in the
Electronic Era. Voter’s Time. New York : The
Twentieth Century Fund, 1969.
Whittaker, James 0. "Cognitive Dissonance and the
Effectiveness of Persuasive Communications,7
Public Opinion Quarterly, 28:547-555, Winter,
1964.
Wyer, Robert S., Jr. and Sandra Schwartz. "Some
Contingencies in the Effects of the Source of a
Communication on the Evaluation of that Communica
tion ," Journal of Personality and Social
Psychology, 11:1-9, 196 9.
Zimbardo, Philip G. "Involvement and Communication
Discrepancy as Determinants of Opinion Conformity,
Journal of Abnormal and Social Psychology, 6 0:86-
94, January, 196 0.
APPENDIXES
APPENDIX A
ISSUES
90
APPENDIX A
ISSUES USED
A panel of ten judges, composed of students at
California State University, Fullerton, rated copy from
each of the twelve commercials according to how popular
they felt the stands presented in those commercials were.
Subsequently, additional ratings were given by the same
panel of judges according to attractiveness or image of
each of the three candidates. This was accomplished by
showing the panel each of the three candidates presenting
each of three different stands (for a total of nine
commercials) on one issue, the ’ ’energy crisis.” This
Appendix contains a copy of the form used to obtain these
ratings, a Table listing the results of the commercial
copy evaluated, and another Table listing the ratings
for the respective images•
91
INTRODUCTION TO PANEL
On the following pages are political TV commercials
dealing with four different issues. There are three
commercials on each issue and each commercial takes a
different stand. Please read all three commercials on an
issue and rate them 1, 2, and 3, according to how popular
you feel the stand would generally he with Cal State,
Fullerton students. Just write the ranking number at the
top of each page. For example, if the first commercial you
read seemed to be more popular with Cal State Fullerton
students than the other two commercials in that issue,
place the number one at the top of the page.
Thank you for your cooperation.
92
Subject: Campaign Spending
INTERVIEWER
CANDIDATE :
VOICE OVER:
I guess you^re aware of the enormous sums of
money that were illegally donated in the last
Presidential campaign by large corporations.
I think it !s a crucial issue, and I M like to
hear your stand on that.
This is a complex situation and I’ll give you
my short answer on this , . . Public financ
ing of campaigns and new disclosure laws.
Giving people a tax write-off for contributing
to state and Federal campaigns is absolutely
the only way to get vested interests out of
government. The end result would be that
candidates would be far more responsive to
the voter and the voter would be better
informed by the increasjed financing of
advertising for the elections. To further
discourage any illegal large contributions,
maximum limits would be set as to how much a
candidate could spend in each of the major
media such as television, newspapers, etc.
Only a candidate without^vested interests
can take a stand promoting public financing.
Is your candidate responsive to the needs of
the people? Be sure he is by electing Herman
Ackheimer, Governor. Paid for by the
■Ackhe.imer for Governor Committee. 9 3
Subject: Campaign Spending
INTERVIEWER
CANDIDATE:
INTERVIEWER
CANDIDATE:
If you could pinpoint the one biggest
problem in dishonest government right now,
what do you think that would be?
There can be only one realistic, workable
solution to this dilemma— de-regulàtiôn of
campaign spending. Present laws in this area
are too stringent because they force corpora
tions to commit illegal acts in order to
participate in the political process. To
show the absurdity of trying to control such
laws, one only has to look as far as the
recent presidential election in which both
the Republican and Democratic parties
accepted large amounts of illegal funds. In
this age of mass media, it takes ever-
increasing amounts of money to conduct an
effective and informative campaign. Any cuts
in funding decrease the amount of
desperately needed information getting to the
voting public in the form of campaign adver4
tising so that the voter would be even less
able to make an intelligent decision.
So you feel that corporations can actually
provide a service?
Exactly. Why take more money from the
94
already over-burdened taxpayer to try a
shaky experiment in public financing of
campaigns?
