Much of the interest in empirical studies of judges lies in the comparison of actual to ideal behavior to reach conclusions about judicial competence. We may decompose any empirical study that attempts to address the competence of judges or the quality of judging into three basic components: (1) the specification of a normative benchmark; (2) the conversion of the benchmark into testable form and judicial behavior into measurable units; (3) the interpretation of the results of any comparison to draw appropriate conclusions about the descriptive-normative gap. This chapter considers complications at each stage in the comparison process, with illustrations from existing studies of judicial competence and studies from psychology that examine the gap between behavior and norms of rational judgment and decision making.