A New Typology of VAA Users

Voting advice applications (VAAs) have become an integral part of election campaigns in many democracies. However, surprisingly little is known about how VAAs are actually used. Existing typologies of VAA users have been based exclusively on users' self- reported behaviors and attitudes, but not on objective measures of actual behavior. In this article, we bridge this research gap by analyzing actual VAA behavior in a demographically representative sample of Danish voters (n = 1,011). First, we identify classes of VAA users based on objective behavioral metrics using latent class analysis (LCA). Second, we use pre- and post-VAA survey responses of the users, which include demographic and political attitudes, to predict their membership in the created latent classes.