Graphing logistic regressions: the well-switching data described in Section 5.4 are in the folder arsenic. (a) Fit a logistic regression for the probability of switching using log (distance to nearest...

Graphing logistic regressions: the well-switching data described in Section 5.4 are in the folder arsenic.

(a) Fit a logistic regression for the probability of switching using log (distance to nearest safe well) as a predictor.

(b) Make a graph similar to Figure 5.9 displaying Pr(switch) as a function of distance to nearest safe well, along with the data.

(c) Make a residual plot and binned residual plot as in Figure 5.13.

(d) Compute the error rate of the fitted model and compare to the error rate of the null model.

(e) Create indicator variables corresponding to dist <><> 200. Fit a logistic regression for Pr(switch) using these indicators. With this new model, repeat the computations and graphs for part (a) of this exercise.