Semantic Persuasion: Exploring Message Effects of Attribute Degree Centrality and Attribute Tie Strength on Decision Making
Three studies were conducted to explore how semantic network analysis can be utilized to define the quality of arguments and to study novel persuasion effects. Drawing on semantic network theory, the present research introduces attribute degree centrality and attribute tie strength as two new criteria of argument quality. In Study 1, analyses of semantic network data (i.e. degree centrality and tie strength scores) collected on over-the-ear headphone attributes from both the environment and cognitive representations are reported. Semantic network structure was found in the environment and participants were sensitive to this structure. Study 2 assessed the effects of an attribute’s weighted degree centrality on decision making. Attributes chosen from Study 1 varying in degree centrality were embedded in advertising arguments to test the novel hypothesis that the number and strength of attribute connections in a semantic network affect persuasion and decision making. As expected, the degree-centrality hypothesis was confirmed: Advertising arguments based on highly central attributes were more persuasive, and increased choice confidence, when compared with arguments based on lowly central attributes. Attributes high in centrality were also perceived as better arguments than attributes low in centrality which supports the claim that degree centrality provides a theory-based criterion of argument goodness. Centrality did not systematically affect the latency of choices or the perceived credibility of the message source. Study 3 replicated the findings of Study 2 and explored the main and interaction effects of degree centrality and tie strength by manipulating both attribute dimensions through a semantic network learning task. An interaction effect of degree centrality and tie strength on choice behavior and main effects of degree centrality on perceived argument quality, and perceived source credibility were revealed.
Reimer, Purdue University.
Off-Campus Purdue Users:
To access this dissertation, please log in to our