The impact of deliberation on social and semantic networks: Citizens' mental models of bioethical issues in genetics using automated textual analysis

Young Min Baek, University of Pennsylvania


Recent democracy theorists unanimously agree with the importance of citizens' discursive activities about public issues, and argue that deliberation contributes to the formation of informed, balanced, and well-considered opinions of high quality. In this dissertation, discursive activities are conceptualized as the interactions between citizens in a society and messages in an issue discourse, and define the citizen-by-message interactions as two-mode network in formal network analysis. In deliberation, therefore, discourse ties citizens and citizens knit discourses at collective level. ^ In order to investigate effects of deliberation participation and qualitative features of the deliberation on discursive practices among citizens, this dissertation borrows theories of normative political philosophy, political sophistication in empirical political science, and mental model approach in cognitive psychology. Methodologically, this dissertation attempted Wordscores ®, an automated content analytic method, to identify main social themes in a public discourse of an issue, and applied formal network analytic methods (e.g., node centrality, network density, and MRQAP) measuring a citizen's and a theme's properties as nodes in a network within-group solidarity and interconnection relationships between groups whose issue positions are contrasting under multi-level modeling framework. ^ From the perspective of automated content analysis, results show that Wordscores® generated reliable and valid computerized coding comparable with manual coding. From the perspective of theoretical findings, results show that (1) deliberation participants form more direct ties with social themes and increase their ability to mediate two other citizens in the citizen-by-message interaction than non-participants, (2) after deliberation, public-oriented themes are better interconnected with other themes and become more central in semantic network than private-oriented themes, (3) people of the same issue position are more densely interconnected when the deliberation they experienced is more filled with social themes buttressing their issue positions, (4) deliberation partially mediates before-deliberation citizen-by-message network with after-deliberation citizen-by-message network. ^ Possible limitations are critically examined and both theoretical and methodological implications are discussed. ^ Keywords: deliberation, public opinion, computerized content analysis, network analysis, multi-level model ^

Subject Area

Speech Communication|Health Sciences, Medical Ethics|Information Science

Recommended Citation

Baek, Young Min, "The impact of deliberation on social and semantic networks: Citizens' mental models of bioethical issues in genetics using automated textual analysis" (2011). Dissertations available from ProQuest. AAI3475893.