VOICE OVER: It's your money. You make the decision.
Vote for Athur Wilcox, to be your next
governor. Paid for by the committee to
elect Arthur Wilcox.
9 5
Subject: Air Pollution
CANDIDATE:
VOICE OVER
There is no perfect solution which will allo%
us to have a clean envix’ oiiinent and also have
enough energy. We have to have a trade-off,
and I feel that when it comes to putting
people out of work or else having to increase
our pollution problems, the best answer is
to sacrifice the clean environment. All
pollution and smog control devices on autos
presently should be removed because they use
too much fuel. Cities should be permitted
to burn high sulfur oil in order to alleviate
the electricity shortages. Strip mining and
off-shore drilling are necessary evils and
should be tolerated if we are to keep the
economy from going into a severe depression.
Nuclear power plants must be constructed at
a faster rate in order to keep up with
ever-increasing demands for more energy.
It is simply a matter of ordering priorities
There is no perfect solution to every
problem. With recession on the horizon we
desperately need a governor who clearly sees
what^s most important. We need Arthur
96
Wilcox. He's able to make the tough
decisions. The preceding announcement was
paid for by the citizens for Arthur Wilcox.
97
Subject: Air Pollution
CANDIDATE
VOICE OVER
By now, most people are aware of the immedi
ate and personal effects of the energy crisis
These include a scarcity of gas for cars,
reduced speed limits, and some recent
electricity cutbacks. These problems are
certainly not easy to live with, but when
compared to the health and ecological
hazards of pollution, they must be treated
secondarily. What I am saying is that
decreasing pollution is more important than
is saving energy. This means that there
should be no off-shore drilling of any kind,
the Alaskan pipeline should be halted, and
planned emission control standards for
autos should be carried out even though all
of these things spell an even greater energy
sacrifice. With the development of alternate
clean forms of energy, we will realize that
the energy crisis is only temporary, but the
damage done -to ecology and health by a
poisoned environment is forever.
Clean up with Henry Freedman. Make him
Governor of California. Paid for by the
Committee to elect Henry Freedman.
98
Subject: Air Pollution
CANDIDATE: Today, many people feel that certain methods
of solving our pollution problems cause a
greater use of badly needed energy, especi
ally when it comes to auto pollution devices
and the burning of high sulfur fuels for
electricity. I feel strongly that in this
time of the energy crisis we must not be
swayed by the furor of the moment to make
hasty, irrational decisions. We should not
compromise our present gains in solving the
pollution problem and continue to find
hetter means of improving the environment
for ourselves, our children, and future
generations. There will always be minor
problems , like a temporary scarcity of gas
for cars and lower speed limits, but these
sacrifices will be far outweighed by the
benefits of pushing for a cleaner world.
On the other hand, certain concessions have
had to be made such as the-start of the
Alaskan pipeline, offshore drilling, and
coal strip mining in order to kfeep this
country running. I, for one, will make
every effort to insure that these conces
sions are carefully controlled in order to
99
minimize any environmental damage. In
order to do this, I need your backing.
VOICE OVER: Vote for Herman Ackheimer, Governor, to
make our future a clean future. Paid for
by the Committee to elect Herman Ackheimer.
100
Subject: Inflation
CANDIDATE :
VOICE OVER:
CANDIDATE:
INTERVIEWER:
Inflation must be stopped] You know it and
I know iL. Fortunately it can be, if we are
willing to make a few small sacrifices.
We're talking to Herman Ackheimer,
candidate for Governor.
I see a two-fold solution to the problem.
Sure, decrease government spending and
tighten credit, but give the people who are
put out of work by these measures a job.
Since oil is our biggest source of inflation,
I suggest we first establish realistic modes
of public transportation which will
alleviate the demand for oil and, secondly,
push for greater development of alternative
forms of energy. This will cut energy costs
and save us from foreign "energy black
mail." But, more importantly, it will create
new jobs. This tremendous work project
