Linguistic markers of association as persuasive devices in mediated appeals.

The framework developed and tested in this study argues that the success of mediated persuasive appeals can be partially explained by the interaction among linguistic and extra-linguistic variables. It was predicted that sources would be most persuasive when the topic was of little importance and the source was liked by the viewer and that sources using intense language would be evaluated as believing the arguments presented more than sources using less intense language. These predictions were supported. The study also supported a predicted interaction between familiarity and liking such that familiarity works to the advantage of liked sources and to the disadvantage of disliked sources. It was also posited that intensity would have a differentially effective role for liked and disliked sources such that it works to the advantage of liked sources and to the disadvantage of disliked sources. This hypothesis was solely supported in proattitudinal appeals when intensity was operationalized using structural criteria and in counterattitudinal appeals when intensity was operationalized as perceived by the viewer. Finally, a 3-way interaction was predicted suggesting that familiarity and intensity would combine such that high familiarity results in increases in attitude change with the use of language that is more intense than expected for liked sources and decreases in attitude change for disliked sources. This hypothesis was not supported. In fact, both highly familiar sources benefited from use of language that was less intense than expected