← Publications

Journal article · P&G · 2026

Estimating the Recommendation Certainty in Candidate-Based Voting Advice Applications

Fynn Bachmann, Daan van der Weijden, Cristina Sarasua, Abraham Bernstein

Politics and Governance, 2026

Abstract

Voting advice applications typically require users to answer questionnaires before receiving party or candidate recommendations. As users answer more questions, the recommendations naturally become more accurate. However, when users do not complete the questionnaire, the certainty of these recommendations is unknown. In this work, we develop and present a measure to quantify this certainty by introducing an algorithm that estimates the candidate recommendation accuracy—the overlap between early and final recommendations—after each question.

PDFDOI
Two line charts against number of questions answered. Left: estimation error for several methods, with the Posterior sampling lines lowest and Linear and Historic highest. Right: grey per-voter accuracy curves with average, linear and FastCRA lines on top.
Figure 2 from the paper: (a) error of methods that estimate how accurate a voting advice application's early candidate recommendations are, by number of questions answered out of 75. Methods that simulate the voter's remaining answers (Posterior, One-shot) err less than a linearly growing progress bar (Linear) or the historic average (Historic). (b) The true accuracy of each voter's early recommendations (grey lines) varies widely from voter to voter.

Cite

@article{bachmann2026estimating,
  title={Estimating the Recommendation Certainty in Candidate-Based Voting Advice Applications},
  author={Bachmann, Fynn and van der Weijden, Daan and Sarasua, Cristina and Bernstein, Abraham},
  journal={Politics and Governance},
  volume={14},
  year={2026},
  doi={10.17645/pag.11256}
}