Description
In the current Dutch political landscape, with huge numbers of parties competing for voters' favor, there is a pressing social issue of how to inform voters about politics. This problem is critical, as low levels of political knowledge result in low voter turnout rates, affecting the functioning of democracy. Moreover, being informed also has ramifications for other aspects of good democratic citizenship, such as having more stable political attitudes across time, experiencing increased trust in politics, and respecting democratic values (e.g. Galston, 2001). A pressing societal question, therefore, is how to inform and engage citizens in politics. Not only is this question societally relevant, but it is also scientifically challenging as research shows that citizens are typically reluctant to invest significant effort into politics (e.g., Kamoen & Holleman 2017)In this research talk, we will present several studies about how Conversational Agent Voting Advice Applications (CAVAAs) may help to “spark democracy” and can assist voters in informing themselves about politics during elections. CAVAAs represent an evolution of conventional Voting Advice Applications like Stemwijzer and Kieskompas, wherein voters respond to attitude questions on political issues and receive tailored voting recommendations. The distinctive feature of CAVAAs lies in the incorporation of a Conversational Agent—a chatbot—through which users can access information about the political topics presented in the VAA statements while engaging with the tool. During the colloquium talk, we will showcase various studies on (CA)VAAs, including preliminary results from a recent experimental study involving a diverse population of Dutch voters conducted during the Dutch national elections (Tweede Kamerverkiezingen) on 22 November 2023.
Period | 14 Dec 2023 |
---|---|
Held at | Language, Communication and Cognition |
Degree of Recognition | Local |
Keywords
- conversational agents
- voting advice applications
- elections