Hovland Propaganda Attitude Changes Sev9c

:@
;.
1
chapter
9
Theories of Persuasion
I
I
Persuasion is only one type of mass communication, but it is a type in which
many people are interested. The advertiser using mass communication to sell
soft drinks, headache remedies, or automobiles is engaged in persuasion. So is
the nuclear power industry when it hires public relations experts to help it
convince the public that nuclear power is safe. So are the political candidates
who buys newspaper ads, the public health organization that prepares radio
spots to encourage people to stop smoking, and the reIigious organization that
puts evangelical messages on television. AU of these people are attempting to use
mass communication messages to produce some kind of change in other people.
Persuasion has been defined as attitude change resulting from exposure to
information from others (Olson & Zanna, 1993, p. 135). Persuasion has probably
always been a part of human life. It seems inevitable that people will try to
influence other people, even their closest friends and family members. As Eagly
and Chaiken (1993) have noted, persuasion is particularly likely to make up part
of life in a democracy, where attitudinal influence is the form of control that is
most relied on.
For centuries people must have operated on the basis of intuition and
common sense in their attempts to persuade. Aristotle was one of the first to
try to analyze and write about persuasion, in his classic works on rhetoric. Years
later, particularly when mas5 communication became more widespread, people
began to study persuasion even more systematically. The Institute for Propaganda
Analysis, with its identification of seven techniques of propaganda, was doing
some of this early work.
Part of the motive for the Institutes careful study of persuasion was
obviously fearthe war-inspired fear that propaganda could win the hearts and
I II
minds of people. The Institute was operating in that panicky period just before
1
1 8 0 THE SOCIAL-PSYCHOLOGICAL APPROACH
World War 11. A few years later, the same war
was to produce the first careful
scienttilc studies of persuasion, or attitude change, as it became ~own. This work
was done by psychologist Carl Hovland and his associates, all of whom were
working for the Research Branch of the U.S. Armys Information and Education
Division. This work was so original and influential that it has been called the
most important fountainhead of contemporary research on attitude change
(Insko, 1967, p. 1). The Hovland work was based on controlled experiments in
which variables were carefully manipulated in order to observe their effects.
Some earlier work on attitude change was done before Hovland, but rather
poorly. A study sometimes cited as the first attitude change study was an
investigation by Rice and Willey of the effects of William Jennings Bryans address
on evolution at Dartmouth College in 1923 (described in Chen) 1933). A group
of 175 students indicated their acceptance or rejection of evolution on a five-
point scale. The students were asked to give their attitudes after hearing the
speech, and, from retrospection, their attitudes before hearing the speech. They
found that more than onequarter of the students showed substantial change in atti-
tude, but the use of the retrospective report makes the finding highly questionable.
THE CONCEPT OF ATTITUDE
The concept of attitude has been described by psychologist Gordon Allport
(1954) as probably the most distinctive and indispensable in contemporary
American social psychology (p. 43). Allport points out that the term came to replace
in psychology such vague terms as instinct, custom, social force, and sentiment.
Some recent definitions of attitude include the following:
Attitude is primarily a way of being set toward or against certain
things. (Murphy, Murphy, & Newcomb, 1937, P. 889)
A mental and neural state of readiness, organized through experience,
exerting a directive or dynamic influence upon the individuals re-
sponses to all objects and situations with which it is related. (Allport,
1954, p. 45)
An enduring, learned predisposition to behave in a consistent way
toward a given class of objects. (English & English, 1958, p. 50)
An enduring system of positive or negative evaluations, emotional
feelings and pro or con action tendencies with respect to a social
object. (Krech, Crutchfield, & Ballachey, 1962, p. 177)
HOVLANDS ARMY RESEARCH
Hovlands approach to attitude change was essentially a learning theory or
reinforcement theory approach. He believed that attitudes were learned, and that
they were changed through the same processes that occurred when learning took
9 / THEORIES OF PERSUASION
place. Hovland had studied and worked at Yale with Clark Hull, whose
181
theory
was probably the most influential theory of learning between 193o and 195o
(Hilgard & Bower, 1966).
During World War II, the U.S. Army began using films and other forms of
mass communication on an unprecedented scale. Most of this material was used
in the training and motivation of U.S. soldiers. The Experimental Section of the
Research Branch of the War Departments Information and Education Division
was given the task of evaluating the effectiveness of these materials. Much of
the research of the Experimental Section is reported in the volume Experiments
on Mass Communication, first published in 1949 (Hovland, Lumsdaine, &
Sheffield, 1965). The section did two basic types of research: evaluation studies
of existing films and experimental studies in which two different versions of
the same film (or message) were compared. The section had to do much of the
first type of research because it suited the practical purposes of the Army. The
researchers felt, however, that the second type of research was really more useful
because it could lead to general principles of attitude change. These experimental
studies, in which certain variables were manipulated, really constituted the
beginning of attitude change research. But, the evaluation studies of existing films
also made some useful contributions to communication theory.
One of the first tasks the section took on was to evaluate the first four films
of a series of films called W@ We Fight. This series was produced by Frank Capra,
the famous Hollywood filmmaker who would later direct Zts a Wonderful L@e.
The W@ We Fight films were designed as motivational films to be used in the
training and orientation of American soldiers. They were based on the assump-
tions that many draftees did not know the national and international events that
led to Americas entrance into World War II and that a knowledge of these events
would lead men to accept more easily the transition from civilian life to that of
a soldier.
One of the films studied in great detail was The Battle of Britain, a 50-
minute film with the purpose of instilling greater confidence in Americas British
allies. Hovland and his associates designed research to determine the films
impact in three main areas: specific factual knowledge gained from the film,
specific opinions concerning the Battle of Britain, and acceptance of the military
role and willingness to fight. The research procedure was simply to have an
experimental group that saw the film and a control group that did not, and then
one week later to give both groups a questionnaire that appeared unrelated but
measured knowledge and opinions on subjects related to the film. These Army
studies were conducted with military units and therefore ended up with large
sample sizesthe Battle of Britain study involved 2,100 people.
The results showed that the film was quite effective in conveying factual
itiormation about the air war over Britain in 1940, that it was somewhat effective
in changing specific opinions about the conduct of the air war, and that it had
essentially no effect at M on motivation to serve or in building increased resentment
of the enemy. Thus the film failed in its ultimate objective, increasing soldiers
motivations. Similar results showed up for the other Wby We Fight films studied.
I
1 i
I II
182 THE SOCIAL-PSYCHOLOGICAL APPROACH 9 / THEORIES OF PERSUASION 183
This research on the why We Fight series became part of the growing body
of evidence indicating that a single mass communication message is unlikely to
change strongly held attitudes. Similar evidence comes from other studies as
different as the Cooper and Jahoda investigation of antiprejudice cartoons
(Chapter 4) and research by Lazarsfeld and his associates on political campaigns
(Chapter 11).
ONE-SIDED AND TWO-SIDED MESSAGES
Hovland and his associates turned to the second type of researchthe same
message is produced in two versions that differ in only one variablein an
experiment on the effectiveness of one-sided and two-sided messages. On many
issues there are arguments on both sides. Which is the better strategyto
mention only the arguments on the side you are pushing or to mention the
arguments on both sides but focus on the ones on the side you are pushing?
This is essentially the old question of the effectiveness of card stacking, one of
the propaganda devices identified by the Institute for Propaganda Analysis.
Hovland and his associates were trying to answer this question because they
faced a real communication dilemma. After the defeat of Germany in 1945, many
soldiers apparently felt the war was almost over. The Army wanted to get across
the idea that there was still a tough job ahead in defeating the Japanese.
The researchers realized that there were arguments for each strategy. A one-
sided presentation can be defended on the basis that a two-sided presentation
raises doubts in the minds of people unfamiliar with the opposing arguments.
A two-sided presentation can be defended on the basis that it is more fair and
that it will help prevent people who are opposed to a message from rehearsing
counterarguments while being exposed to the message. A specific purpose of
the study was to measure the effectiveness of the two kinds of message presen-
tation on two kinds of audience membersthose initially opposed to the
message and those initially sympathetic to the message.
Two versions of a radio message were prepared. Both presented the general
argument that the war would take at least two more years. The one-sided message
was 15 minutes long and brought out arguments such as the size of the Japanese
army and the determination of the Japanese people. The two-sided message was
19 minutes long and brought out arguments on the other side, such as the
advantage of fighting only one enemy, but it focused mostly on the arguments
that the war would be a long one.
One week before the presentation of the radio message, subjects were given
a preliminary questionnaire on which they expressed their estimates of how long
the war in the Pacific would take. Then one group made up of eight platoons
heard the one-sided message, a second group of eight platoons heard the two-
sided message, and a third group heard neither message and served as the control
group. Then all three groups received another questionnaire differing from the
first one in its form and its announced mumose. but again askina for an estimate
TABLE 9.1 Soldiers who estimated a war of more than 1 l/z years after hearing
one-sided and two-sided messages
Group 1 Group 2 Control Group
(8 platoons) (8 platoons) (8 platoons)
Preliminary survey 37% 387. 36%
Exposure to message one-sided two-sided none
Follow-up survey 59?. 59yo 34?40
S OUR C E : Adapted from C. 1. I+ovland, A. A. Lumsdaine, and F. D. Sheffield, Experiments on Mass
Corrrnrurrication (New fork: Wiley, 1965), Vol. Ill, Studies in Social Psychology in World War II, p, ZI O.
Copyright 1949, @ renewed 1977 by Princeton University Press. Reprinted by permission of Princeton
University Press.
of how long the war in the Pacific would take. All questionnaires were anony-
mous, but the before and after questionnaires for the same person could be
matched on the basis of the answers to questions about date of birth, schooling,
and so forth.
Looked at in general for all groups, the results (Table 9.1) indicated that
both kinds of presentations produced clear opinion change in comparison with
the control group, but that neither presentation was more effective than the other.
The researchers had anticipated that the two-sided presentation might work
better with an audience initially opposed to the message, so they proceeded to
check out this possibility. They did this by dividing each test group into subjects
initially opposed to the message and subjects initially favorable to the message.
The men who had given initial estimates that the war would take 1 l/z years or
less were considered to be initially opposed to the message, while those who
gave initial estimates of more than 1 /, years were considered to be initially
favorable to the message. Results of this analysis are shown in Table 9.2. Results
are presented in terms of net eflect, or the percentage in a group who increased
their estimate minus the percentage in that group who decreased their estimate.
TABLE 9.2 Effects on one-sided and two-sided messages on men
who were initially either opposed or favorable to the message
Initially Opposed Initially Favorable
(%) (%)
one-sided 36 52
Two-sided 48 73
Note: The number in the table is the net effect, or the percentage in a group who
increased their estimate minus the percentage in that group who decreased their
estimate.
S OUR C E : Adapted from C. 1. Hovland, A, A. Lumsdaine, and F. D. Sheffield,
Experiments on Mass Communication (New York: Wiley, 1965), Vol. Ill, Studies in
Social Psychology in World War 11, p. 213, Copyright 1949, 0 renewed 1977 by
Princeton University Press. Reprinted by permission of Princeton University PreSS.
. . .
A
184 THE SOCIAL-PSYCHOLOGICAL APPROACH
TABLE 9.3 Effects of the one-sided and two-sided messages
on men who graduated from high school and men who did not
Didnt Graduate Did Graduate
(%) (%)
One-si ded 46 35
Two-sided 31 4 9
Note: The number in the table is the net effect, or the percentage in a group
who increased their estimate minus the percentage in that group who decreased
their estimate.
