Government documents and statistics provide another source of information and evidence, particularly for examining public policy, across and within countries and over time. This chapter reviews the different kinds of evidence which are commonly used to assess political science theories. It deals with the most obvious contrast – between qualitative and quantitative evidence. The chapter shows how demographic data developed out of qualitative descriptions. Running checks on the accuracy of the procedures used to produce them are useful as data collection and production proceed. The most direct checks on error are those directed at the final estimates themselves – for replicability, internal consistency and above all validity. Error also crops up when we relate evidence to theory within a statistical framework. The multi-measure multi-variable matrix gives powerful results because it combines and focuses several ways of assessing the error in estimates produced by measurement techniques.