Return to the anemia data set. (a) Suppose that the investigator wants the time scale to be weeks rather than days. Re-fit the main effects model (i.e., AGE, CSP.MTX, LAF) and compare your results with those previously obtained. (b) Fit a main effects model, with AGE, CSP.MTX, and LAF as covariates to the censoring hazard. Based on the results from this model, do you have evidence that the noninformative censoring assumption is violated? Briefly defend your response