S OU R C E : Adapted from C. 1. Hovland, A. A. Lumsdaine, and F. D. Sheffield,
Experiments on Mass Communication (New York: Wiley, 1965), Vol. Ill, Studies
in Social Psychology in World War 11, p. 214. Copyright 1949, 0 renewed 1977 by
Princeton University Press. Reprinted by permission of Princeton University Press.
This examination of results according to initial attitude shows the one-sided
message is most effective with persons initially favorable to the message and the
two-sided message is most effective with persons initially opposed to the
message. This is what the researchers had predicted.
Hovland and his associates also found that a one-sided message is most
effective with people of less education and the two-sided message is most
effective with people of greater education (see Table 9.3).
Both additional analysesthe one by initial opinion and the one by edu-
cation levelshow that the kind of presentation that is most effective depends
on the characteristics of the audience. These results brought out the complexity
of attitude changethat variables in the message sometimes interact with other
variables, such as personal characteristics of the audience. This is part of the
evidence that led psychologist Roger Brown (1958), in his analysis of propaganda,
to conclude that the propaganda devices are contingently effective rather than
invariably effective (p. 306).
THE Y=COMMUNICATION RESEARCH PROGRAM
After the war, Hovland returned to Yale University, where he had been a faculty
member, and continued his research on attitude change. This program had the
purpose of developing scientific propositions which specify the conditions under
which the effectiveness of one or another type of persuasive communication is
increased or decreased (Hovland, Janis, & Kelley, 1953, P. v). The project had
three characteristics: (1) it was primarily concerned with theoretical issues and
basic research; (2) it drew upon theoretical developments from diverse sources,
both within psychology and related fields; and (3) it emphasized testing Propo-
sitions by controlled experiment (Hovland, Janis, & Kelley, 1953).
The first book from the Yale program, Communication and Persuasion
(Hovland, Janis, & Kelley, 1953), dealt with a number of topics that would later
_.
SOURCE CREDIBILITY
9 / THEORIES OF PERSUASION 185
receive entire volumes or else become topics investigated extensively by later
researchers. TWO of these topicssource credibility and fear appealsare
particularly important because they led to many later studies. The book also
reported some further research on one-sided and two-sided messages.
One of the variables in a communication situation over which communicator
typically has some control is the choice of the source. And, judging from many day-
today examples of communication campaigns, there appears to be a widespread
belief that having the right source can increase the effectiveness of your message.
In the presidential election of 1992, President Bush was endorsed by such
popular country music stars as Ricky Skaggs, the Judds, Marty Stuart, Lee
Greenwood, and Tanya Tucker. Rock musicians such as Frank Zappa, Natalie
Merchant, and Madonna gave their support to rival candidate Bill Clinton. In
another illustration of a deliberately chosen message source, Nickelodeon, the
television network for children, centered a program on AIDS prevention around
Magic Johnson, the basketball superstar who had announced that he was HIV-
positive. The program, titled Special Edition: A Conversation with Magic: was
hosted by journalist Linda Ellerbee.
The selection of an effective source to speak for your idea or product is
essentially the propaganda device of the testimonial. But the effectiveness of this
technique was not really investigated by the Institute for Propaganda Analysis.
Hovland and Weiss (1951) designed an experiment to test the effectiveness of
source credibility. They apparently became interested in the possible impact of
source credibility after learning of a phenomenally successful radio program
involving entertainer Kate Smith. In an 18-hour program during World War II,
Kate Smith received pledges for an astounding $39 million for war bonds.
Researchers who studied the Kate Smith broadcast concluded that key elements
in her success were her perceived sincerity and trustworthiness.
Hovland and Weiss designed an experiment in which the same messages
would be presented to some people as coming from a high-credibility source
and to other people as coming from a low-credibility source. This would allow
them to determine the effect of the source variable alone. The experiment was
done with four messages on four different topics. Each subject received a booklet
containing four articles. Each article was on a different topic. The subjects
opinions on the four topics were measured with questionnaires before getting
~he communication, immediately after getting it, and four weeks after getting
it. Each article was presented with a high-credibility source for half the subjects
and a low-credibility source for the other half. The four topics were controversial
ones at the time and revolved around the following opinion questions:
1. Should antihistamine drugs continue to be sold without a doctors
Prescription? The high-credibility source on this issue was the New England
J ournal of Baology and Medicine. The low-credibility source was identitled in
the research report as a mass circulation monthly pictorial magazine.
2. Can a practicable atomic-powered submarine be built at the present
time? The high-credibility source was J. Robert Oppenheimer, the head of
the team of scientists that developed the atomic bomb. (This was before
Oppenheimers security clearance investigation, which undoubtedly damaged his
credibility.) The low-credibility source was Pravda, the Russian newspaper.
3. Is the steel industry to blame for the current shortage of steel? The high-
credibility source was the Bulletin of National Resources Planning Board. The
low-credibility source was identified as an anti-labor, anti-New Deal, rightist
newspaper columnist .
4. As a result of W, will there be a decrease in the number of movie theaters
in operation by 1955? The high-credibility source was Fortune magazine. The
low-credibility source was identified as a woman movie-gossip columnist .
The design was counterbalanced so that every source argued both pro and
con on his or her topic, although each subject would see only the pro or the
con message.
The results for the immediate aftertest (see Table 9.4) show that the high-
credibility source did produce more opinion change on three of the four topics.
The exception was the topic of the future of movies, where the results show
slightly more opinion change for the low-credibility source.
The retest of opinion after four weeks produced a striking finding. Results
for this retest are presented in Figure 9.1 for all four topics combined.
The figure shows that when the subjects were retested after four weeks,
the amount of opinion change retained was approximately equal for the high-
credibility and low-credibility sources. But for the low-credibility source, there
appeared to be greater opinion change after four weeks than there was immediately
after receiving the communication. This was the second occurrence of what
TABLE 9.4 Subjects who changed their opinion in direction of communication for
high- and low-credibility sources
High-Credibility Source Low-Credibility Source
(A) (%)
Antihistamines 23 13
Atomic submarines 36 0
Steel shortage 23 - 4
Future of movies 13 17
Note: The number in the table is the net percentage of subjects who changed their opinions in the
direction of the communication, or the percentage who changed in the direction of the communication
minus the percentage who changed in the opposite direction.
S OU R C E : Adapt ed f r om C. I, Hovland, I. L. J anis, and H, H, Kelley, Cofrrnrunic afiorr and Persuasi on (New
Haven, Corm.: Yale LJniver~itY press, 1953), p 30 CoP~fight @ 1 g53 by Yale University press. Reprinted
by permission,
24
22
20
18
16
14
12
10
8
1
()~
Immediate 4 weeks
Time Interval
FIGURE 9.1 Changes in extent of agree-
ment with high-credibility and low-credibility
sources after four weeks
S OU R C E : From C. 1. Hovland and W. Weiss, The influ-
ence of Source Credibility on Communication Effective-
ness, Pub/it Opiniorr Quarrerfy 15 (1951): 646.
Copyright 1951 by Princeton University Press. Reprinted
by permission of the University of Chicago Press.
Hovland, Lumsdaine, and Sheffield earlier had called a sleeper effect. Hovland
and Weiss did some further research and found that this was not due to the
forgetting of the source, suggested in the earlier study, but to a tendency after
the passage of time to dissociate the source and the opinion.
Much research since the Hovland and Weiss experiment has gone into
attempting to find the dimensions of source credibility. Hovland and Weiss had
suggested that the dimensions of expertness and trustworthiness might be
imPOrtant. Many of these studies used factor analysis of rating scales applied to
speakers to trY to find the common dimensions used in the ratings. In one of
the more comprehensive of these studies, Whitehead had subjects rate two
speakers on 65 semantic differential scales solely on the basis of tape-recorded
introductions (Whitehead, 1968). Whitehead found four dominant faCtOI%:
trustworthiness, professionalism or competence, dynamism, and objectivity. The
.-
trustworthiness factor was based on the scales right-wrong, honest-dishonest,
trustworthy-untrustworthy, and just-unjust. The professionalism or competence
factor was based on the scales experienced-inexperienced and has professional
mannerlacks professional manner. The dynamism factor was based on the
scales aggressive- meek and active-passive. The objectivity factor was based on
the scales open minded-closed minded and objective-subjective.
Whiteheads results are similar to the suggestions of Hovland and Weiss in
that the y showed trustworthiness to be an important dimension. The profes-
sionalism or competence dimension is also similar to Hovland and Weisss
dimension of expertness, although it differs in dealing more with manner of
presentation than with the actual knowledge that a person might possess.
Whiteheads research suggests that source credibility is more complicated than
that, however, with dynamism and objectivity being important components.
If a high-credibility source is effective in producing attitude change, does
that source lose or gain effectiveness if it becomes associated with a number of
messages? This problem can come up in the field of advertising, where celebrities
may be hired to endorse a number of different products. For instance, basketball
superstar Michael Jordan was at one time endorsing products for 14 companies.
An experiment by Tripp, Jensen, and Carlson (1994) exposed viewers to advertise-
ments that involved endorsements of Visa credit cards, Kodak film, Colgate
toothpaste, and Certs breath mints by Dustin Hoffmann and Matthew Broderick.
When one of these celebrities endorsed four products, he was perceived as less
trustworthy and less of an expert than when he endorsed one or two. The
attitude toward the ad also became more negative with four products than with
one or two. So there is evidence that making multiple endorsements can reduce
the effectiveness of a high-credibililty source.
Still other research has been done on source credibility since the original
Hovland and Weiss study. Some researchers have challenged the existence of a
sleeper effect. Gillig and Greenwald (1974) were unable to produce a sleeper
effectthat is, a statistically significant increase in opinion change for a group
exposed to a low-credibility sourcein seven replications of an experiment
designed to show this effect. Furthermore, their review of the literature indicated
no previous study had really shown that kind of sleeper effect. What the earlier
studies, including that of Hovland and Weiss, had shown was a significant
difference in the effects of high- and low-credibility sources over time, but that
is not the same as a significant increase in opinion change for a group exposed
to a low-credibility source.
Other research has shown additional support for a sleeper effect, however.
Cook and Flay (1978) used the term absolute sleeper effect to refer to the kind
of change mentioned abovea statistically significant increase in attitude change
over time for a group exposed to a low-credibility source. They report that
demonstrably strong tests of the absolute sleeper effect have recently been
conducted, and they repeatedly result in absolute sleeper effects (p. 19).
In applying source credibility research, one should remember that the same
source will not have high credibility for all audience members. In an effort to
get newly eligible (teen-age) voters to vote, the Rock the Vote campaign enlisted
Madonna and Rapper Ice-T for their television public service announcements.
The same sources probably would not work with older audience members.
FEAR APPEALS
Another common tactic in mass communication is to threaten or arouse some
fear in the audience. Films shown to teenagers to promote safe driving sometimes
show terrible traffic accidents and what they do to people. A television commer-
cial for an insurance company arouses fear by saying, You need something to
help keep these promises, even if youre not there.
