Visualizing Opinion Space in Voting Advice Applications: A User Study

Keywords

VAAs, visualization, interaction, dimensionality reduction, social information, voting

Abstract

We investigate the impact of visualization methods and interaction mechanisms in Voting Advice Applications (VAAs) by comparing two designs: the traditional design in Kieskompas and an alternative design composed by a visualization -- metaphoric map -- based on dimensionality reduction of opinion data and an interaction mechanism allowing to adjust answers on the questions. In a user study with 382 participants from the Netherlands, we assessed their effects on vote choice, political knowledge, ease of use, and interpretation of opinion space. The results show no evidence that visualization and interaction mechanisms influence vote choice. The metaphoric map led to lower perceived and factual political knowledge and was harder to interpret, likely due to its higher visualization literacy demands. Notably, participants using metaphoric map offered more objective, non-ideological interpretations of their political position, suggesting potential to foster unbiased political comprehension. Supplementary materials are available at: https://osf.io/zvhsg/?view_only=ee2d7d315c254b358b9be077214cc18e.