Testing Key Hypotheses in Panel Analysis

It remains true, as was observed a few years ago, that "attention to panel analysis has continuously increased; the sixteenfold table has especially intrigued various writers because of the surprising richness of data it provides" (Lazarsfeld and others, 1972, p. 329). Cases in point are the recent contributions of Davis (1978), Goodman (1973), Kessler (1977), Lazarsfeld (1972), and Singer and Spilerman (1977). Each of these papers is concerned with the general strategy of panel analysis. This chapter has more modest objectives: it shows how to test two particular hypotheses that are often of interest in studying a 24 table. Illustrative data are presented in Table 1.