Abstrakt
This dissertation contains empirical analyses where subjective data are used. It consists of two parts.
In part I analyses concerning anchoring vignette method are presented. Specifically, we want to validate this method and answer to the following questions: Do different vignettes within a health domain help to make self-assessments closer to objective reality? Do we get the same conclusion for different health domains? Are the main underlying assumptions of the method fulfilled in different domains? And how could we formally test them? Is the commonly used parametric model for anchoring vignettes correctly specified? Are the statistical assumptions for this model fulfilled? In part II we use stated preference method to analyze preferences of Dutch people for early, late and gradual retirement.