The book Communication and Persuasion (Hovland et al., 1953) describes
a classic experiment by Janis and Feshbach aimed at investigating the effective-
ness of fear appeals in producing attitude change. On the basis of learning theory,
a key element in the Hovland approach, it can be predicted that a strong fear
appeal would lead to increased attitude change because it would increase arousal
and bring about greater attention and comprehension. Motivation to accept the
recommendations of the communication would also be increased. In reinforce-
ment theory terms, learning and practicing the recommended procedure should
become associated with the reinforcement of reduced fear and anxiety. On the
other hand, the researchers realized that a high degree of emotionaI tension
could lead to spontaneous defensive reactions and the possibility of the audience
distorting the meaning of what is being said. Part of their research purpose was
to investigate this potentially adverse effect of a strong fear appeal.
Janis and Feshbach designed an experiment that was based on three different
messages with three different levels of fear appeal. They selected dental hygiene
as their topic. The subjects were the entire freshman class of a large Connecticut
high school. The class was randomly divided into four groups, three of which
were to get the different fear messages and one of which was to be a control
group. The basic message, common to all three fear levels, was a standard lecture
on dental hygiene. The level of fear was varied primarily through changing the
material used to illustrate the lecture. In the minimal fear appeal message, the
illustrative material used x-rays and drawings to represent cavities, and any
photographs used were of completely healthy teeth. In the moderate fear appeal
version, photographs of mild cases of tooth decay and oral diseases were used.
In the strong fear appeal version, the slides used to illustrate the lecture included
very realistic photographs of advanced tooth decay and gum diseases. The strong
fear condition also contained some personalized threats, such as the statement,
This can happen to you. The control group received a lecture on the structure
and function of the human eye.
Subjects were given a questionnaire asking specific questions about their
dental hygiene practices one week before the lecture and one week after.
Comparison of these questionnaires would show whether subjects changed their
dental hygiene behavior after being exposed to the various types of messages.
TABLE 9.5 Conformity to dental hygiene recommendations in subjects who
received messages with different levels of fear
Strong Moderate Minimal
Fear Fear Fear Control
Appeal Appeal Appeal Group
(%) (%) (%) (%)
Increased conformity 28 44 50 22
Decreased conformi ty 20 22 14 22
No change 52 3 4 38 56
S OUR C E : Adapted from C. 1. Hovland, 1. L. Janis, and H. H. Kelley, Communication and Persuasion (New
Haven, Corm.: Yale University Press, 1953), p. 80. Copyright 01953 by Yale University Press. Reprinted
by permission,
The results in Table 9.5 show that the minimal fear appeal was the most effective
in getting the students to follow the dental hygiene recommendations in the
lecture. The strong fear appeal was the least effective. This was definite evidence
that a fear appeal can be too strong and can evoke some form of interference
that reduces the effectiveness of the communication.
This experiment had several strengths that have not always been present
in later attitude change studies. One is that the message was shown to have an
effect on reported behavior, and not just on a paper-and-pencil measure of a
hypothetical attitude. The field of attitude change research was involved in a
controversy a few years later in which many studies were criticized for producing
slight changes in unimportant attitudes. Second, the study showed that the
persuasive messages used by Janis and Feshbach produced long-term attitude
change. Another criticism of some later attitude change studies is that they dealt
only with short-term attitude change, often measured immediately after the
message. Janis and Feshbach went back to their subjects a year later and still
found the differences in attitude change between their experimental groups.
The Janis and Feshbach study was the first of a number of studies on fear
appeals. Not all these studies have agreed with the finding that strong fear
produces less attitude change. One possible explanation for the findings of Janis
and Feshbach is that the recommendation of brushing your teeth properly was
not seen as a believable recommendation for preventing the kinds of horrible
consequences presented in the strong fear appeal message. Other studies have
shed some light on this possibility. Leventhal and Niles (1964) presented a
message to audiences at a New York City health exposition recommending that
they get a chest x-ray and that they stop smoking. The message was presented
to different groups with differing levels of fear: high fear (featuring a color movie
of removal of a lung), medium fear (featuring the same color movie but without
the graphic scene of a lung removal), and low fear (with no movie). They found
that the amount of reported fear in audience members was correlated with stated
intentions to stop smoking and to get a chest x-ray. These results suggest that
-.
fear facilitates attitude changethe opposite of the Janis and Feshbach
191
finding.
-----
mat could account for this difference? Possibly the difference was due to the
degree to which the recommendations appeared to be effective. Toothbrushing
maY not have seemed adequate to prevent the rotted teeth and bloody gums
seen in the Janis and Feshbach experiment. In contrast, stopping smoking may
appear to be a believable recommendation for preventing lung cancer.
On the basis of the Janis and Feshbach experiment and other research, Janis
(1967) formulated a model suggesting that the relationship between fear appeal
and attitude change is curvilinear. This model specifies that low and high levels
of fear in a message will lead to small amounts of attitude change and that
moderate levels of fear will lead to the greatest amount of attitude change. This
inverted U-shaped curve was the dominant view of the relationship between fear
and attitude change for many years.
A different approach to the role of fear in persuasion, protection motivation
theory, was developed by Rogers (1975) as an alternative to Janiss curvilinear
theory. Advocates claim that the PM model is superior to the curvilinear
approach because it provides a clearer prescription for how to develop messages
that can influence adaptive behavior (Tanner, Hunt, & Eppright, 1991, p. 37).
Building on the Leventhal and Niles research as well as on some other
research of his own, Rogers (1975) developed a model that summarizes three
key elements in the operation of a fear appeal: (1) the magnitude of noxiousness
of a depicted event, (2) the probability of that events occurrence, and (3) the
efficacy of a recommended response. Each component brings about a process
of cognitive appraisal, and these cognitive appraisal processes then determine
the amount of attitude change (see Figure 9.2). That is, when audience members
receive a fear appeal, they weigh it in their minds. If the portrayed noxiousness
or horribleness of the event is not believed, or the event is thought to be
unlikely, or the recommended actions are not believed to be adequate to deal
with the threat, then attitude change is not likely. Rogers calls his model a
protection motivation theory of fear appeals and attitude change. Attitude change
is said to be a function of the amount of protection motivation aroused by the
cognitive appraisal that the audience member goes through.
Reardon (1989) discussed how fear appeals might best be used in a commu-
nication about AIDS that is addressed to teenagers (who typically perceive the
threat of death as remote), She suggested that mass media messages aimed at
adolescents emphasize the more immediate consequences of the disease, includ-
i ng mental problem5, skin rashes and sores, and the negat i ve ef f ect on a
teenagers social lif
e
. She suggested, furthermore, that the media campaigns be
Combined with interpersonal question-and-answer sessions, in which discussion
can bring out i~ormation regarding methods for avoiding the problems described
or depicted in the media messages.
The effectivene55 of fear aPPea15 in condom advertisements that stress the
prevention of AIDS was studied by Hill (1988). He found that subjects had more
positive attitudes toward a moderate fear appeal commercial (stating that sex
can be a risky business) than either a nonfear appeal (stressing the sensitivity
Components of
Fear ;ppeal
1
Magrdtude
of noxiousness
t
I
Cognitive Mediating Processes
I
r
__ -
1
A
Appraised 1
I
severity
1- -_
7
J
r
_
t
1
I r
-1
A
Expectsney Protection 1-
of exposure
1 motivation I
L J LJ
r
_
7
A
Befief in efficacy
of coping response
I
I___-d
Attitude Change
FIGURE 9.2 A model of protection motivation theory
S OUR C E : From R. W. Rogers, A Protection Motivation Theory of Fear Appeals and Attitude
Change, Jouma/of PsycFro/ogy91 (1975): 99. Reprinted with permission of the Helen
Dwight Reid Educational Foundation. Published by Heldref Publications, 4000 Albemarle St.,
N. W., Washington, D.C. 20016. Copyright @1975.
of the condom and saying nothing about AIDS) or high fear appeal (mentioning
the possibility of death). Hill speculated that a nonfear appeal may appear
inappropriate in an AIDS environment, but that a high fear appeal may be viewed
as too threatening when it is combined with the individuals existing level of
AIDS-related anxiety.
Fear appeals are attempting to bring about changes in peoples motivations,
a tricky enterprise at best. As Ross (1985) noted, Audience motivation through
symbolic interaction is terribly complex; we should be wary of infallible motive
appeals (p. 48).
RESISTANCE TO COUNTERPROPAGANDA
With so much energy being devoted to changing attitudes, it might be extremelY ;
I
useful to discover some methods of making attitudes resistant to change.
Lumsdaine and Janis report a study in Communication and Persuasion (Hovland ,
et al., 1953) that deals with building resistance of an attitude to change. Their
experiment follows up on the earlier work on one-sided and two-sided messages ;
by Hovland, Lumsdaine, and Sheffield.
~
Groups not exposed to Groups exposed to
later counterpropaganda later counterpropaganda
Lumsdaine and Janis produced one-sided and two-sided messages arguing
that Russia would be unable to produce large numbers of atomic bombs for at
least five years. This was a realistic issue for differences of opinion in the early
1950s. The one-sided message argued that the Russians lacked some crucial
secrets, that their espionage was not effective, and that Russia was lacking in
industry. The two-sided message added brief mentions of the arguments that
Russia had uranium mines in Siberia, that it had many top scientists, and that
its industry had grown since the war. Several weeks before the messages were
presented, all subjects were given a questionnaire to determine their initiaI
opinions. One group received the one-sided message and another received the
two-sided message. A week later, half of each group was exposed to an opposing
communication from a different communicator arguing that Russia had probably
already developed the atomic bomb. This counterpropaganda brought out some
new arguments not included in the two-sided message. Both the initial messages
and the counterpropaganda were presented in the form of recorded radio
programs. Finally, all subjects were given another questionnaire.
The key question, asked in both the initial and final questionnaires, was this:
About how long from now do you think it will be before the Russians are really
producing large numbers of atomic bombs?
The net opinion change from initial to final questionnaire for those who
received counterpropaganda and those who did not in both the one-sided and
two-sided conditions is presented in Figure 9.3. The results show that for those
receiving no counterpropaganda, the one-sided and two-sided messages were
about equally effective. This replicates the finding of the earlier study by Hovland,
FIGURE 9.3 Comparison of the effectiveness of Programs I and 11:
Changes in opinions concerning the length of time before Russia pro-
duces large numbers of atomic bombs
S OURC E : From A, Lumsdaine and 1, Janis, Resistance to Counterpropaganda Produced
by One-Sided and Two-Sided Propagnda Presentations, Pub/it Opinion Quarterly 17
(1953): 316. Copyright O 1953 by Princeton University. Reprinted by permission of the
University of Chicago Press.
Net Change i n the Posi ti ve LJ i rechon
Program I
(one side)
P 1
2%
I
I
v
1~
Difference in Favor of RO~ 11:
5%
I
Lumsdaine, and Sheffield. The results show a striking difference for those
receiving counterpropaganda, however. Those receiving a one-sided message
showed almost no remaining attitude change after they were exposed to counter-
propaganda. In contrast, those receiving a two-sided message showed almost as
much attitude change remaining after counterpropaganda as they did when they
werent exposed to counterpropaganda.
One of the advantages of the two-sided message over the one-sided message,
then, is that it is more effective in building resistance to later persuasive efforts.
Lumsdaine and Janis speak of the recipient of the two-sided message as
becoming inoculated. This is a medical analogy that William McGuire and
Demetrios Papageorgis drew upon later in developing their inoculation theory.
