Year: 2018
Type: dissertation
Abstract: For the study of politics, the recent growth in the size, affordability, and availability of data and computational resources has been staggering. Large and novel data sources are now beginning to figure prominently in the work of some of the most impressive current research. This dissertation builds on these developments by demonstrating how the growing availability of large-scale survey data can open up new opportunities to answer questions that have, until now, remained difficult to address. To this end, this dissertation (1) proposes new methods for the analysis of large-scale survey data, and (2) seeks to answer theoretically important questions that can benefit from these methods and/or data. In the first paper, I seek to answer why some survey respondents answer truthfully to sensitive survey questions, while others do not. I propose a solution to this problem by developing a multivariate regression-based technique to model whether survey respondents provide one response to a sensitive item in a list experiment, but answer otherwise to a direct question. The method is applied to a large-scale list experiment to investigate misreporting about prejudice concerning women in politics. In the second paper, I examine how researchers can measure the ideology of ordinary citizens and political actors on the same scale. I propose a Bayesian Aldrich-McKelvey scaling method that both incorporates survey respondentsâ ideological self-placement into the scaling model and accounts for the fact that typical ideological placement scales are ordinal. The model is applied to large-scale data from recent Cooperative Congressional Election Studies to investigate whether Bernie Sanders was ideologically out of step with the national electorate during recent Presidential elections. In the final paper, I examine the effects of major terrorist attacks on public attitudes toward refugees. Using data from a large-scale natural experiment, the paper shows that the 2015 Islamic State terrorist attacks in Paris increased (1) anxiety over refugee resettlement; (2) perceptions of refugees as a security and cultural threat; and (3) opposition to resettlement. The findings are highly relevant to our understanding of public reactions to major terrorist attacks, and the responses of political entrepreneurs in their aftermath.
Source: TSpace (University of Toronto)
Author: Gregory Eady
Language: English
Subfield: Statistics and Probability
Field: Mathematics
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W3090739836