Page 254 - The Handbook of Persuasion and Social Marketing

230 The Handbook of Persuasion and Social Marketing

Empirical generalizations may also help in preparing for the process
evaluation and summative research done to measure how a program is
performing and what has been ultimately accomplished (Balch & Sutton,
1997). Detecting small effects reliably requires larger samples (Cohen,
1992), which must be considered at the planning stage. Resources should
be allocated for reliable monitoring of a program’s progress and for evalu-
ating its ultimate success. Otherwise, inadequate research may lead to
missed opportunities for improvement, continuance of a failing program,
or failure to complete or expand on highly effective programs.

Conclusion

This review offers empirical generalizations about quantitative effect sizes
that have been derived from meta-analytic research in numerous disci-
plines. Dillard (1998, p. 267) suggested that “it will usually be better to
treat meta-analytic findings as guidance for future research rather than
problems that have been fully solved.” Effect sizes can vary over time, top-
ics, settings, and so on, and findings from smaller research streams may be
less reliable than those from larger bodies of literature. Therefore, such
generalizations can be both useful and dangerous. “To generalize is to be
an idiot,” according to William Blake (c. 1808), and “all generalizations are
false” (attributed to Alexander Chase, Albert Einstein, and Mark Twain).
Certainly, generalizing without evidence or assuming that the general ap-
plies to every particular is unwise. However, refusing to generalize is refus-
ing to learn. As Georg Wilhelm Friedrich Hegel stated, “An idea is always
a generalization, and generalization is a property of thinking. To generalize
means to think” (1833, p. 35). If used appropriately, the empirical gener-
alizations summarized here from more than 80 meta-analyses based on
thousands of effect sizes and millions of observations should have useful
implications for future social marketing efforts.

References

Albarracín, Dolores, Blair T. Johnson, Martin Fishbein, and Paige A. Muellerleile.
Theories of Reasoned Action and Planned Behavior as Models of Condom Use:
A Meta-Analysis. Psychological Bulletin 127, no. 1 (2001): 142–161.
Allen, Mike. Comparing the Persuasive Effectiveness of One-and Two-Sided
Messages, in Persuasion: Advances through Meta-Analysis, ed. Mike Allen and
Raymond W. Preiss (New York: Hampton Press, 1998): 87–98.
Andreasen, Alan R. Social Marketing: Its Definition and Domain. Journal of Public
Policy & Marketing 13, no. 1 (1994): 108–114.