Value-affirmative and value-protective processing of alcohol education messages that include statistical evidence or anecdotes

This study of persuasion processes in a value-relevant context tests effects of the presence or absence of statistical evidence and the presence or absence of anecdotal evidence, crossed across three base messages regarding different alcohol use issues. Results suggest that a variant of central processing as described by Petty and Cacioppo (1986) was used: Involvement predicted greater message-relevant responses only when the message was congruent with recipients' own values regarding alcohol use. Among recipients for whom the message was value-congruent, messages with statistical evidence were rated more persuasive, more believable, and better written; anecdotal evidence had no effect. Among recipients for whom the message was value-discrepant, messages with anecdotal evidence were rated more persuasive, more believable, and (marginally) better written, and statistical evidence had no effect. Path analyses also suggest that peripheral-processing strategies are employed when the message is value-discrepant, and central-processing strategies are used when the message is value-congruent.