Content-Type: text/html
Issue Involvement and Argument Strength
as Mediating Factors in the
Third-Person Effect
Abstract
This paper explores the possible relationship among issue involvement, argument
strength and the
Third-Person Effect. The study examines Petty and Cacioppo's Elaboration
Likelihood Model
and the Third-Person Effect hypothesis. A laboratory experiment was conducted
in which it was
hypothesized that perceived issue involvement in others would interact with
argument strength in
a written persuasive message to affect the nature of the Third-Person Effect.
That is, it was
hypothesized and found that people tend to believe that "others" will be more
affected than
themselves by a persuasive message that contains weak argumentation, but that
"others" will be
less affected than themselves by a persuasive message that contains strong
argumentation.
Issue Involvement and Argument Strength
as Mediating Factors in the
Third-Person Effect
A Paper Submitted to the
Communication Theory and Methodology Division of
The Association for Education in Journalism and Mass Communication
Washington, D.C.,Conference
August, 1995
H. Allen White
Assistant Dean
College of Fine Arts and Communication
Murray State University
P.O. Box 9
Murray, Ky 42071-0009
(502) 762-4669
Issue Involvement and Argument Strength
as Mediating Factors in the
Third-Person Effect
Introduction
Two areas of study concerning the formation of public opinion have received
a great deal
of attention during the past 15 years. Petty and Cacioppo's (1986) "Elaboration
Likelihood
Model" (ELM) posits that receivers of persuasive messages may be persuaded in
one of two
ways, either by message content or by non-content cues. Which of these two
routes predominates
depends on a variety of individual and situational factors.
Davison's (1983) description of the "Third-Person Effect," hypothesis, on
the other hand,
argues that receivers of persuasive messages tend to believe that message impact
will be greater
on "other" people than on themselves.
While the ELM models human communication, showing the process through which
message receivers go in dealing with persuasion, the Third-Person Effect
hypothesizes about what
is likely to occur. The present study is an attempt to meld the two areas of
research and offers
elements of the ELM as an explanation for the Third-Person Effect.
Literature Review
Elaboration Likelihood Model
Petty and Cacioppo's Elaboration Likelihood Model (1986) postulates that:
Issue Involvement, Argument Strength, and the Third-Person Effect
Although people want to hold
correct attitudes, the amount
and nature of issue relevant elaborations in which they are
willing or able to engage to evaluate a message vary with
individual and situational factors.
This suggests that people will think about a message (persuasive or otherwise)
as a function of
their motivation and of their ability to do so.
The ELM also states that people's mental dealings with messages occur in
one of two
ways -- through what the model calls the "central route" or through what it
calls the "peripheral
route."
Central route information processing requires the refining of information
in a manner such
that persuasive message arguments would be considered, weighed, and judged
"using thoughtful
consideration of the information at hand." The peripheral route, on the other
hand, requires less
cognitive effort in making decisions. That is, the peripheral route depends on
factors such as
sender characteristics and non-content message cues. These factors are called
"peripheral cues."
The ELM also postulates that:
As motivation and/or ability to process arguments is
decreased, peripheral cues become relatively more important
determinants of persuasion. Conversely, as argument scrutiny
is increased, peripheral cues become relatively less important
determinants of persuasion.
Thus, the ELM models persuasive communication as a process in which people
can deal
with information in one of two ways. Those who are able and who have some sort
of motivation
Issue Involvement, Argument Strength, and the Third-Person Effect
to do so will actually think about message content when making their decisions.
Those who lack
either ability or motivation will default to relying on some peripheral cue(s)
when making their
decisions. The ELM argues that strong content is more persuasive than weak
content, and
positive peripheral cues are more persuasive than negative peripheral cues.
Researchers have operationalized motivation in a number of ways. For
example, Petty and
Cacioppo (1979) and Petty, Cacioppo, and Goldman (1981) used issue involvement
as an
operational definition of motivation. They found that for subjects highly
involved with an issue at
hand, strong (logical) argumentation was more persuasive than weak (illogical)
argumentation.
