Electoral choices can be complex with voters being able to choose from numerous parties. In this project, we assume that voters manage this complexity by first ruling out certain parties based on a few heuristics. They then engage in more effortful processing to settle on a final choice, should multiple alternatives remain. I develop statistical tools for estimating two-stage choice models, look at individual level heterogeneity in choice processes, and assess implications for understanding party competition.
With the growing prominence of cross-national survey data, differential item functioning has become a central concern. Are cross-national differences indicative of trait differences across countries or do they merely reflect differences in response behaviour? If the latter is true, then how does this affect, among other things, causal inference and multilevel models? These are the central questions of the project.