INOCULATION THEORY
McGuire and Papageorgiss theory rests on the medical analogy that is suggested
by its name. They point out that most people have many unchallenged beliefs,
and that these beliefs can often be easily swayed once they are attacked because
the person is not used to defending them. The situation is similar to that in the
medical field when a person is brought up in a germ-free environment and is
suddenly exposed to germs. That persons body is vulnerable to infection because
it has not developed any resistance. Such a person can be given resistance either
by supportive treatmentgood diet, exercise, rest, and so forthor by inocu-
lation, a deliberate exposure to a weakened form of the germ that stimulates
the development of defenses. In the medical area the inoculation approach has
been more effective than supportive treatment in producing resistance. The word
immunization can be applied to either of these methods of building immunity
the supportive approach or the inoculation approach.
McGuire and Papageorgis have conducted a number of experiments to test
this theory. One of the first (McGuire & Papageorgis, 1961) tested the basic
prediction that the supportive approach of preexposing a person to arguments
supporting basic beliefs would have less immunizing effectiveness than the
inoculation approach of preexposing the person to weakened, defense-stimulating
forms of arguments attacking the beliefs. It also tested a second hypothesis that
active participation during exposure to a defense should be less effective than
passive participation in producing immunity to later persuasion. The researchers
made this prediction because they theorized that subjects would not be accus-
tomed to active participation in defending their basic beliefs and so would not
do it very well. Furthermore, they thought that active participation might
interfere with the reception of any defensive material presented. (There was also
a complicated third hypothesis that need not concern us here.)
McGuire and Papageorgis selected for their study some beliefs that were
hardly ever attacked in our culture, which they called cultural truisms. The four
beliefs were these: Everyone should get a chest x-ray each year in order to detect
any possible tuberculosis symptoms at an early stage. The effects of penicillin
91
have been, almost without exception, of great
THEORIES OF PERSUASION 195
benefit to mankind. Most forms
of mental illness are not contagious. Everyone should brush his teeth after every
meal if at all possible. These cultural truisms were so widely believed that
control groups of subjects rated them at an average level of 13.26 on a scale
ranging from 1 for definitely false to 15 for definitely true.
Subjects took part in two one-hour experimental sessions held two days
apart. The first exposed subjects to the two types of immunizing material
designed to make the basic beliefs (cultural truisms) more resistant to change;
the second exposed subjects to strong counterarguments attacking the basic
beliefs. Questionnaires were administered at the end of each session to measure
strength of acceptance of beliefs.
The two major types of immunizing material presented to subjects were
supportive and refutational. The supportive material was made up of argu-
ments supporting the cultural truisms. The refutational material consisted of
possible counterarguments against the cultural truisms together with refutations
of these counterarguments. The amount of participation in the defense was varied
primarily by having subjects write in a high-participation condition and read in
a low-participation condition. Each subject was tested on one cultural truism
for which he or she received no immunization but did receive the later counter-
arguments. The average scale position for these beliefs after they were attacked
was 6.64, compared with the average level of 13.26 prior to attack. This result
shows that the cultural truisms were highly vulnerable to attack if no immu-
nization was given.
McGuire and Papageorgis found, as they had predicted, that the refutational
defenses were more effective in making the cultural truisms resistant to change
than were the supportive defenses. After the supportive defenses, the counter-
arguments were able to reduce the belief in the cultural truisms to an average
rating of 7.39, only slightly better than the 6.64 level achieved when there was
no prior prepamtion at all. After the inoculation defenses the counterarguments
were able to reduce the beliefs in the cultural truisms only to an average scale
rating of 1().33. The authors also found support for their second hypothesis: the
Passive (reading) conditions had a greater effect in making beliefs resistant to
Persuasion than did the active (writing) conditions.
The McGuire and papageorgis experiment was Iimited in one respect that
needed further investigation. The attacks on the cultural truisms that were
Presented and then refited in the inoculation were the same attacks that were
Presented in the next session when the cultural truisms were assailed. It was
not clear whether presenting and refuting one set of attacks would also provide
later immunity to a different set of attacks. This question was investigated in
another experiment by papageorgis and McGuire (1961).
papageorgis and McGuire predicted that a kind of generalized immunity
ould develop when people were exposed to attacks on basic beliefs and
refutations of those attacks. That is, they predicted that this procedure would
eveloP a general resistance that would make the basic belief unlikely to change
Ven
when it was exposed to attacks that were not the same. They expected
I
I
this result for two reasons: (1) The experience of seeing the first attacks refuted
could lower the credibility of the later attacks. (2) Preexposure to attacks may
make a person more aware that his or her beliefs are indeed vulnerable and
motivate the person to develop additional supporting arguments. Their results
showed that inoculation led to an immunity to differing counterarguments that
was almost as strong as the immunity to the same counterarguments. In fact,
the final attitude positions in these two conditions were not significantly
different. This, of course, increases the potency of an inoculationthe developers
of the inoculation program do not have to anticipate all the attacks on a belief
to which a person might later be exposed.
KATZS FUNCTIONAL APPROACH
The two major theoretical approaches to attitude changethe learning
theory approach, primarily associated with Hovland, and the consistency theory
approach, primarily associated with Festinger, Newcomb, Heider, and Osgood
and Tannenbaum existed side by side with little apparent relation to one
another for some time. But eventually researchers became interested in recon-
ciling these rather different ways of dealing with attitude change. Daniel Katz
and his colleagues, Irving Sarnoff and Charles McClintock, tackled this problem,
and it led them to develop the functional approach to attitude change.
These authors were trying to bring together two different models of human
behavior that have been presented over the yearsthe rational model and the
irrational model. The irrational model suggests that human beings are nonthinking
creatures whose beliefs are easily influenced by people around them and who
even can have their perception of reality influenced by their own desires. The
rational model suggests that human beings are intelligent and critical thinkers
who can make wise decisions when given ample information. How can both of
these models be true? Katz and his associates suggest that the answer to this
dilemma is that human beings are both rational and irrational, depending on the
situation, the motivations operating at the time, and so forth. And they argue
that this tendency for people to operate with different ways of thinking has
important implications for understanding attitude change.
Katz argues that both attitude formation and change must be understood
in terms of the functions that attitudes serve for the personality. As these
functions differ, so will the conditions and techniques of attitude change. Katz
points out that much of the earlier research on mass communication dealt with
factors that are not really psychological variables, such as exposure to a motion
picture. Since being exposed to a motion picture can serve different functions
for different individuals, Katz argues that the researcher dealing only with
exposure to a film is not really able to understand or predict attitude change.
Katz makes the key point that the same attitude can have a different motivational
basis in different people. He suggests that unless we know the psychological
need which is met by the holding of an attitude we are in a poor position to
predict when and how it will change (Katz, 196o, p. 170).
Katz identifies the following four major functions that attitudes can serve
for the personality:
1. Tbe instrumental, adjustive, or utilitarian function. Some attitudes are
held because people are striving to maximize the rewards in their external
environments and minimize the penalties. For instance, a voter who thinks taxes
are too high might favor a political candidate because that candidate promises
to reduce taxes.
2. The ego-defensive function. Some attitudes are held because people are
protecting their egos from their own unacceptable impulses or from knowledge
of threatening forces without. Feelings of inferiority are often projected onto a
minority group as a means of bolstering the ego, This would be an example of
an attitude of prejudice serving the ego-defensive function.
3. The value-expressive function. Some attitudes are held because they allow
a person to give positive expression to central values and to the kind of person
one feels he or she is. For instance, a teenager who likes a particular rock and
roll group is expressing his or her individuality through this attitude,
4. The knowledge function. Some attitudes are held because they satisfy a
desire for knowledge or provide structure and meaning in what would otherwise
be a chaotic world. Many religious beliefs serve this function, as do other
attitudes such as the shared norms of a culture.
Katz presented a table (see Table 9.6, p. 198) summarizing the origin and
dynamics, the arousal conditions, and the change conditions for attitudes serving
each of the four functions.
Katz warns that an attempt to change an attitude may baclctiie if it is not
based on an understanding of the functions the attitude is serving. For instance,
an attempt to change attitudes of prejudice by presenting factual information
on the accomplishments of minority group members would be an attempt to
change the attitudes as if they were serving the knowledge function. It is not
likely to succeed if the attitudes of prejudice are held for ego-defense reasons.
ATTITuDES AND BEHAVIOR
Despite all the research on attitude change, researchers for a long time neglected
an important question. In its more general form, the question is whether attitudes
as they are measured by social science methods have any real relation
to behavior. In its more specific form, the question is whether attitude change
produced by persuasive messages is accompanied by any meaningful change
in behavior.
One early study had indicated that attitudes might not bear much of a
elationship to behavior. A social scientist, Richard LaPiere, traveled in the early
1930s around the United States with a young Chinese couple. They made 251
TABLE 9.6 Determinants of attitude formation, arousal, and chance in relation to
tvi3e of function
Function Origin and Dynamics Arousal Conditions Change Conditions
Adj ust ment Utiiity of attitudinal
object in need
satisfaction.
Maximizing external
rewards and
mi ni mi zi ng
puni shment s
Ego defense Protecting against
internal conflicts and
external dangers
Value
expression
Maintaining self-
identity; enhancing
favorable self-image;
self-expression and
self-determination
Knowl edge Need for under-
standing, for
meaningful cognitive
organization, for
consi stency and
clarity
1,
2.
1.
2.
3.
4.
1.
2.
3
1
Activation of needs
Saiience of cues
associated with
need satisfaction
Posing of threats
Appeals to hatred
and repressed
impulses
Rise in frustrations
Use of authoritarian
suggestions
Saiience of cues
associated with
vaiues
Appeals to
individual to
reassert self-image
Ambiguities which
threaten self-
concept
Reinstatement of
cues associated
with old problem or
of old problem itself
1.
2.
3.
4.
1.
2,
3.
1.
2
1,
2,
Need deprivation
Creation of new
needs and new
ievels of aspiration
Shifting rewards
and punishments
Emphasis on new
and better paths for
need satisfaction
Removal of threats
Catharsis
Development of
seif-insight
Some degree
of dissatisfaction
with self
Greater appropri-
ateness of new
attitude for
the self
3. Control of all
environmental
supports to
undermine old
values
Ambiguity created
by new information
or change in
environment
More meaningful
information about
probiems
S OUR C E : From D, Katz, The Functional Approach to the Study of Attitudea, Public Opinion Quarter/y
24 (1960): 192. Copyright 1960 by Princeton University, Reprinted by permission of the University of
Chicago Press.
visits to hotels and restaurants, and in only one case were they refused service.
Six months later, LaPiere sent a questionnaire to each establishment asking: Will
you accept members of the Chinese race as guests in your establishment? He
received replies from 128 of these businesses. The responses from 92 percent
of the restaurants and 91 percent of the hotels were no. Only one person gave
a definite Yes (Lapiere. 1934). This classic study, then, provided some evidence
that peoples verbal reports of their attitudes might not be very good predictors
of their actual behavior.
In an address in 1963, Leon Festinger, the psychologist who developed the
theory of cognitive dissonance, raised some basic questions about attitude change
experiments and subsequent behavior. Festinger (1964) said he had been reading
a manuscript by Arthur R. Cohen when he came across the statement that very
little work on attitude change had deah explicitly with the behavior that may
follow a change in attitude. Festinger was intrigued by this notion and attempted
to find as many studies as he could that showed an effect of attitude change on
subsequent behavior change. He found only three. One of these was the Janis
and Feshbach study of fear appeals. Their study did not investigate actual
behavior change, but it did look at verbal reports of toothbrushing behavior and
other dental hygiene behavior. Festinger was willing to accept this verbal report
since it did purportedly deal with actual behavior. In all three of the studies
that Festinger found, there seemed to be a sIight inverse relationship between
attitude change and behavior change. For instance, in the Janis and Feshbach
study, the individuals who indicated the most concern about their teeth after
receiving the persuasive messages showed the least change in their reported
behavior. Festinger argued that this inverse relationship indicates that the
relationship between attitude change and behavior is not a simple one.
