Page 180 - Media Effects Advances in Theory and Research

7. MASS MEDIA ATTITUDE CHANGE 169

ing, Staats and Staats, 1958). Or, the source of a message can trigger a rela-
tively simple inference or heuristic such as “experts are correct” (Chaiken
1987) that a person can use to judge the message. Similarly, the responses
of other people who are exposed to the message can serve as a validity cue
(e.g., “if so many agree, it must be true”; Axsom, Yates, & Chaiken, 1987).
In the first half of the past century, the Institute for Propaganda Analysis,
in a report on propaganda techniques, listed a number of “tricks” that
speakers of the time used to persuade their audiences that relied on
peripheral cues (e.g., the “bandwagon” effect was giving the sense that
most other people already supported the speaker; see Lee & Lee, 1939).
We do not mean to suggest that peripheral approaches are necessarily
ineffective. In fact, they can be quite powerful in the short term. The prob-
lem is that over time, moods dissipate, peoples’ feelings about sources can
change, and the cues can become dissociated from the message. These fac-
tors would then undermine the basis of the attitude. Laboratory research
has shown that attitude changes based on peripheral cues tend to be less
accessible, enduring, and resistant to subsequent attacking messages than
attitudes based on careful processing of message arguments (see Petty et
al., 1995). In sum, attitudes changed via the central route tend to be based
on active thought processes resulting in a well-integrated cognitive struc-
ture, but attitudes changed via the peripheral route are based on more
passive acceptance or rejection of simple cues and have a less well articu-
lated foundation. 4
The tendency for simple cue processes to dissipate over time along with
the tendency for argument-based persuasion to persist can lead to inter-
esting effects. For example, one such phenomena is the often cited but
infrequently found (Gillig & Greenwald, 1978) “sleeper effect” (Gruder,
Cook, Hennigan, Flay, Alessis, & Halamaj, 1978; Hovland, Lumsdaine, &
Sheffield, 1949; Peterson & Thurstone, 1933). The sleeper effect can occur
when a persuasive message is followed by a discounting cue (e.g., you
learn that some information was reported in the National Enquirer after
exposure). The effect is that although the discounting cue suppresses atti-
tude change initially, over time the message can increase in effectiveness—
opposite to the typical decay pattern found. The ELM predicts that such
an effect should be most likely to occur under conditions in which the ini-
tial message is very strong, processed carefully, and then discounted. If
the message was processed carefully and a simple cue follows message

4 For expository purposes, we have emphasized the distinction between the central and the
peripheral routes to persuasion. That is, we have focused on the prototypical processes at the
endpoints of the elaboration likelihood continuum. In most persuasion situations (which fall
somewhere along this continuum), some combination of central and peripheral processes
are likely to have an impact on attitudes.