Date of Award

Summer 2012

Document Type


Degree Name

Master of Arts (MA)



First Advisor

Chattopadhyay, Sumana

Second Advisor

Griffin, Robert

Third Advisor

Ugland, Erik


This thesis presents a conceptual and methodological approach to researching preferences for political conversation. The thesis contends that although real-world political discussion is not deliberative, insofar as it fails to satisfy the rigorous requirements deliberative theorists have laid out, the lack of empirical evidence is not cause to reject deliberation as a viable political theory. To connect the theoretical and empirical, this thesis presents a “quasi-deliberation” framework. Quasi-deliberation, for the purpose of this thesis, is the state of political discourse shaped by the choices made when ideal deliberative qualities conflict in the real world. Quasi-deliberation suggests that the differences between the real world and the theoretical are described by preferences regarding different qualities of political conversation. These qualities, drawn from the deliberation literature (Moy & Gastil, 2006) are “dominance during political conversation,” “clarity” of opinion expression, use of “reason, logic, and evidence,” and “understanding of other conversants' views” (p. 448). The thesis tests the exploratory supposition that these choices are predictable outcomes of antecedent political characteristics of respondents, using a nationwide online survey instrument distributed to the non-random membership of a website. The characteristics measured in this thesis are culturally-informed worldview (Kahan, Slovic, Braman, & Gastil, 2006), attributes of personal discursive networks (Moy & Gastil, 2006), and political information efficacy (Kaid, McKinney, & Tedesco, 2007). These three sets of measures are used as independent variables to describe the unique, discursively relevant characteristics of the respondent. Each is then tested as a predictor of the relative priorities ascribed to each deliberative quality. Saaty’s (1980) Analytic Hierarchy Process is used to create the dependent priority ranking variables. Respondents provide a preference for each quality vis-à-vis each other, producing a preference matrix, from which a single priority vector is derived.

Included in

Communication Commons