For subjects not involved with the issue, they found that the effects of
argument strength were
much less pronounced. Petty and Cacioppo argued that these findings were
evidence that
involved (motivated) subjects engaged in central route processing while
uninvolved (unmotivated)
subjects did not.
Cacioppo and Petty's (1982) Need for Cognition (NFC) is another
manifestation of being
motivated to engage in central route processing. Operationalized as scores on a
"Need for
Cognition Scale," NFC is defined as a person's proclivity for and liking of
effortful thinking.
Cacioppo, Petty, and Morris (1983) argued for and found evidence to support a
relationship
between high levels of NFC and using central route processing. As with high
levels of issue
involvement, subjects high in NFC were found to react to argument strength in
ways that subjects
low in NFC did not. That is, high NFC subjects tended to be persuaded by strong
and dissuaded
by weak argumentation. Those low in NFC were not influenced by argument
strength.
Issue Involvement, Argument Strength, and the Third-Person Effect
In
short, the ELM posits that a number of variables interact determining whether a
message receiver will be persuaded. High levels of motivation interact with
argument strength
when the central route is used. Low levels of motivation interact with
peripheral cues when the
peripheral route is used.
Third-Person Effect
Davison (1983) described the Third-Person Effect as a hypothesis that
contends:
. . . individuals who are members of an audience that is
exposed to a persuasive communication (whether or not this
communication is intended to be persuasive) will expect the
communication to have a greater effect on others than on
themselves. And whether or not these individuals are among
the ostensible audience for the message, the impact that they
expect this communication to have on others may lead them
to take some action. Any effect that the communication
achieves may thus be due not to the reaction of the
ostensible audience but rather to the behavior of those who
anticipate, or think they perceive, some reaction on the part
of others.
As Perloff (1993) points out, a number of studies have found support for
the Third-Person
Effect hypothesis. In a study of political advertising during the 1988
presidential campaign,
Cohen and Davis (1991) reported that support was found for the Third-Person
Effect. Their
experimental subjects reported that negative political advertising had little
effect on their opinion,
but that the opinion of others would be affected.
Griswold (1992) found support for the Third-Person Effect in economic
speeches delivered by
Vice President Bush during the 1992 presidential campaign. Also conducting
studies Issue
Involvement, Argument Strength, and the Third-Person Effect
in the political context, Mutz (1989) and Rucinski and Salmon (1990) found
support for the
hypothesis. The Mutz study found that the Third-Person Effect existed for
highly charged issues
that were also deemed to be personally important to the subjects. Rucinski and
Salmon reported
that the Third-Person Effect was present when respondents considered political
polling
information.
Additionally, Cohen, et al, (1988) and Gunther (1991) found support for
the
Third-Person Effect when subjects considered defamatory news accounts. Finally,
Gunther
(1991), Perloff, et al, (1992), and Lasorsa (1989) found support for the
hypothesis in studies
involving entertainment programming and pornography.
In addition to these contextual variables, two other variables, which also
are related to the
ELM, have been found to influence the presence of the Third-Person Effect.
Perloff (1989)
found that ego-involvement magnified the Third-Person Effect in a study of
televised news
coverage of the 1982 war in Lebanon. (As noted above, involvement has been used
as an
operational definition for motivation in ELM studies.) This study used
pro-Israeli,
pro-Palestinian, and a non-partisan control group as subjects. The study found
that groups highly
involved with the issue being covered tended to believe neutral viewers would
acquire negative
attitudes about their side of the conflict and positive attitudes about the
other side. The author
offered an elaborated version of the interest-value hypothesis by Collins, et
al., (1988) as an
explanation:
Issue Involvement, Argument Strength, and the Third-Person Effect
According to this view,
individuals more closely attend to
schema-discrepant information if task conditions allow it. . . .
partisans might have directed more attention to the
schema-discrepant information presented in the news
videotape -- that is, events that cast their side in a negative
light -- than to schema-consistent information -- material that
portrayed their side positively. To the extent that this
discrepant information captured their attention, partisans
should have regarded it as more interesting and colorful than
schema-consistent information. According to the
interest-value hypothesis, they should have therefore believed
that these unfavorable depictions of their side would be more
persuasive (Perloff, 1989).
