History of the sleeper effect: Some logical pitfalls in accepting the null hypothesis.

Abstract

The history of research on the sleeper effect prior to 1978 can be divided into five stages: (a) initial discovery of the effect, (b) development of the underlying theory, (c) widespread acceptance of the effect and of the discounting cue explanation of it, (d) realization that past operational definitions of the effect were not isomorphic with the conceptual definition, and (e) repeated failures to demonstrate the effect once operational definitions were employed that corresponded to the conceptual definition (Gillig & Greenwald, 1974). These failures resulted in an invitation to accept the null hypothesis and to "lay the sleeper effect to rest." This article illustrates why it is not justifiable to accept the null hypothesis about the sleeper effect. We suggest that provisional acceptance of the null hypothesis depends on assuming that all the necessary theoretical, countervailing, statistical, and procedural conditions for an adequate test of the effect have been demonstrably met. We further suggest that none of the empirical studies prior to 1978 demonstrably succeeded in meeting these conditions. However, adequate tests following the guidelines we have described for provisionally accepting the null hypothesis have recently been conducted, and the effect has been repeatedly found. A deductive model of the logical factors that should guide provisional acceptance of the null hypothesis is contrasted with a current model that stresses induction and statistical power analyses.Most statistics texts assert that there is search findings to be clear about the criteria no formal basis for accepting the null hy-that help distinguish between a conclusion pothesis. The rationale for this assumption such as "We can be reasonably certain that is basically a restatement of the well-known no meaningful difference exists" and a conposition in philosophy that inductive knowl-elusion such as "Although no statistically edge is not logically possible. However, prac-reliable difference was found, we cannot be ticing scientists are often forced to act as at all certain whether a meaningful differthough the null hypothesis were true even ence exists." when they know that there is no compellingThe present article uses the history of epistemological basis for their actions. Given research on the sleeper effect to illustrate the apparent conflict between practice and some pitfalls in accepting the null hypothelogic, it is important in the analysis of re-sis. It also uses this research to describe and justify a model that might help in deciding when no-difference findings warrant the ten-' °a t t h U tative conclusion that no difference exists as C. an B rR ,a a the University of Waterloo, Waterloo, Canada, opposed to the conclusion that no decision The authors would like to thank Anthony G. Green-about differences is warranted. The sleeper wald and Daniel Romer for their comments on a effect is useful for this illustration, since a previous draft of this article^ recent da j m was made ^ ^ effect does Requests for reprints should be se...

Redressing the Sleeper Effect: Evidence for the Favorable Persuasive Impact of Discounting Information Over Time in a Contemporary Advertising Context

“…This aligns with a meta-analysis confirming the logical requirement of a strong discounting cue to find the sleeper effect (Cook et al 1979;Kumkale and Albarracín 2004), and counters more recent research using only weak negative information to enhance the persuasiveness of positive information (Ein-Gar et al 2012). Moreover, we can argue that dissociation occurred, because the increase in attitudes over time in the message + post-discounting group compared to the no message baseline implies that the discounting cue is less accessible than the message topic in the delayed post-test (Kumkale and Albarracín 2004).…”

“…It replicates earlier sleeper effect findings within a contemporary advertising context and with more stringent logical requirements (e.g., a strong discounting cue) (Cook et al 1979). The findings counter assertions from previous research and address criticisms of the effect by showing that an absolute sleeper effect can be found, that the effect occurs in a relevant context, and that it is more likely to occur when the discounting cue follows the message.…”

“…In response to contrary findings, Gruder et al (1978) and Cook, Gruder, Hennigan, and Flay (1979) detail the assumptions of the discounting cue/dissociation hypothesis, distinguish between the absolute and relative sleeper effect and outline the necessary conditions for the sleeper effect:…”

“…The absolute sleeper effect, in which attitude change increases over time in groups exposed to a discounting cue versus a no message control group. The relative sleeper effect, in which attitude change increases over time in the discounting cue group versus the message only group (Cook et al 1979;Pratkanis et al 1988). If there is no increase in persuasion compared to the no message group, then maybe attitudes are decaying to baseline attitudes rather than an actual change in attitudes.…”

The Sleeper Effect in Persuasion: A Meta-Analytic Review.

“…As such, this phenomenon has stimulated a large amount of research about the possibility of increased persuasion over time, as well as the potential decrease and lack of longitudinal change in persuasion (for reviews, see Cook, Gruder, Hennigan, & Flay, 1979;Eagly & Chaiken, 1993). Moreover, if one can understand the conditions that elicit increases, decreases, and stability in persuasion, one should be able to explain the Copyright 2004 by the American Psychological Association, Inc.…”