Using Twitter to Improve the Design of Voting Advice Applications

During the design phase of a VAA, a task of top importance is the selection of the statements which will be included in the final version of the tool. The selected topics should reflect the major dimensions of electoral competition, and they should have adequate discriminating power to facilitate the discrimination between the political parties. There are many ways to create the list of statements: some VAAs cooperate with the political parties, other research teams ask a random sample of voters to propose issues and other use mixed methods. In this paper, I explore the possibility of using the analysis of tweets made by the electorate as an additional measure that can facilitate the VAA designers on their effort to determine the importance of the issues they plan to include in their applications. Finally, I explore the possibility of using the analysis of tweets made by the candidates to estimate the position of their parties on specific political issues.