In another interesting study, Cohen, et al., (1988) used a source
characteristic (bias) and
combined it with the notion of social distance to test the Third-Person Effect.
These researchers
asked Stanford students to read a defamatory news account about a public figure.
These subjects
were then asked to judge the effects the defamatory content would have on
people's opinions of
the public figure. The "third persons" in this study were "other Stanford
students," "other
Californians," and "public opinion at large." The researchers predicted and
found that as the
social distance from the subjects' "self" increased, so did the Third-Person
Effect. That is, the
effect was greater for "other Californians" than for "other Stanford students,"
and the effect was
greatest of all for "public opinion at large." Further, there was a
Third-Person Effect for bias --
the source characteristic. That is, when the news account was attributed to a
source that was
either biased in favor of or against the public figure, the subjects said the
direction of the bias
would have little impact on their own opinions of the public figure. But, the
negative source bias
was judged to have a strong negative impact on the third persons' opinion of the
public figure
Issue Involvement, Argument Strength, and the Third-Person Effect
when compared with the positive source bias. The "bias" variable in this study
would be a
peripheral cue in ELM terms.
Hypotheses
While the ELM and Third-Person Effect hypothesis are different in many
respects, a
number of studies exploring the two theoretical issues share some common
attributes. Issue
involvement and source characteristics are two such attributes.
Perloff (1993) notes that the Third-Person Effect is likely to be manifest
under a number
of conditions. Among these conditions are messages that advocate behavior that
will not be
beneficial for the self and "statements that give rise to the perception that
'it is not smart to be
influenced by that message.'"
This becomes particularly clear when one considers the contextual issues
that have been
used in a number of Third-Person Effect studies. Among these contextual issues
are: negative
political ads (Cohen and Davis, 1991), defamatory news stories (Cohen, et. al.,
1988),
pornography (Gunther, 1991), and harmful media content (Tiedge, et. al., 1991).
It remains an empirical question as to whether a Third-Person
Effect would emerge for positive messages that make
recommendations that are perceived to benefit the self or for
messages that contain socially desirable recommendations
(Perloff, 1993).
This latter statement gives rise to a possible modification to
the Third-person Effect
hypothesis. Rather than hypothesizing that the "'other' will be judged to be
influenced more than
Issue Invo lvement, Argument Strength, and the Third-Person Effect
the
'self''," it may be profitable to hypothesize that the "'other' will be
influenced differently from
how the 'self' is influenced." After all, some persuasive messages are deemed
to be well thought-
out, logical discussions of some issue that may very well be beneficial to the
self. If this weren't
the case, how could anyone ever rationalize being persuaded to do anything --
anything from
buying a car to eating a healthful diet?
Gunther and Mundy (1993) suggest that the Third-Person Effect may be
attributed to a
"human tendency to see the world though optimistic or self-serving lenses."
These researchers
concluded that:
People are likely to consider themselves smarter and more
resistant to a message when they feel the topic is one that has
little benefit, or even potentially harmful consequences, for its
audience. If the potential benefit from a message is high,
however, then people consider themselves just as much
influenced as others. In some cases, they may anticipate even
more effect on themselves.
Gunther and Mundy's conclusions suggest that the Third-Person
Effect will manifest itself
as subjects' indicating something like: "I can tell the difference between good
and bad ideas, but
others cannot." Also, the Cohen, et. al., (1988) study involving social
distance suggests this
phenomenon will become greater as the "other" becomes less like the "self."
H1: A message evaluated as being weak in terms of logical
persuasive impact will be judged to have a greater persuasive
impact on others than the self.