could be administered in a way similar to
that of the WPA back in the 1930 Vs, with
even those with no technical skills at all
being given a place in the construction
process.
How will this project be paid for?
101
CANDIDATE :
VOICE OVER
It will more than pay for itself in the
service it provides the public and the ■ -
revenues it generates and in the fact that it
will help combat inflation.
A vote for Herman Ackheimer is a vote for
the future. Paid for by the Committee to
elect Herman Ackheimer.
102
Subject i ; Inflation
CANDIDATE:
VOICE OVER:
CANDIDATE:
INTERVIEWER:
CANDIDATE :
VOICE OVER:
It has been proven time and again that
tampering with our present economic system
leads to disaster.
We^re talking to Henry Freedman, candidate
for Governor.
Things could be far worse than they are
presently. Thanks to skillful handling of
the economy, the United States has enjoyed
a trade surplus for the first time in many
years and the dollar has taken a resurgence
abroad. America has done better économie
cally than almost every other free-world
country and, therefore, despite some infla
tion and unemployment, I feel that the
present course, which has been tested and
proven to be viable, should be continued.
Could you specify some of those present
measures?
Inflation can best be fought with tight
credit, just as is now being done, and we
should continue exporting agricultural
commodities ;to help support the value of
the dollar abroad.
A strong America is essential to a strong
California, so give your vote to Henry lo
Freedman for Governor and elect a man who
knows how to solve our present economic
problems without,creating new ones. Paid
for by the Committee to elect Henry
Freedman.
104
Sub ject : Inflation
CANDIDATE:
VOICE OVER:
CANDIDATE:
INTERVIEWER’
CANDIDATE :
VOICE OVER:
You have to remember that inflation is not
all bad. When inflation's up, so is
employment. And, right now, real income is
at an all time high.
We're talking to Arthur Wilcox, candidate
for Governor.
A tight monetary credit policy by the
Federal Reserve Board would curb inflation,
but it would also tighten credit spending
and put a lot of people out of work. That's
why I think it should stay high. Inflation
itself is nothing, it's real income that's
important.
Just what is real income, and how does it
relate to the working man?
Real income is the amount you earn minus
what you lose due to inflation. Sure, you
lose buying power through inflation, but
you have an increasingly large paycheck
which more than compensates for that. The
net result is that our standard of living is
constantly on the rise.
A vote for Arthur Wilcox is a vote for
prosperity. Insure your future with the man
who knows what's real. Paid for by the
commit tee to elect Arthur Wilcox.__________
Subject: The Energy Crisis
CANDIDATE:
INTERVIEWER:
CANDIDATE :
INTERVIEWER:
You know and I know that there is only one
real solution to this falsely manufactured
Energy Crisis, and that is strict regula
tion of the oil industry, such as there is
of public utilities like the electric and
telephone companies. The oil industry has
consistently refused even to provide
precise petroleum inventory figures to the
government.
What do you think they are trying to hide?
Simple. They are concealing reserve stocks
that do not show up in the American petrol
eum institute’s statistics. The oil industry
has repeatedly been granted concessions which
allow it to act in monopolistic ways, so that
any nationalization of this industry will
not be a quelling of the free enterprise
system, but rather a restructuring of an
already noncompeting body of international
cartels which have no allegiance to this
country, but rather"operate strictly on the
profit motive.
For instance ... .
106
CANDIDATE :
VOICE OVER
Well, if you think the Présidentes tax
deductions were questionable • • • the oil
industry deducted over two billion dollars
from their returns which they claim they
paid in foreign taxes, leaving their tax
hill at only seven hundred million dollars.
Herman Ackheimer knows the facts. If you
know the facts, you'll give him your vote
for governor. Paid for by the committee
to elect Herman Ackheimer.
107
Subject: The Energy Crisis
CANDIDATE: Of course there's an Energy Crisis. We see
it every day with cold homes and gas less
Sundays:. However, the only thing which has
caused the Energy Crisis was the recent Arab
oil Embargo. But, even with this Embargo
considered, there is a far greater shortage
