Message variation and source credibility in advertising
This thesis examines the empirical literature on the effects of source credibility and message variation on persuasive communication. Much of the credibility literature on the persuasive effects of source has been contradictory, for four main reasons. First, the dimensions used to define credibility are often unclear. Sometimes sources vary on more than one dimension. Alternatively, the independence of the factor the experiment is attempting to manipulate is often confounded with unrelated or extraneous dimensions. Second, given that certain persuasive effects are induced, the process by which this occurs is unclear - the literature is generally weak in dealing with mediating processes. Third, extensive research shows that source credibility operates most effectively when it is interacting with other variables, not in isolation. Fourth, many of the observed effects have been obtained from social persuasion experiments. These may not generalise to an advertising context. This thesis attempted to extend research in this area by addressing itself to some of the above problems. In particular, it examined the interactive effect of source credibility and an important variable, message variation, in inducing attitudinal and behavioural change in an experimental advertising situation. Findings from the experimental study are discussed in relation to (1) the difficulties involved in providing independent manipulations of source credibility; (2) findings from previous empirical research; and (3) the practical implications for advertising practice.