TRUNCATED LOGISTIC-REGRESSION

Truncated binary data occurs when a group of individuals, who each hav e a binary response, are observed only if one or more of the individua ls has a positive response. In this paper the group will be taken to b e a motor vehicle accident and the binary response taken to be surviva l or death. We compare two regression techniques that can be used for truncated binary data. The first procedure, conditional logistic regre ssion (Breslow and Day, 1980, Statistical Methods in Cancer Research. 1: The Analysis of Case-Control Studies. No. 32. Lyon: IARC) condition s on the actual number of deaths, and has been previously used for thi s type of data. The second procedure, truncated logistic regression, c onditions on there being at least one death. It is computationally sim pler than conditional logistic for groups of size greater than two and can be considerably more efficient. A major difference between the tw o methods is that only truncated logistic regression requires a knowle dge of group level covariates and allows estimation of group level eff ects.