One reason attitude change might not be automatically followed by behavior
change, Festinger (1964) suggested, is that the environmental factors that had
produced an original attitude would usually still be operating after that attitude
was changed. Thus there would be a tendency for an attitude to revert to its
original position after exposure to a persuasive message.
Festinger was suggesting to attitude change theorists the disturbing possi-
bility that they had conducted hundreds of experiments on variables that make
very little difference in terms of human behavior.
Realizing that the prediction of a specific behavior depends on a number
of factors in addition to some kind of measure of attitude, Martin Fishbein
attempted to develop a model that would include all the important factors. The
model (see Ajzen & Fishbein, 1970) takes the form of the following equation:
B - BI = [A,C,] WO + [NB(MC )] W,
Although the model looks complicated, it becomes easier to understand
when it is put into words. The letters in the equation can be translated as follows:
BI =
A
act
=
NB .
MC =
WO and w, =
overt behavior
the behavioral intention to perform that behavior
attitude toward performing a given behavior in a given
act situation
normative beliefs, or beliefs that significant others think
one should or should not perform the behavior
motivation to comply with the norm C
regression weights to be determined empirically
The equation can be rephrased in the following English sentence: A persons
intention to perform a given behavior is a function of (1) the persons attitude
toward performing that behavior and (2) the persons perception of the norms
governing that behavior and the individuals motivation to comply with those norms.
This model brings in some of the key situational factors, particularly the
beliefs that other people have about the behavior and the individuals motivation
to conform to those beliefs. If precise measurements could be made of all the
variable quantities in the model, it should be possible to make rather exact
predictions of behavioral intention and then of actual behavior. Fishbein (1973)
reports that a number of experiments using the model to predict behavioral
intention have produced multiple correlations of about .80, which are quite high.
These experiments also found correlations between behavioral intention and
overt behavior of .70, so all the key parts have been supported. In a continuation
of the Fishbein research, Ajzen (1971) has used the Fishbein model to demon-
strate behavioral change as the result of persuasive communication, the pheno-
menon that Festinger had difficulty finding in 1963.
One significant development in attitude change research is that many
researchers are now including behavioral measures in their studies. For instance,
many fear appeal studies now include behavioral measures such as the disclosing
wafer test of how well teeth are actually cleaned (Evans, Rozelle, Lasater,
Dembroski, & Allen, 1970) or the actual act of going to get a shot or vaccination
(Krisher, Darley, & Darley, 1973). Similarly, the Stanford project aimed at reducing
heart disease through communication used such behavioral measures as blood
pressure, cholesterol level, weight, and number of cigarettes smoked (Maccoby
& Farquhar, 1975).
CLASSICAL CONDITIONING OF ATTITUDES
Some other researchers besides Hovland have also attempted to apply learning
theory to attitude change. In particular, Staats and Staats have applied classical
conditioning to the learning of attitudes (Staats, 1968).
Staats and Staats begin by demonstrating the application of classical condition-
ing to the learning of the emotional meaning of language. They point out that
in our everyday experience certain words are systematically paired with certain
emotional experiences. For instance, words like joy, happy, play, dinner, pretty,
and good are typically paired with positive emotions, while words like angry,
hurt, dirty, awful, sick, sad, and ugly are typically paired with negative emotions.
In the terms of classical conditioning, the emotional stimuli can be considered
the unconditioned stimuli that elicit emotional responses. When a word stimulus
is systematically paired with such an unconditioned stimulus, the word should
become a conditioned stimulus and also elicit the emotional responses.
In an experiment to test this possibility, researchers exposed two groups
of subjects to a list of spoken words (Staats, Staats, & Crawford, 1962). For the
experimental group, 9 of the 14 times a subject was presented with the word
large, it was followed by a negative stimuluseither a loud noise delivered by
earphone or a shock to the right forearm. Both noise and shock were set at a
level where they were uncomfortable but not painful. Members of the control
group also received the negative stimuli 9 times, but they were paired with
different filler words other than large. Results showed that those subjects who
experienced the word large being paired with the aversive stimulus came to
display an emotional reaction, as measured by the galvanic skin response (GSR),
when presented with this word. This was not true for the control group.
Furthermore, measurements with semantic differential scales showed that large
had acquired a negative rating on the evaluative scale for the experimental group
that it did not have for the control group.
Staats and Staats (1957) went on to hypothesize that this conditioning of
meaning should work from word to word as well as from a physical stimulus to
a word. In another experiment, nonsense syllables such as yof, laj, xeb, wub,
giw, and qug were presented visually on a screen while words were presented
aurally. For one group of subjects, two of the nonsense syllables were always
paired with words that had high loadings on evaluative meaning, such as beauty
win, gtft, sweet, and honest. A different high evaluation word was used in every
pairing, so that subjects would not associate particular pairs of words. The other
four nonsense syllables were paired with words that had no systematic meaning.
For another group of subjects, the procedures were identical except that the
two nonsense syllables were always paired with words with negative emotional
meaning, such as tbieJ bitteq ugly, sad, and worthless. Subjects were later given
semantic differential scales to measure their evaluative meanings for the nonsense
syllables, and those scales showed positive ratings in general for the group receiv-
ing positive stimulus words and negative ratings in general for the group receiving
negative stimulus words.
Staats and Staats argue that attitude is nothing more than this kind of
emotional meaning for a word that has been established by classical conditioning.
In another experiment (Staats & Staats, 1958), names of nations and familiar
masculine names were used as the conditioned stimulus rather than nonsense
syllables. The national names were Dutch or Swedish, and the masculine names
were Tom and Bill. For all four of these, subjects would be expected to have
existing attitudes on the basis of prior experience. Thus, the experiment was
reallY a study of attitude change. The experiment showed that pairing any of
he four words with either positive or negative words could condition the
subjects attitudes in either a positive or a negative direction. The Staats and
taats research provides a theoretical explanation for some of the propaganda
evices, such as glittering generality; which attempts to link a person or idea
[o a virtue word, or <name calling, which attempts to associate a person or idea
with a bad label.
This conditioning of attitudes also seems to be related to much of what goes
n n advertising. Many product names, such as Ipana or Qantas, are essentially
onS
ense syllables to the public when they are first introduced. A major goal of
advertising is to associate them with positive words or experiences, which,
202 T
through CO1
IE SOCIAL-PSYCHOLOGICAL APPROACH
ditioning, might give them a positive meaning. A slogan such as
Coke is the real thing is attempting to transfer positive associations to Coke.
In the cases of some other products, such as Fab or Sprite, the product name
might be chosen because of positive associations that it already has.
TECHNIQUES OF PERSUASION
We now turn to four important techniques commonly used in persuasion: use
of pictures, appeals to humor, appeals to
sex, and extensive repetition of an
advertising message. Audiences and communicators need to understand their
applicationsand their potential misuse.
Use of Pictures
A common technique in communication, particularly in advertising, is to use
pictures to accompany a textual verbal message.
Pictures can add to the persuasiveness of a message in the following ways
(Miniard, Bhatla, Lord, Dickson, & Unnava, 1991):
1. They can help attract attention to the message.
2. They can enhance learning of the messages content.
3. A positive emotional response to the picture might transfer to other
aspects of the message.
The use of pictures with persuasive messages is probably more complicated
than that, however. We probably need to look in greater detail at some charac-
teristics of the picture and at the cognitive state of the individual receiving the
message. One important aspect of the picture is whether or not it is relevant
to the persuasive message. One important aspect of the cognitive state of the
individual is the level of involvement in processing the message.
Miniard et al. (1991) looked at the effects of receiver involvement and
picture relevance in advertisements on attitudes toward the product and purchase
intentions. The two variables interacted, with irrelevant pictures having their
greatest impact on attitude change when receiver involvement was low and
relevant pictures having their greatest impact when receiver involvement was high.
Appeal s t o Humor
The use of humor is a popular technique in communication. Many public
speakers obviously believe in the importance
of beginning their talks with a
humorous story. Studies have suggested that 15 to 20 percent of television
commercials contain some element of humor (Kelly & Solomon, 1975; Duncan
& Nelson, 1985).
9 / THEORIES OF PERSUASION 2 0 3
Humor could also be used to help create a positive mood when attempting
to communicate about a serious toPic, such as AIDS. One company has used a
humorous approach to selling condoms with its Rubber Ducky condoms
campaign (Frankenberger & Sukhdial, 1994). In a similar approach, a film strip
trying to promote AIDS preventive behavior featured a character called Captain
Condom (Frankenberger & Sukhdial, 1994).
In the typical study of the effects of humor on attitude change or other
variables in the hiemrchy of effects, different groups are exposed to different
versions of the same messageone with humor and one without. For instance,
Brooker (1981) examined the effects of humor in two commercialsone for a
toothpaste and one for a flu vaccine. Examples of the humorous appeals used
in his study appear in Table 9.7.
When attitude change or persuasion is the dependent variable of interest,
most studies have not found a signillcant effect due to humor (Gruner, 1965,
1967, 1970, 1972; Brooker, 1981). Other studies of the effectiveness of humor
indicate that it has more of an effect on lower-order communication effects
(responses lower in the response hierarchy) than on higher-order communication
TABLE 9.7 Humor appeals used in Brookers experiment
Toothbrush:
Vaccine:
Toothbrush:
Vaccine:
Toothbrush:
Vaccine:
Toothbrush:
Vaccine:
Pun
Heres an idea with some teeth in it.
Song of the spring camper: Were tenting tonight on the old damp ground.
Li meri ck
If your lady friend turns aside her nose
Whenever you begin to propose
The halitosis demon
Might be what sends her screamin
And your toothbrush could help to solve your woes.
There was an old lade of Crewe
Who was horribly frightened of flu.
She spoilt her complexion
Through fear of infection
Having fixed on her gas mask with glue.
Joke
Detecting decay in the tooth of a beautiful young woman, the dentist said,
Whats a place like this doing in a girl like you?
A little boy waa found watching a movie by the manager in the morning.
(Why arent you in s c hool? he asked the boy, Its O. K., mi ster, Im just
9ettin9 over the flu.
One-l i ner
Oscar Levant once said, The first thing I do in the morning is brush my
teeth and sharpen my tongue.
Many flu sufferers have said, 1 hope Im really sick. Id hate to feel like this
if Im well.
OURCE: G. W. Brooker A ComPari~on of the persuasive Effec ts of Mild Humor and Mild Fear Appeals,
Jouma/ of ,4dveflisjng Io, no, 4 (I 981): 32. Used by permission.
effects (Gelb & Pickett, 1983; Duncan & Nelson, 1985). That is, humor is more
effective in attracting attention, generating liking for the communicator, and so
forth, than it is in producing attitude change or changes in behavior.
Not all studies agree, however, that humor is even effective in generating
liking for the communicator. One study showed that a woman speaker was liked
less when she used humor than when she did not (Taylor, 1974). The author
suggests that the speaker was perceived as
trying too hard to curry favor.
Similarly, another study showed that college teachers who use humor are
perceived with suspicion and hostility because they are acting contrary to
student expectations that a teachers behavior will be controlling and evaluative
(Darling & Civikly, 1984).
