Spatial Maps in Voting Advice Applications: The Case for Dynamic Scale Validation

Fernando Mendez
Jonathan Wheatley
Uwe Serdült
Acta Politica 50(2): 214–38.
Low‐dimensional spatial representations of political preferences are a widespread feature of voting advice applications (VAAs). Currently, VAA spatial maps tend to be defined on the basis of a priori reasoning. This article argues that VAA spatial maps should be empirically validated to safeguard fundamental psychometric properties – in particular, unidimensionality and reliability. We suggest dynamic scale validation as a pragmatic method for improving measurement quality in VAA spatial maps. The basic logic of dynamic scale validation is to exploit early user data as a benchmark against which ex‐ante defined maps can be evaluated. We draw on data from one of the most institutionalised VAA settings, Switzerland, to illustrate this dynamic approach to scale validation.
DOI: 10.1057/ap.2014.3
Germann, Micha, Fernando Mendez, Jonathan Wheatley, and Uwe Serdült. 2015. “Spatial Maps in Voting Advice Applications: The Case for Dynamic Scale Validation.” Acta Politica 50(2): 214–38.
@article{spatial-maps-in-vaa,
   author = {Germann, Micha and Mendez, Fernando and Wheatley, Jonathan and Serd{\"u}lt, Uwe},
   year = {2015},
   title = {{Spatial Maps in Voting Advice Applications: The Case for Dynamic Scale Validation}},
   pages = {214--238},
   volume = {50},
   number = {2},
   journal = {{Acta Politica}},
   abstract = {Low-dimensional spatial representations of political preferences are a widespread feature of voting advice applications (VAAs). Currently, VAA spatial maps tend to be defined on the basis of a priori reasoning. This article argues that VAA spatial maps should be empirically validated to safeguard fundamental psychometric properties -- in particular, unidimensionality and reliability. We suggest dynamic scale validation as a pragmatic method for improving measurement quality in VAA spatial maps. The basic logic of dynamic scale validation is to exploit early user data as a benchmark against which ex-ante defined maps can be evaluated. We draw on data from one of the most institutionalised VAA settings, Switzerland, to illustrate this dynamic approach to scale validation.},
   doi = {10.1057/ap.2014.3},
   url = {https://doi.org/10.1057/ap.2014.3},
   status = {personal}
}