than one would expect.t
INTERVIEWER: What do you see as the reason for this?
CANDIDATE: It is likely that the oil companies have
taken advantage of the Embargo to gain for
themselves large price increases by further
decreasing the amount of oil available 1 ; ;
through monopolistic practices.
INTERVIEWER: How can we deal with this?
CANDIDATE: A windfall profits tax on the oil companies
is the only way to discourage hoarding of
oil. No matter how much prices increase in
oil, companies would collect the same
income. This will once again get oil flowing
in this country, though there will still be
a few minor spot shortages due to the
problem caused by the Arah Embargo.
VOICE OVER: Look over all the candidates for governor,
then, make your decision. We know you'll
go with Henry Freedman because he cares
10 8
about your future. Without him, there may
not be a future. Paid for by the citizens
for Henry Freedman Committee.
109
Subject: The Energy Crisis
CANDIDATE: I feel very strongly that the energy crisis
is real, and that it was not in any way
precipitated by the oil companies in order
to gain increased profits.
INTERVIEWER: Then just who is responsible for the present
situation?
CANDIDATE: Two basic problems brought about the energy
crisis— namely, the Arab embargo and the
ever—increasing demand of U.S. consumers.
The only real solution is, then, to cut back
on demand since we never know when the
Arabs; may stop the flow of oil again.
Present measures such as mandatory fuel
allocations and the 55 mile per hour speed
limit are helping. Unfortunately, gas
rationing will likely be the best solution
in the future, as far as the oil problem is
concerned. The best thing citizens can do
is cooperate with present and future
regulations and also comply with various
voluntary measures for cutting fuel usage.
The worst thing that can happen is if
people go back to their higher levels of
consumption.
110
VOICE OVER: In hard times, you need a real leader.
That's why you should go with Arthur Wilcox
for Governor. Paid for by the Committee to
elect Arthur Wilcox.
Ill
RATINGS FOR ISSUES*
Stand Ideal Rating Average Rating
Campaign
Spending
Neutral
Unpopular
1.1
2.0
2.9
Air
Pollution
Popular
N eutral
Unpopular
1. 3
1.7
3.0
Inflation
Popular
Neutral
Unpopular
1.1
1.4
1.6
The Energy
Crisis
Popular
Neutral
Unpopular
1.6
1.8
2.6
*Ratings ranged from 1 to 3 with 1 equalling very popular
112
RATINGS FOR IMAGES*
Candidate Stand Ideal Rating Average Rating
Pretty Boy
Popular
Neutral
Unpopular
1.2
1.0
1. 3
Middle Boy
Popular
Neutral
Unpopular
2.0
1.9
2 .0
Ugly Boy
Popular
Neutral
Unpopular
2.8
2.1
2.7
*Ratings ranged fromll to 3 with 1 equalling very
attractive
113
APPENDIX B
NEWSPAPER ADVERTISEMENT AND
QUESTIONNAIRES
114
NEWSPAPER ADVERTISEMENT
i Iron: i l f
Sjave Rivej.
a incliMlod
during rt-
)m locom o
r l «aid. The
ick bunker
contUtency
ompany In-
a to treat
Ail. kccord*
irtin e d t ùt
ra## !*•
j were not
andle peak
,e were oc-
ihe s)«iem
and oily
charged to
j however,
Ovcd untU
said.
'ft 11M2. the
I to convert
imotivee to
| r the con-
■ said the
from the
I contained
DII n t a of
led during
Lions.* The
{successful,
(lowed into
k .
{diesel en-
d serious
) l e m 6 in
items. As a
ids added
ie radiator
Is. too. was
I the river-
om ium be
ll in down- i
water U di sewage origin verbal protests 3l a grow- lied oil in the waste.- tnrougn ine plant, ano me me nant#^
VOTE FOR
ROBERT LGGAM
FOR GOVERNOR OF CALIFORNIA
EXPERIENCE COUNTS
Robert Logan wants your vote for governor. But, he also wants
more than that. He wants you to look at the records of all the
candidates for governor because, if you do, you will find that he
is by far the most qualified individual available for the job. A law
degree from Stanford, four years as a city manager in Fresno,
three terms in the California legislature, and two years as a district
court judge in San Diego have made Robert Logan the ideal
candidate for governor. Robert Logan knows California. He has
the knowledge to know what needs doing and the experience to
get the job done. Throughout bis extensive public service career,
not once has there been even the slightest blemish on his record.
Many concerned California citizens realize this and believe that
if it is pointed out to you, you will make your vote a vote for
integrity. Vote for Robert Logan. You know where he stands, and
he'll stand up for you.
Paid for by the Robert Logan for governor committee