H2: A message evaluated as being strong in terms of logical
persuasive impact will be judged to have a greater impact on
the self than on others.
Issue Involvement, Argument Strength, and the Third-Person
Effect
These first two hypotheses predict an interaction between person (self/others)
and
argument strength. Hypotheses 3 and 4 predict the nature of the interaction.
H3: Weak persuasive messages will be judged to have more
impact on others that are dissimilar to the self than on others
that are more similar to the self.
H4: Strong persuasive messages will be judged to have less
impact on others that are dissimilar to the self than on others
that are similar to the self.
In addition to these four hypotheses, the present study derived a
fifth hypothesis from
the logical underpinnings of the ELM and from observations made by Davison
(1983). Davison
noted that World War II propagandists were adept at sending messages to an
ostensible
audience with the intent of actually affecting the behavior of some third
person(s).
From the standpoint of a propagandist or other persuasive
communicator, on the other hand, the third persons are
those who are in some way concerned with the attitudes and
behavior of the ostensible audience. Indeed, the
propagandist may try to manipulate the behavior of these
third persons by apparently seeking to influence someone
else (Davison, 1983).
What Davison is indicating here is not only that people have a
tendency to judge the persuasive
impact a message is likely to have on third persons, but also an ability to
evaluate how involved
these third persons are with the issue at hand. (Davison used the term
"concerned.")
Interestingly, the ELM discusses issue involvement in ways that are pertinent
here. Issue
involvement and argument strength were found to interact in a number of ELM
studies.
Combining Davison's reasoning with ELM findings leads to positing that people
who project
Issue Invo lvement, Argument Strength, and the Third-Person Effect
issue
involvement on others will also project interactions similar to those found in
an ELM
context.
H5: Others perceived to be involved with an issue will be
judged to be more influenced by argument strength than
others perceived to be uninvolved with an issue.
Methods
One hundred eighteen subjects participated in a laboratory experiment
employing a
repeated measures design. The subjects were undergraduate students attending a
Southeastern
university. They were volunteers who received no inducements to participate.
The subjects were randomly assigned to two treatment groups. Each group
was
exposed to one of two versions of a written persuasive message advocating a
tuition increase at
the university as well as at other state-supported schools. One treatment
condition was
designed so that it contained "strong," logical reasons for the tuition
increase. The other
version contained "weak," illogical reasons for the proposed increase. The
messages were
identical in all respects except for the argument strength manipulation. The
strong and weak
argument versions of the message had been pretested on a different group of
undergraduate
students attending the same university. The pretest results were significant at
traditional levels
of statistical significance, with the "strong" arguments being judged as more
logical than the
"weak" arguments.
After exposure to the persuasive message, subjects were asked to evaluate
the message's
persuasive impact on:
Issue Involvement, Argument Strength, and the Third-Person Effect
Themselves
Other students attending their university
Other students attending other universities in their state
Other state residents "not involved" with higher education.
The items used for this task were seven-point Likert-type items.
Additionally, the order of the
Likert items was randomized with half the subjects responding to the items in
the order noted
above. The remaining half responded to the items in reverse order. This was
done to control
for any possible order effect.
Thus, the independent variables of interest were argument strength and a
person
(self/others) variable that was created by the repeated measures design. The
dependent
variables were responses to the four Likert-type items.
The General Linear Model in SAS was used to create analysis of variance
tests with a
repeated measured design.
Results
The data gathered supported each of the study's five hypotheses.
As with the pretest results of the argument strength variable, there was a
statistically
significant difference between the weak and strong version of the persuasive
message. Across
persons, the strong version was judged more persuasive than the weak version
(df=1, Type III
SS = 58.8, p<.01). A test of order effect was not significant. Thus, the
subjects responded to
the weak and strong versions of the persuasive message as had been predicted by
the pretest.
Although precautions were taken for a possible order effect, none was found.
