Rethinking the 16-fold table problem

This paper presents a reconceptualization of Lazarsfeld's 16-fold table method of evaluating rival causal hypotheses over time. In the first section we review the problem and the various methods proposed to deal with it. In the second section we discuss the conventional methods for analyzing the 16-fold table and present a series of descriptive statistics which are computationally and conceptually more appropriate to the rigorous assessment of rival causal hypotheses. Finally, in the last section, we present significance tests for each of these descriptive measures. (March 1977)