The research on the effectiveness of humor that has been conducted so far
should be interpreted in light of its limitations,
however. One limitation is that
the settings of the studies have often been classrooms or laboratories, which
might not be representative of the settings where humor is expected. Another
limitation is that the research has tended to be nontheoretical, with little
discussion of why humor might or might not be effective in achieving various
effects. Markiewicz (1974) has suggested that learning theory and distraction
theory are two promising theories for understanding the relationship of humor
to persuasion. A learning theory approach might suggest that humor would
provide reinforcement and thus lead to greater attitude change. A distraction
theory approach might make the prediction that humor would be distracting.
This distraction, in turn, might lead to greater attitude change by preventing
counterarguing (Festinger & Maccoby, 1964). Or distraction might lead to less
attitude change by interfering with attentiveness to the message.
It has also been suggested that the use of humor needs to be studied in
relation to other variables (Kelly & Solomon, 1975). For instance, in advertising)
is the humor more effective when it relates to the topic or when it does not?
In a commercial, should the humor come at the beginning, at the end, or
throughout the commercial?
Appeal s to SXX
The use of sexy models and other sexual appeals is a common technique in
advertising. One study indicated that more than one-fourth of magazine ads
contain obviously alluring female models (Sexton & Haberman, 1974). Further-
more, these kinds of ads are on the increase. The same study showed that ads
with obviously alluring models increased from 10 percent in 1951 to 27 percent
in 1971. Many advertisers apparently believe that sex sells. But does it?
Theoretically, the use of sex in a message could have an impact on Per
-
suasion in at least three ways (Gould, 1994):
1. The sexual material could increase attention to the message, a
necessary condition for attitude change.
2.
3.
At
9 / THEORIES OF PERSUASION 2 0 5
The sexual material could lead to arousal, and the pleasantness of the
arousal condition could transfer to the product or recommendations
of the message, leading to attitude change.
The product, institution, or recommendations in the message could
become sexualized, or pick Up a sexual charge, through their
association with the sexual stimulus. This sexual charge could then
make the product, institution, or recommendations more acceptable.
least one study suggests that a sexy model can affect the perception or
image of a product, even if there is very little logical connection between the
model and the product. Smith and Engel (1968) prepared a print ad for an
automobile in two versions. In one version, a female model clad in black lace
panties and a simple sleeveless sweater stood in front of the car. She held a
spearon the assumption that the spear might be regarded as a phallic symbol
and might lead the model to be seen as more aggressively seductive. In the other
version, there was no model. When the car was pictured with the woman,
subjects rated it as more appealing, more youthful, more lively, and better
designed. Even objective characteristics were affected. When the car appeared
with the woman, it was rated as higher in horsepower, less safe, more expensive
by $340, and able to move an average 7.3 miles per hour faster. In general, male
and female subjects responded the same way to the ads.
In contrast to the Smith and Engel study, however, a number of studies
investigating the effects of sexy models on brand recall have shown either no
effect or less recall with the sexy model (Chestnut, LaChance, & Lubitz, 1977;
Alexander & Judd, 1978). It appears that the sexy models distract the viewers
attention away from the portion of the ad presenting the product or company name.
One study suggests that for certain products, an attractive female might not
be as effective in stimulating sales as an attractive male (Caballero & Solomon,
1984). This study changed the displays for a brand of beer and a brand of tissue
that appeared at the end of an aisle in a Tom Thumb supermarket. They found
that overall, the male models tended to stimulate more beer sales among both
male and female customers than either the female stimulus or the control (no
model) treatment. There are also some clear-cut age differences in responses to
ex aPPeals in advertising, with more approval from younger people than from
older people (Wise, King, & Merenski, 1974).
Even though sex in advertising is common, it appears that there are some
isk
s in using it. Appeals to sex might be disapproved of by some audience
embers) might be misperceived or missed by others, and might distract still
others from the real purpose of the ad. Few, if any, studies exist that show a
positive effect of sex in advertising on brand recall or product sales. While the
mith
and Engel study shows a sew model having the effect of increasing the
avorable evaluation of an automobile in an ad, it did not test for brand recall
after eeing the ad It is possible that the subjects did not recall the brand name
f he automobile any better with the sexy model than without, and this would
defeat the purpose of the ad.
Jib Fowles, author of an article Advertisings Fifteen Basic Appeal% draws
As a rule, though, advertisers
this conclusion about appeals to the need for sex.
have found sex to be a tricky appeal, to be used sparingly. Less controversial
and equally fetching are the appeals to our need for affectionate human contact
(1982, p. 278).
Effects of Repeti ti on
Many mass communication messages, particularly advertisementswhether
commercial or politicalare repeated extensively. There are a number of reasons
why this might be a good idea. Not all audience members will be watching at
the same time, or, in the print media, not all readers will see a single printing
of an advertisement. Another advantage of repetition is that it might remind the
audience of a source for a message from a high-credibility source, and thus
prevent the drop-off in attitude change from a high-credibility source found over
time by Hovland and Weiss. Repeating a message might help the learning of
attitudes and emotional meanings for words discussed by Staats and Staats, since
a repeated association of the two stimuli is part of the process of conditioning.
Repetition might help the audience remember the message itself. Zielske (1959)
showed that advertising is quickly forgotten if not continuously exposed.
Krugman (1972) presents the intriguing argument that three exposures
might be all that are needed for a television advertisement to have its desired
effect. But he adds the important qualification that it might take 23 exposures
to get the three that produce the particular responses that are needed. Krugman
suggests that the first exposure to an ad is dominated by a cognitive What is
it? response. The second exposure is dominated by an evaluative What of it?
response. And the third exposure is a
reminder, but also the beginning of
disengagement. Krugman points out a fundamental difficulty, however, in that
people can screen out television ads by stopping at the What is it? response
without further involvement. Then, on perhaps the 23rd exposure, they might,
or might not, move on to the What of it?
response. Thus, Krugmans analysis
is stating that three exposures to an advertisement might be enough under ideal
circumstances, but that it might take a number of repetitions to achieve those three.
Too much repetition can also have some undesirable effects, however. In
one study, three groups of subjects were presented with one, three, or five
repetitions of a persuasive message (Cacioppo & Petty, 1979). The researchers
found that the message repetition led at first to increasing agreement with the
advocated position, but that after a certain point it led to decreasing agreement
with the advocated position. They found repetition led to decreasing, then
increasing, counterarguing against the message by the recipient of the message
And they found that any amount of repetition led to increasing topic-irrelevant
thinking. This kind of curvilinear relationship between repetition and commufli-
cation effects was also found in a study of political advertising. Becker an
d
Doolittle (1975) found that both liking for a candidate and seeking of informatio
n
about a candidate were highest with a moderate amount of repetition but
2 0 7
declined with high repetition. Another study found that humor ratings declined
steadily with repetition of ads (Gelb & Zinkham, 1985). A change in the creative
execution of the ad was found to boost the humor ratings back up.
THE NEWER PROCESS MODELS OF PERSUAS1ON
One of the newer developments in the field of persuasion has been the creation
of models of persuasion that emphasize persuasion as a process. Three major
process models of persuasion are McGuires information processing theory,
Andersons information integration theory, and Petty and Cacioppis elaboration
likelihood model.
These models share the following characteristics:
1. They present attitude change or persuasion as a process that takes
place through several steps and over time.
2. They involve an emphasis on cognition or information-processing.
3. They give a more active role to the receiver as an information-processing
agent than earlier conceptions of persuasion or attitude change.
McGuires Informati on-Processi ng Theory
McGuires theory suggests that attitude change involves six steps, with each step
being a necessary precedent for the next (McGuire, 1968). The steps are as follows:
1. The
2. The
3. The
b. The
5. The
6. The
persuasive message must be communicated.
receiver will attend to the message.
receiver will comprehend the message.
receiver yields to and is convinced by the arguments presented.
newly adopted position is retained.
desired behavior takes place.
McGuire notes that any independent variable in the communication situation
can have an effect on anY one or more of the six steps. A variable such as
intelligence, for instance, might lead to less yielding, because the more intelligent
Person is better able to detect flaws in an argument and is more willing to
aintain an opinion not held by others. But it might lead to more attention
because the more intelligent person has a greater interest in the outside world.
McGuire also points out that it is typical for independent variables to affect
ne step in a positive way and another step in a negative way. A fear appeal,
or nstance, might increase attention to the message but interfere with yielding.
ln a later article, McGuire (1976) presented eight steps in the information-
processing theory: exposure, perception, comprehension, agreement, retention,
etrieval~ decision making and action. It is obvious that this list of steps is built
p
on
the earlier six-step sequence but with some of the earlier steps recon-
ceptualized and some additional st~ps added.
F
In a still later article, McGuire (1989) presented twelve steps in the output
or dependent variable side of the persuasion process: 1. exposure to commu-
nication; 2. attending to it; 3. liking or becoming interested in it; 4. compre-
hending it (learning what); 5. skill acquisition (learning how); 6. yielding to it
(attitude change); 7. memory storage of content and/or agreement; 8. information
search and retrieval; 9. deciding on basis of retrieval; 10. behaving in accord with
decision; 11. reinforcement of desired acts; and 12. post-behaviod consolidating.
As initially presented, McGuires theory tended to deal almost exclusively
with the dependent variables in the persuasion process, splitting them intO more
and more categories until there were 12.
In other works, such as his 1989
chapter, McGuire also discusses the roles of independent variables.
McGuires information-processing theory gives us a good overview of the
attitude change process, reminding us that it involves a number of components.
Few previous theories have addressed all of these components, and few, if any,
attitude change studies looked at the effects of independent variables on all of
these steps. In fact, as McGuire pointed out,
most of the extensive attitude
change literature has probably focused on the step of yielding or agreement.
Finally, McGuires theory reminds us of the difficulty of attitude change. The
theory suggests that many independent variables tend to cancel themselves out
in their overall effects by having a positive effect on one step in the process of
attitude change but a negative effect on other steps in the process of attitude
change. Furthermore, we must face the fact that successful attitude change attempts
need to accomplish the desired effects specified by each of the various steps.
Anderson s Informati on Integrati on Theory
Information integration theory is a general theory developed by Norman Anderson ~
to explain how human beings bring together different pieces of information :
(Anderson, 1981). Anderson suggests that the process involves a kind of icogni-~
tive algebra and can be represented mathematically. The theory was originally ,
developed to explain how people can integrate a few basic personality traits to.
a
arrive at an overall impression of a person. But it was found to be applicable to
many areas of psychology, including attitude change. $
Information integration theory describes attitude change as a process
of integrating new information with old information (Anderson, 1971). The ,~
old information consists of the present attitude, and the new information COn
sists of the persuasive message. Each piece of information comes with Wo
attributesa scale value (represented by s) and a weight (represented by w).
The scale refers to the favorability rating assigned by the receiver to the piece
of information. The weight refers to the importance or relevance of the piece
of information.
The ratings (scale values and weights) for the various pieces of information
can be combined by the individual processing the information in several waYs
Anderson suggested that processing involves averaging the scale values an
d
weights. Another researcher, Martin Fishbein, proposed a similar model but
i
2 0 9
suggested that processing involves summing or adding the scale values and
weights (Fishbein, 1967). In practice, these procedures can lead to some rather
different outcomes.
As an example, let us take a presidential election campaign. The campaign
sttif is interested in changing the attitudes of the voters in a direction that is
favorable to their candidate. Imagine the following scale values are held by a
receiver about various positions taken by the presidential candidate. To simplify
the discussion, assume that all weights are equal.