Issue Involvement, Argument Strength, and the Third-Person Effect
Hypotheses 1 and 2 predicted an interaction between person (self/others)
and argument
strength. As Figure 1 indicates, these two hypotheses were supported (F[3,112]
= 2.82, p<.05).
Hypotheses 3 and 4 were enhancements of hypotheses 1 and 2. They predicted that
as others
become progressively less like the self, weak arguments will be judged to have
an increasing
persuasive impact, and strong arguments will be judged to have a decreasing
persuasive impact.
Again, as Figure 1 indicates, the nature of the interaction is as predicted.
____________________
Figure 1 About Here
____________________
Finally, hypothesis 5 predicted that others perceived to be involved with
college tuition
will be judged to be more influenced by argument strength than others perceived
to be
uninvolved with this issue. Again, this hypothesis was supported. The
univariate tests reported
by SAS showed that there was a statistically significant difference in the
persuasive impact that
subjects ascribed to the weak and strong versions on:
Themselves (df = 1, Type III SS = 36.38, F = 13.53, p< .001)
Other students attending their university (df = 1, Type III SS = 20.82, F =
9.79, p<.01)
Other students attending other universities in their state (df = 1, Type
III SS = 10.1, F =
6.21, p< .05)
Other state residents "not involved" with higher education (df = 1, Type
III SS = 2.45,
F = 1.35, Not Significant)
The final group of "other state residents not involved with higher education" is
the only one that
the questionnaire identified as "not involved" with the issue. The first three
groups were
Issue Involvement, Argument Strength, and the Third-Person Effect
identified as being involved by their status as tuition-paying students.
Conclusions
The results of this study indicate that the basic statement subjects are
making is
something like: "I can tell the difference between good and bad persuasion. As
people get less
and less like me, however, they are less and less able to make this
distinction." The Elaboration
Likelihood Model postulates that people who are motivated in some way will
consider argument
strength when deciding whether an advocated behavior is to be adopted. The
Third-Person
Effect hypothesis states the people make evaluations about "others" in
determining an estimate
of message impact on these others vis-a-vis the "self."
If, as the ELM states, people gauge their own motivation when they are
dealing with a
persuasive message, does it not follow that motivation would also be part of the
mental
calculations that are necessary to compare persuasive impact on "others" with
persuasive impact
on the "self?" Further, the ELM argues for a relationship between motivation
and argument
quality. If this reasoning can also be applied to the Third-Person Effect
hypothesis, then
argument quality should also be factored into one's mental calculations when
estimating
persuasive impact on others.
The results of this study support the above arguments. These data suggest
that people
do not always estimate persuasive impact to be greater on other people than on
themselves.
The data reported here suggest the phenomenon is more complicated. An
interpretation of
Issue Involvement, Argument Strength, and the Third-Person Effect
these data says people estimate that the persuasive impact on other people will
be different from
what it is on themselves. The nature of this different impact depends on the
interaction of a
number of variables.
The hypotheses supported in this study explored the variables of issue
involvement (an
operational definition of motivation) and argument strength. Future studies in
this area should
explore other variables in these terms. For example, individual differences
defined as Need for
Cognition or Need for Orientation (Weaver, 1980) may provide fertile ground for
future
studies.
For this study, the data reveal that it is a somewhat shallow
interpretation of the Third-
Person Effect to merely define it as people's tendency to believe others are
more vulnerable to
persuasive messages than they are themselves. A more accurate articulation of
the effect is to
say people believe others are affected by persuasive messages in ways that are
different from
their own. The nature of this difference depends on a nexus of interacting
variables.
Figure 1
Estimated Persuasive Impact by
Person by
Argument Strength
Estimated
Persuasive
Impact
Positive
Impact > 4
Neutral
Impact = 4
Negative
Impact < 4
Person
________________________________________________________________________________
_______
Repeated Measures ANOVA:
Source DF F p > F
Person 3, 112 2.89 0.038
Person by Argument Strength 3, 112 2.82 0.042
References
Cacioppo, J.T. and Petty, R.E. (1982). The Need for Cognition. Journal of
Personality and Social Psychology 42, 116-31.
