This chapter examines many different analyses and presents some of the principles on which readers shall base the judgments. It discusses technical terms and shows the kinds of statistical configurations readers have in mind when these terms are used. The typical question that the empirical investigator faces is how to test such propositions, that is how to collect and treat empirical data so as to make reasonable statements about causal hypotheses. The randomization allows the experimenter to deal with unwanted or extraneous causal variables and thus to satisfy the requirement for making a causal inference: that the association between the independent and dependent variables does not result from their having a common cause. In the empirical social sciences there is general agreement on the criteria for evaluating such statements. The central task, then, is to apply these criteria to the analysis of empirical data.