93950 or contact Carol M il- , _ l" bojh dikes _
BV C L Ï
W ith t
A m trak j
b:gger rp
For cr!
tb e 2<-|
from ■ In
whole nc*
ca in a ■ '
lu xurioui
snd. let',
time# fri:
Recenl
west Urt,
Super Q i
1 becamt
the West
experte ni
or behin'
.w-
L itt
1 boarc!
Lim ited
Friday ei
in Kansa;
m orning!
the road
private i
more th
less tha
compart;
Skim I
and see I
the astrc
country I
geu
Los A i
nardino:,
dinner t
lion, pi
served t
Jableclo'
lim e tra|
fresh ro!
a re St
Glimpse:
on front'
116
QUESTIONNAIRES
INSTRUCTIONS
On the following pages are listed political issues
which are followed by pairs of words which can describe
the issues. In between each pair of words is a set of
seven spaces, representing a scale. First, place a +
wherever you feel it belongs in the spaces between each
pair of words if you find that position to be acceptable
to you. Next, circle the plus sign in that position which
you feel is most acceptable. Place a - in all positions
which you feel are unacceptable. Last, put a Q in every
other position which as yet has no markings. Complete
each of the scales, ranging from one to seven, based upon
how well you feel that the pairs of words describe your
feelings about the concept at the top of each set of
scales. An example of how the spaces are?filled in
appears below. First place the t in each space which you
accept.
SEX EDUCATION IN SCHOOLS:
Good____ ^ : + : + :____ : :__ _ :____Bad
Then circle the most acceptable position.
Good____ • : + :____ :____ :____ :_____Bad
Place a - in each unacceptable position.
Good - : ft) : t : - : - :_________ Bad
Place the 0 in all remaining positions.
Good - : : + : - : - : Q : 0 Bad
Complete the entire list of scales for each of the
various concepts presented on the following yellow colored
pages.
Control of Political Campaign Spending is:
good
ugly _
clean
large
weak
heavy
sharp_
passive_
fast
Environmental Pollution is
ugly
bad
clean
heavy
weak
large
fast
passive
sharp
Bad
beautiful
dirty
small
strong
light
dull
active
slow
beautiful
good
dirty
light
strong
small
slow
active
dull
119
Inflation is:
ugly : ; : beautifu:
good : : : bad
clean : : : dirty
large : :
small
weak....: : : strong
heavy : : : light
passive : : . ; : active
sharp : : : dull
fast : :
slow
The Energy Crisis is:
beautiful : :
ugly
bad : : :
. .
good
clean : : :
• .
dirty
large : : : small
strong : : :
weak
heavy : : : light
passive : : : active
sharp ' : ■ : ^ : dull
fast : : slow
1 2 0
Now, simply place an X in the spaces
words to describe the candidates you
Use one X per pair of words, placing
fitting position.
between each pair of
just learned about,
it in the most
ARTHUR WILCOX
reliable ; : : : :
IS:
: unreliable
uninformed : : : : : : informed
qualified : : : : : ; unqualified
intelligent : : : : : unintelligent
valuable : : ; : : worthless
expert : : : : : : inexpert
dishonest : : : : : : honest
unfriendly : : : : : friendly
pleasant : : : : : : unpleasant
unselfish : : : : : : selfish
awful : : : : nice
virtuous : : : : : : sinful
1 2 1
ROBERT LOGAN IS:
reliable_
uninforined_
qualified_
intelligent_
valuable^
expert_
dishonesty
unfriendly_
pieasant_
unselfishy
awfuly
virtuous
unreliable
informed
unqualified
yUnintelligent
worthless
inexpert
honest
friendly
unpleasant
selfish
nice
sinful
122
PLEASE CHECK THE APPROPRIATE BOXES
I am:
1. CSF Freshman / ^ '
Junior £_
Senior / / __
Graduate Student / /
2 . Male /£7
Female /~~T
3. 17/27 18/27 19/27 2 0/27 21/27 ' 22/77 23/77
24 or over /~T
4. Are you a registered voter? yes/77 no/77
5. If yes, with which party?
Republican 7~7
Democrat / 77 '
Other I~1 Tspecify)
5. Have you ever donated money to a political party?
Yes/~7
No/ 7 7
6. If you voted in the 19 72 presidential election, for
whom did you vote?
Nixon £77
McGovern l~~T
Other / / (specify) ____________________________
Did not vote / /
7. Do you have a television at home in working order?
Yes /~7
No £27
8. Family^s approximate income:
Under $5,000 I~1
5.000-9,999 £77
10.000-19,999 /“7
20.000-29,999 77^
30,000 or over / 7
9. Race:
White !'~f
Black / /
Spanish-American /~7
Asian /~7 Other £77 (specify) ________ ^2
APPENDIX C
BALLOT
124
ARTHUR WILCOX;
I would-dëfinitely vote for him /~7
I would prohably vote for him / /
I would probably not vote for him /~~7
I would definitely not vote for him f~T
ROBERT LOGAN:
I would definitely vote for him
I would probably vote for him / /
I would probably not vote for him /~7
I would definitely not vote for him l~~T
125

