Voting Aid Applications (VAA) provide information on political options during campaigns, and match users' policy preferences with the political offer of parties. In this way, VAAs may influence the process of electoral decision-making. But do VAAs actually influence citizens in forming a political preference, or changing their vote choice? We still do not know precisely what are the effects of VAAs on individual voting behavior, as most observational studies of VAAs run into selection and causal inference issues. In the current study, I use a randomized field experiment to evaluate the effect of a Voting Aid Application on users' voting preferences, during the 2014 Quebec election. It is hypothesized that, given Quebec's multidimensional political space and multiparty system, VAAs would be well adapted to assist less educated citizens in making an informed decision, while the more educated and politically sophisticated citizens would not be influenced by VAAs' information.
