Comments about Ch. 2 of Applied Logistic Regression, 2nd Ed.

1. (p. 31) Note that (2.1) corresponds to a strong assumption which may not be true --- the logit need not be a linear function of the predictor variables.
2. (p. 34) In general, one needs the expectation of the 2nd-order derivatives to get an information matrix, but in this case the 2nd-order derivatives do not involve any random variables.