|
You might find it useful to read the early sections of Chapter 21 for this exercise.) The extramarital affairs data analyzed in Section 22.3.7 can be reinterpreted in the context of a binary choice model. The dependent variable in the analysis is a count of events. Using these data, first recode the dependent variable 0 for none and 1 for more than zero.Now, first using the binary probit estimator, fit a binary choice model using the same independent variables as in the example discussed in Section 22.3.7. Then using a semiparametric or nonparametric estimator, estimate the same binary choice model. A model for binary choice can be fit for at least two purposes, for estimation of interesting coefficients or for prediction of the dependent variable. Use your estimated models for these two purposes and compare the two models. |
| New search. (Also 1294 free access solutions) |