Scal e
Candidate favors increased welfare
+2
Candidate opposes capital punishment
+3
Candidate favors equal rights for women +3
The candidates campaign is considering putting out a message that the
candidate is in favor of increased gun control.
Assume that for our hypothetical receiver, this piece of information would
have this scale value:
Candidate favors increased gun control
+2
Using the addition rule for information integration, the attitude would be
+8 before the new piece of information and +10 after.
Using the averaging rule for information integration, the attitude would be
8/3 or +2.67 befor the new piece of information and 10/4 or +2.5 after.
This example suggests that in the practical application of persuasion, it will
make a difference which rule for information integration is correct. There are
situations where the addition rule would suggest that attitude change would take
Place in a positive direction and the averaging rule would suggest that attitude
change would take place in a negative direction. Although the issue is not
completely resoIved, the research evidence at this point provides more support
for the averaging rule (McGuire, 1976).
The Elaboration Likelihood Model
ost People living in contemporary society are bombarded by mass media
essage% many of them attempting to persuade them of something. It is
ObviouslY impossible for a receiver to deal with all these messages at great length.
~pical~Y, we select some messages for detailed examination and deal with others
in a more peremptory fashion, if at all. A model of persuasion that acknowledges
hese two different means of processing messages is petty and Cacioppos
elaboration likelihood model (petty & Cacioppo, 1986).
he elaboration likelihood model suggests that there are are two routes to
attitude change_ the central route and the peripheral route. The central route
s nvO1ved when the receiver actively processes the information and is persuaded
I
I
I
by the rationality of the arguments. The peripheral route is involved when the
receiver does not expend the cognitive energy to evaluate the arguments and
process the information in the message, and is guided more by peripheral cues.
These cues can include source credibility, the style and format of the message,
the mood of the receiver, and so forth.
When the central route to persuasion is active, the receiver is said to be
involved in high elaboration. When the peripheral route to persuasion is active,
the receiver is said to be involved in low elaboration. Elaboration refers to
cognitive work involved in processing a persuasive message. Petty and Cacioppo
(1986, p. 7) state that elaboration refers to the extent
0
hich a
pe r s on
carefully thinks about issue-relevant information.
Elaboration involves attending
carefully to the appeal, attempting to access relevant information (from memory
or external sources), scrutinizing and making inferences about the arguments,
drawing conclusions about the merits of the arguments, and reaching an overall
evaluation of the recommended position.
Persuasion can take place under either a high degree of elaboration or a
low degree of elaboration, or at any point in between, but the model suggests
that the process of attitude change will be very different at different degrees of
elaboration. When persuasion occurs through the central route, it is usually
because high-quality arguments are being strongly presented. With the central
route, persuasion is most likely to occur when the receiver is led to have
predominantly favorable thoughts about the advocated position. So a key ques-
tion becomes, what factors lead the receiver to have either favorable or un-
favorable thoughts about the recommended position? TWO factors seem to be
important. The first is the agreement between the receivers initial position and
the recommended position. If an advocated position is one toward which a
receiver is already inclined, presumably the receiver will be favorably disposed
to the message. The second factor is the strength of the argument. The stronger
or more carefully defined the argument, the more likely it is that the receiver
will be favorably disposed to the message.
Under the peripheral route, persuasion will not depend on thoughtful
consideration of the message but on the receivers use of simple decision rules,
or heuristics. These principles are activated by cues in the persuasion situation.
The three major heuristics are credibility, liking, and consensus (0 Keefe, 1990,
pp. 186- 187). The credibility heuristic refers to the tendency for people to
believe sources that have credibility. The liking heuristic refers to the tendencY
for people to agree with people they like. The consensus heuristic refers to the
tendency for people to agree with positions that a lot of other people support
An example of persuasion through the peripheral route might involve
someone who wants to vote in a local or state election but doesnt want to take
the time to gather information about all the candidates and come to an origina
l
decision. Such a person might take the endorsements of candidates from an
editorial in the local newspaper and just vote straight down the list. Or a perso
n
might choose a straight party ticket, voting for all the Democrats or all the
Republicans. In these case% the decision rule being used is Vote the way this
trusted source recommends.
What kinds of factors determine which route will be taken, central or
peripheral? Two main factors influence the degree of elaboration by a receiver:
the receivers motivation to engage in elaboration and his or her ability to engage
in elaboration (O Keefe, 1990). Overall, the elaboration likelihood model helps
to account for a variety of attitude change studies by bringing them together in
the same model. We begin to see that rational models of persuasion, such as
information processing theory, and less rational models, such as the source
credibility model, can both be true. The model also helps us to reconcile
conflicting results of studies of the same factor in persuasion. It may be the case
that a high-credibility source will lead to attitude change in one situation but
not in another if attitude change is taking place through the peripheral route
in one situation and the central route in the other.
The elaboration likelihood model also gives a more active role in attitude
change to the receiver than we have seen in some earlier models of persuasion.
Under the elaboration likelihood model, the receiverto some degree, at least
decides whether to take the central route or the peripheral route. And certainly
if the central route is taken, the receiver becomes active in analyzing the message.
One problem with the elaboration likelihood model centers around the
concept of the quality of arguments. This concept is basic to the process of
persuasion through the central routehigher-quality arguments should lead to
more attitude change. To some extent, the theory defines strong or high quality
arguments as those that are persuasive. If researchers then conduct experiments
that show that strong arguments produce more attitude change than weak
arguments, we are not really finding out anything new. There seems to be some
circularity in the presentation of the theory at this point.
CONCLUSIONS
The field of attitude change research has expanded greatly since the early days
when the learning theory approach and the consistency theory approach were
dominant. Katzs functional approach was developed specifically to reconcile
hese two divergent views and fit them both into a larger picture. The Katz
a
PProach has drawn particular attention to the problem of changing attitudes
serving the ego-defensive functionnot an easy kind of attitude to change.
McGuire and Papageorgiss inoculation theory provides a nice kind of
balance to the many studies of attitude change. While others have been trying
0
discover the best means of persuading people, McGuire and Papageorgis have
bee
n investigating the best means of making people resistant to persuasion.
Festinger raised the important issue of whether attitude change produced
by Persuasive messages was accompanied by any real behavior change. About
he ame time researchers began a serious study of whether attitudes in general
as they were measured by researchers were useful in predicting behavior. One
of the beneficial results of all this questioning is that many attitude change
studies now incorporate behavioral measures as well as attitude measures.
Staats and Staatss theory that attitudes are learned through classical con-
ditioning suggests a strategy for use in advertising and other persuasive efforts.
Under this theory, the goal of a persuasive message is to cause the learning of
a positive or negative response to a word.
We also discussed four techniques that are common in persuasive commu-
nicationpictures, humor, sex, and repetition. The research on humor and sex
suggests that they should be used carefully, because they can be misunderstood
or can distract from the message, and there is little evidence that they actually
bring about attitude change. Repetition has its pros and cons and should also
be used carefully. It increases the chances of penetrating through audience
indifference or resistance, and it can lead to greater learningof a message, of
a relationship between a product name and positive associations, or of the
connection between a credible source and particular message. But it can also lead
to increased counterarguing and increased thinking about other irrelevant topics.
Finally, we discussed three newer models of persuasion that emphasize
persuasion as a process and the active role of the receiver. These models are
McGuires information processing theory, Andersons information integration
theory, and Petty and Cacioppis elaboration likelihood model. These models
remind us that persuasion is a complex endeavor. They also offer promise of
bringing togeth~r
into more unified
some of the diverse findings of research on attitude change
theories.
DISCUSSION
1.
2.
3.
4.
::
Much of the early attitude change research was based on the notion
that attitudes are learned responses. What are some other ways to
think of attitudes?
Research has shown that attempts to persuade that are based on fear
appeals can backfire. What are some alternatives to the use of fear
appeals in messages?
Evaluate Rogerss protection motivation theory. What are its advan-
tages over the notion of a curvilinear relationship between fear and
attitude change proposed by Janis?
What are the advantages of the new process models of persuasion
over the older Hovland approach to attitude change theory?
What are the disadvantages of the new process models of persuasion?
Newspaper reporters can be thought of in terms of low-credibility
or high-credibility sources. Why do you think the public has low
confidence in newspaper reporters, as public opinion poll results
sometimes suggest?
7.
8.
What are some of the likely consequences of the low confidence
ratings that the public assigns to newspaper reporters?
pick a worthwhile social issue, such as prevention of AIDS. Drawing
upon your knowledge of persuasion theory, design a communication
campaign to deal with the issue.
REFERENCES
Ajzen, I. (1971). Attitudinal vs. normative messages: An investigation of the differential
effects of persuasive communications on behavior. Sociometry 34: 263-280.
Ajzen, I., and M. Fishbein (1970). The prediction of behavior from attitudinal and
normative variables. J ournal of Experimental Social Psychology 6:466-487.
Alexander, M. W., and B. Judd (1978). Do nudes in ads enhance brand recall? J ournal of
Advertising Research 18 (no. 1): 47-50.
Allport, G. W. (1954). The historical background of modern social psychology. In
G. Lindzey (cd.), Handbook of Social Psychology, vol. 1, pp. 3-56. Reading, Mass.:
Addison-Wesley.
Anderson, N. H. (1971). Integration theory and attitude change. Psychological Review
78: 171-206.
Anderson, N. H. (1981). Integration theory applied to cognitive responses and attitudes.
In R. E. Petty, T. M. Ostrom, and T. C. Brock (eds.), Cognitive Responses in Persuasion,
pp. 361-397. Hillsdale, N.J.: Lawrence Erlbaum.
Becker, L, B., and J. C. Doolittle (1975). How repetition affects evaluations of and
information seeking about candidates. J ournalism Quarterly 52:611-617.
Brooker, G. W. (1981). A comparison of the persuasive effects of mild humor and mild
fear appeals. J ournal of Advertising 10 (no. 4): 29-40.
Brown, R. (1958). Words and Things. New York: Free Press.
Caballero, M. J., and P. J. Solomon (1984). Effects of model attractiveness on sales
response. J ournal of Advertising 13 (no. 1): 17-23.
Cacioppo, J. T., and R. E. Petty (1979). Effects of message repetition and position on
cognitive responses, recall, and persuasion. J ournal of Personality and f?OCid pSY-
Cbdogy 37: 97-109.
Chm w. (1933). The influence of oral propaganda material upon students attitudes.
Archi ves of Psychol ogy 150:1-43.
Chestnut, R. W., C. C. LaChance, and A. Lubitz (1977). The decorative female model:
Sexual stimuli and the recognition of advertisements. J ournal of Advertising 6 (no.
4): 11-14.
cook, T. D., and B. R. Flay (1978). The persistence of experimentally induced attitude
change. In L. Berkowitz (cd.), Advances i n Experi mental Social Psychology, vol. 11,
PP. 2-57. New York: Academic Press.
arling) A. L., and J. M. CivikIy (1984). The effect of teacher humor on classroom climate.
proceedings of tbe Tenth International Conference on Improving University Teaching,
PP. 798-8o6.
uncan~ C. R, and J. E. Nelson (1985). Effects of humor in a radio advertising experiment.
J ournal of Advertising 14 (no. 2): 33-40, 64.
agy A. H., and Chaiken S . (1993). Tbe Ryc bology of Atti tudes. Fort Worth, Tex.:
Harcourt Brace Jovanovich.
