(a) Describe the intuition behind the maximum likelihood estimation technique used for limited dependent variable models.

(b) Why do we need to exercise caution when interpreting the coefficients of a probit or logit model?

(c) How can we measure whether a logit model that we have estimated fits the data well or not?

(d) What is the difference, in terms of the model setup, in binary choice versus multiple choice problems?