Evaluating the cross-national comparability of survey measures of political interest using anchoring vignettes

Making cross-groups comparisons by using survey instruments has raised substantial scholarly concerns due to the potential risk of incomparability resulting from differential item functioning (DIF). However, not every survey item necessarily suffers from DIF. In this paper, we argue that, unlike many other survey items (e.g., political efficacy), the usual question used to measure political interest is likely to be largely immune to DIF. Our theoretical argument centers on the relative specificity of the item and a corresponding cultural homogeneity (at least in advanced democracies) in what it means to be politically interested or not. Utilizing the anchoring vignettes technique (King et al., 2004; King and Wand, 2007) in our original surveys in the UK, France, and the Netherlands, we demonstrate the size of DIF is small for the standard political interest question.