English, H. B., and A. C. English (1958). A Comprebensiue Dictionary of Psyd?ological
and Psychoanalytical Terms: A Guide to Usage. New York: Longmans, Green.
Evans, R. I., R. R. Rozelle, T. M. Lasater, T. M. Dembroski, and B. 1? Allen (1970). Fear arousal,
persuasion and actual versus implied behavioral change: New perspective utilizing a real.
life dental hygiene prograrn.J ournal of Personality and Social Psychology 16:220-227,
Festinger, L. (1964). Behavioral support for opinion change. Public Opinion Quarterly
28:404-417.
Festinger, L., and N. Maccoby (1964). On resistance to persuasive communications.
J ournal of Abnormal and Social Psychology 68:359-366.
Fishbein, M. (1967). A behavior theory approach to the relations between beliefs about
an object and the attitude toward the object. In M. Fishbein (cd.), Readings in Attitude
Tbeoty and Measurement, pp. 389-400. New York: Wiley.
Fishbein, M. (1973). Introduction: The prediction of behaviors from attitudinal variables.
In C. D. Mortensen and K. K. Sereno (eds.), Advances in Communication Research,
Pp. 3-31. New York: Harper & Row.
Fowles, J. (1982). Advertisings fifteen basic appeals. ETC. 39: 273-290.
Frankenberger, K. D., and A. S. SukhdiaL (1994). Segmenting teens for AIDS preventive
behaviors with implications for marketing communications. J ournal of Public Policy
& Marketing 13:133-150.
Gelb, B. D., and C. M. Pickett (1983). Attitude toward the ad: Links to humor and to
advertising effectiveness. J ournal of Advertising 12 (no. 2): 34-42.
Gelb, B. D., and G. M. Zinkham (1985). The effect of repetition on humor in a radio
advertising study. J ournal of Advertising 14 (no. 4): 13-20, 68.
Gillig, P. M., and A. G. Greenwald (1974). Is it time to lay the sIeeper effect to rest?
J ournal of Personality and Social Psychology 29:132-139.
Gould, S. J. (1994). Sexuality and ethics in advertising: A research agenda and policy
guideline perspective. J ournal of Advertising 23 (no. 3): 73-80.
Gruner, C. R. (1965). An experimental study of satire as persuasion. Speech Monograph
32: 149-153.
Gruner, C. R. (1967). Effect of humor on speaker ethos and audience information gain.
J ournal of Communication 17 (no. 3): 228-233.
Gruner, C. R. (1970). The effect of humor in dull and interesting informative speeches.
Central States Speech J ournal 21:160-166.
Gruner, C. R. (1972). Effects of including humorous material in a persuasive sermon.
Southern Speech Communication J ournal 38:188-196.
Hilgard, E. R., and G. H. Bower (1966). Theories of Learni ng. New York: Appleton-
Century-Crofts.
Hill, R. P. (1988). An exploration of the relationship between AIDS+ related anxiety and
the evaluation of condom advertisements. J ournal of Advertising 17; 35-42.
Hovland, C. I., I. L. Janis, and H. H. Kelley (1953). Communi cati on and Persuasion.
New Haven, Corm.: Yale University Press.
Hovland, C. 1., A. A. Lumsdaine, and F. D. Sheffield (1965). Experi ments on Mass
Communication. New York: Wiley.
Hovland, C. 1., and W. Weiss (1951). The influence of source credibility on communica
-
tion effectiveness. Public Opinion Quarterly 15:633-650.
Insko, C. A. (1967). Theories of Attitude Change. New York: Appleton-Century-Crofts.
Janis, I. (1967). Effects of fear arousal on attitude change: Recent developments in theorY
and experimental research. In L. Berkowitz (cd.), Advances i n Experi mental social
Psychology, vol. 3, pp. 166-224. New York: Academic Press.
I
Katz, D. (196o). The functional approach to the study of attitudes. Public
f)uarterly 24:163-204.
215
Opinion
Kelly, J. P., and P. J. Solomon (1975). Humor in television advertising. J ournal of
Advertising 4 (no. 3): 31-35.
fiech, D., R. S. Crutchfield, and E. L. Ballachey (1962). Individual in Society: A Textbook
of Social Psychology. New York: McGraw-Hill.
Krisher, H. P III) S. A. Darley) and J M. Darley O 973). Fear-provoking recommendations!
intentions to take preventive actions, and actual preventive actions. J ournal of
Personality and Social Psychology 26:301-308.
Krugman, H. E. (1972). Why three exposures may be enough. J ournal of Advertising
Research 12 (no. 6): 11-14.
LaPiere, R. T. (1934). Attitudes VS. actions. Social Forces 13:230-237.
Leventhal, H., and p. Niles (1964). A field experiment on fear arousal with data on the
validity of questionnaire measures. J ournal of Personality 32:459-479.
Lumsdaine, A., and I. Janis. (1953). Resistance to counterpropaganda Produced by one-
sided and two-sided propaganda presentations. Public Opinion Quarterly 17: 316.
Maccoby, N., and J. W. Farquhar (1975). Communicating for health: Unselling heart
disease. J ournal of Communication 25 (no. 3): 114-126.
Markiewicz, D. (1974). Effects of humor on persuasion. Sociometry 37:407-422.
McGuire, W. J. (1968). Personality and attitude change: An information-processing theory.
In A.C. Greenwald, T. C. Brock, and T. M. Ostrom (eds.), Psychological Foundations
of Attitudes, pp. 171-196. San Diego, Calif.: Academic Press.
McGuire, W. J. (1976). Some internal psychological factors influencing consumer choice.
J ournal of Consumer Research 2:302-319.
McGuire, W. J. (1989). Theoretical foundations of campaigns. In R.E. Rice and C. K.
Atkin (eds.), Publ i c Communi cati on Campai gns, 2nd ed., pp. 43-65. Newbury Park,
Calif.: Sage.
McGuire, W., and D. Papageorgis (1961). The relative efficacy of various types of prior
belief-defense in producing immunity against persuasion. J ournal of Abnormal and
Social Psychology 62:327-337.
Miniard, P. W., S. Bhatla, K. R. Lord, I? R. Dickson, and H. R. Unnava (1991). Picture-
based persuasion processes and the mediating role of involvement. J ournal of Con-
sumer Research 18: 92-107.
Murphy, G., L. B, Murphy, and T. M. Newcomb (1$)37). Experimental Social Psychology:
An Interpretation of Research upon tbe Socialization of tbe Indi vi dual , rev. ed. New
York: Harper and Brothers.
OKeefe, D. J. (199o). Persuasi on: Theory and Research. Newbury Park, Calif.: Sage.
Olson, J. M,, and M, p Zanna (1993). Attitudes and attitude change. Annual Review of
Psychology 44: 117-154.
aPageorgis, D., and W. McGuire (1961). The generality of immunity to persuasion
produced by pre-exPo5ure to weakened counterarguments. J ournal of Abnormal and
Social Psychology 62:475-481.
ettY, R. E., and J. T. Cacioppo (1986). Communication and Persuasion: Central and
Peripheral Routes to Atti tude Change. New York: Springer-Verlag.
Reardon, K. K. (1989). The potential role of persuasion in adolescent AIDS prevention.
ln R. E. Rice and C. K. Atkin (eds.), Public Communication Campaigns, 2nd cd., pp.
273-289. Newbury Park, Calif.: Sage.
gers> R. W. (1975). A protection motivation theory of fear appeals and attitude change.
J ournal of Psychology 91:93-114.
Ross, R. S. (1985). Understanding Persuasion: Foundations and PracticeS, 2nd ed.
Englewood Cliffs, N.J.: Prentice-Hall.
Sexton, D. E., and F! Haberman (1974). Women in magazine advertisements. J ournal of
Advertising Research 14 (no. 4): 41-46.
Sherif, M., and C. I. Hovland (1961). Social J udgment: Assimilation and Contrast Effects
in Communication and Attitude Change. New Haven, Corm.: Yale University Press.
Smith, G. H., and R. Engel (1968). Influence of a female model on perceived charac-
teristics of an automobile. Proceedings of tbe 76th Annual Convention of the
American Psychological Association 3:68-682.
Staats, A. W. (1968). learning, Language, and Cognition. New York: Holt, Rinehart,
and Winston.
Staats, C. K., and A. W. Staats (1957). Meaning established by classical conditioning.
J ournal of Experimental Psychology 54:74-80.
Staats, A. W., and C. K. Staats (1958). Attitudes established by classical conditioning.
J ournal of Abnormal and Social Psychology 57:37-40.
Staats, A. W., C. K. Staats, and H. L. Crawford (1962). First-order conditioning of meaning
and the parallel conditioning of a GSR. J ournal of General Psychology 67: 159-167.
Tanner, J. F., J. B. Hunt, and D. R. Eppright (1991). The protection motivation modeI: a
normative model of fear appeals. J ournal of Marketing 55: 36-45.
Taylor, P. M. (1974). An experimental study of humor and ethos. Soutberrz Speech
Communication J ournal 39:359-366.
Tripp, C., T. D. Jensen, and L. Carlson. (1994). The effects of multiple product endorse-
ments by celebrities on consumers attitudes and intentions. J ournal of Consumer
Research 20:535-547.
Whitehead, J. L. (1968). Factors of source credibility. Quarterly J ournal of Speech 54:
59-63.
Wise, G. L., A. L. King, and J. F! Merenski (1974). Reactions to sexy ads vary with age.
J ournal of Advertising Research 14 (no. 4): 11-16.
Zielske, H. A. (1959). The remembering and forgetting of advertising. J ournal of ~ar-
keting 23 (no. 3): 239-243.
chapter 10
Groups and Communication
The Dutch philosopher Baruch Spinoza pointed out 300 years ago that human
beings are social animals. His statement has been strongly reinforced by modern
psychology, which has shown that other people have a great influence on our
attitudes, our behavior, and even our perceptions.
The other people that influence us are in the groups that we belong to,
large or small, formal or informal. These groups can have a great effect on the
way we receive a mass communication message. This was hinted at in Chapter 4,
where we reported Cooper and Jahodas suggestion that group membership can
make attitudes of prejudice hard to change. Groups influence people S commu-
nication behavior in other ways, as we shall see.
The scientific study of the influence of groups on human behavior began
in the 193c)5, primarily with the work of social psychologist Muzafer Sherif.
Solomon Asch, another social psychologist, did some noteworthy work on group
pressures and conformity. ~other important name in the study of groups was
Kurt Lewin, the founder of the field known as group dynamics. The importance
of groups in th
e
formation of political attitudes and the making of voting
deci si ons was brought out in some classic electionstudies conducted in the
lq@s by sociologist Paul Lazarsfeld and his associates.
The following are three of the most important types of groups. A prthzary
group is a group (two or more persons) involving longstanding, intimate, face-
o-face association. Examples are a family, a work group, a team, a fraternity, or
a
military unit. A reference group is a group identified with and used as a
tandard of reference, but not necessarily belonged to. For instance, a student
wishing to belong to a certain fraternity might begin to dress like members and
adO
Pt their attitudes even though he is not a member. A casual group is a one-
ime
group of people who didnt know each other before they were brought
9*7

Hovland Propaganda Attitude Changes Sev9c

Uploaded by

Copyright:

Available Formats

Hovland Propaganda Attitude Changes Sev9c

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hovland Propaganda Attitude Changes Sev9c

Uploaded by

Copyright:

Available Formats

:@

You might also like