Application of minimum logit chi square estimate to a problem of Grizzle with a notation on the problem of 'no interaction'.

Logit analysis is applied to a problem of Grizzle [1961] who used maximum likelihood estimation and the standard Pearson x2. The estimates and the X2's by the two methods for four case problems of Grizzle are compared. Case 1 of Grizzle is equivalent to the 'no interaction' problem for the 2 X 2 X t table and it is noted that the logit analysis applies to that problem. Data of Norton [1945], of Bartlett [1935], and of Woolf [1955] are analyzed. A supplementary note considers the 'no interaction' problem for the 2 X 3 X t table and logit analysis is applied to data presented by Kastenbaum and Lamphiear [1959]. A x2 applied by Plackett [1962] and by Goodman [1963] to these data is identified as the logit x2 corresponding to the minimum logit x2 estimate. Some remarks are made on the relative efficiency of maximum likelihood and minimum logit x2 estimates.