Research


Dissertation

Download PDF

Scholars have researched how voters make decisions for well over a half a century, but these studies are limited in what they are able to say about how voters make decisions because they have focused on the choice rather than the process. Most of these studies have focused on the choice that voters reach or the way their memories are structured, overlooking the importance of the search and acquisition of information. Specifically, scholars in political science have paid little attention to how contextual variations in the information environment affect how voters make decisions.

This dissertation investigates how changes in context affect how voters search for information. I explore three specific contexts: the number of offices on the ballot, the availability of partisan information about the candidates, and the presence or absence of campaign dialogue between two candidates. Indeed, one of the prominent features of American elections is the variation in the number of elections across jurisdictions, the availability of partisan information about candidates, and the amount of campaign dialogue between candidates--the three contexts that I examine in this study.

I conduct three experiments that manipulate each of these contexts, using a dynamic information board that simulates the campaign environment and process tracing methods to track the information subjects chose to view and in what order they chose to view it. Results indicate that context shapes the way voters search for and acquire information. When faced with long ballots, subjects examined information less closely, they compared more information between candidates, and they searched for information less systematically. When subjects were unable to use the partisan cue, they compared less information between candidates and searched for information less systematically. Finally, when there was no dialogue between candidates, subjects searched for less information and had a less systematic search for it. These findings suggest that there are better ways to design elections.