ABSTRACT
We shall investigate the data to try to determine which, if any, of the explanatory variables are predictive of survival. (The analysis presented in this chapter will be relatively straightforward; a far more comprehensive analysis of the Titanic survivor data is given in Harrell, 2001.)
In essence, this is the same question that is addressed by the multiple regression model described in Chapter 4. Consequently, readers might ask: What is different here? Why not simply apply multiple regression to the Titanic data directly? There are two main reasons why this would not be appropriate:
The response variable in this case is binary rather than continuous. Assuming the usual multiple regression model for the probability of surviving could lead to predicted values of the probability outside the interval (0, 1).