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Outside the Whale: Reading the American Political Novel in the Age of Reagan

PDF

Effects of ego-involvement and success-failure on postcompetitive affect in young athletes

PDF

William Lawes: His life and works

PDF

Reaffirmation and transformation of gender in popular film: A feminist approach to mythic rhetoric

PDF

An historical study of Bob Pierce and Worlds Vision's: Development of the evangelical social action film

PDF

Perestroika: The Soviet political economy in transition

PDF

Analysis and simulation of cascaded phase locked loops

PDF

The life and Mantuan masses of Francesco Rovigo (1541/42-1597)

PDF

A structural and psychoanalytic analysis of the symbolic ritual process in three sets of Russian folk tales

PDF

Anglo-Jewish art collectors of the Victorian Period: Patterns in collecting

PDF

An examination of personality type dominating leadership positions in department of defense program management

PDF

An experimental study of the relative effectiveness of certain types of radio and television commercials

PDF

The harpsichord works of Nicolas Le Begue

PDF

Attitudes and knowledge of the aged regarding old age insurance and old age assistance

PDF

SEC's enforcement actions and its effectiveness in an efficient market

PDF

Location Orientation Manipulator: A Physical Tool for the Study of the Kinematics of Assembly of Building Systems

PDF

The effects of the Educational Employment Relations Act and the impact of using alternative bargaining techniques

PDF

Crisis preparation in technical organizations: A study using a multi-dimensional approach

PDF

Women's autobiography and national identity: Natalia Ginzburg, Anna Banti and Renata Vigano

PDF

I'm not much different: Occupation, identity, and spinal cord injury in America

Asset Metadata

Creator Boyd, John M. (author)

Core Title An experimental analysis of source credibility and message discrepancy under differential levels of ego-involvement in political television commercials

Degree Doctor of Philosophy

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag communication and the arts,OAI-PMH Harvest,Social Sciences

Format application/pdf (imt)

Language English

Contributor Digitized by ProQuest (provenance)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c37-40993

Unique identifier UC11632736

Identifier DP71316.pdf (filename),usctheses-c37-40993 (legacy record id)

Legacy Identifier DP71316.pdf

Dmrecord 40993

Document Type Dissertation

Format application/pdf (imt)

Rights Boyd, John M.

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA