| Abstract: | Jackson and Jacobs (1983) have argued that many of the findings in communication research could be specific to the particular messages used in a given study. As a solution, Jackson and Jacobs propose generalizing to message populations by treating messages as a random variable. This article argues that the Jackson and Jacobs solution is inappropriate for complex messages and that meta-analytic techniques represent an alternative that permits the detection of results that are specific to particular messages, researchers, or paradigms. |