Cohen, J., Mutz, D., Price, V., and Gunther, A. (1988). Perceived Impact on
Defamation:
An Experiment on Third-Person Effects. Public Opinion Quarterly 52 161-73.
Cohen, J. and Davis, R.G. (1991). Third-Person Effects and the Differential
Impact in
Negative Political Advertising. Journalism Quarterly 68, 680-88.
Collins, R.L., Taylor, S.E., Wood, J.V., and Tompson, S.C. (1988). The
Vividness
Effect: Elusive or Illusory. Journal of Experimental Social Psychology 24,
1-18.
Davison, W. P. (1983). The Third-Person Effect in Communication. Public
Opinion
Quarterly 47, 1-15.
Griswold, W.F. (1992). Third-Person Effects and Voting Intentions in a
Presidential
Primary Election, Paper presented at the annual meeting of the Association
for
Education in Journalism and Mass Communication, Montreal.
Gunther, A. C. (1991). What We Think Others Think: causes and Consequences in
the
Third-Person Effect. Communication Research 18, 355-72.
Gunther, A.C. (1991). The Third-Person Effect and Support for Restrictions on
Pornography. Paper presented at the annual meeting of the Association for
Education in Journalism and Mass Communication, Boston.
Gunther A.C. and Thornson, E. (1992). Perceived Persuasive Effects of Product
Commercials and Public Service Announcements. Communication Research 19,
574-96.
Gunther, A.C. and Mundy, P. (1993). Biased Optimism and the Third-Person
Effect.
Journalism Quarterly 70, 58-67.
Mutz, D.C. (1989). The Influence of Perceptions of Media Influence: Third
Person Effects and
Public Expression of Opinions. International Journal of Public Opinion
Research 1,
3-23.
Lasorsa, D.L. (1989). Real and Perceived Effects of "Amerika." Journalism
Quarterly 66,
373-8, 529.
Perloff, R.M. (1993). Third-Person Effect Research 1983-1992: A Review and
Synthesis. International Journal of Public Opinion Research 5, 167-84.
Perloff, R.M., Neuendorf, D., Giles, D., Chang, T.K., and Jeffres, L.W. (1992).
Perceptions
of "Amerika." Mass Comm Review 19, 42-48.
Perloff, R.M. (1989). Ego-Involvement and the Third-Person Effect of Televised
News Coverage. Communication Research 16, 236-62.
Petty, R.E., Cacioppo, J.T., and Goldman, R. (1981). Personal Involvement as a
Determinant
of Argument-Based Persuasion. Journal of Personality and Social Psychology
41,
847-55.
Petty, R.E., Cacioppo, J.T., and Morris, K.J. (1983). Effects of need for
Cognition on
Message Evaluation, Recall, and Persuasion. Journal of Personality and
Social Psychology 45, 805-18.
Petty, R.E. and Cacioppo, J.T. (1986). The Elaboration Likelihood Model of
Persuasion.
In L. Berkowitz (Ed.), Advances in Experimental Social Psychology 19,
124-206.
Petty, R. E. and Cacioppo, J.T. (1979). Issue Involvement Can Increase or
Decrease
Persuasion by Enhancing Message Relevant Cognition Responses.
Journal of Personality and Social Psychology 37, 1915-26.
Rucinski, D. and Salmon, C.T. (1990). The "Other" as the Vulnerable Voter:
A Study of the Third-Person Effect in the 1988 U.S. Presidential Campaign.
International Journal of Public Opinion Research 2, 345-68.
Tiedge, J.T., Silverblatt, A., Havice, M.J., and Rosenfeld, R. (1991).
Discrepancy
Between Perceived First-Person and Perceived Third-Person Mass Media
Effects. Journalism Quarterly 68 141-54.
Weaver, D.H. (1980). Audience Need for Orientation and Media Effects.
Communication Research 7, 361-76