Question: Consider again the data in Table. Would it be meaningful to apply any linear smoother to these data? In general, when should we consider using a smoother?
Hastie and Tibshirani (1990) use part of a data set given in Sockett et al. (1987) to illustrate and compare various linear smoothers, but their analysis did not consider either robust smoothers, diagnostics, or multiple neighborhood sizes. We will consider each of these in the problems that follow, and the interested reader may wish to compare the results with the results given by Hastie and Tibshirani (1990).
The objective of the experimenters was to determine the dependence of the level of serum C-peptide on various other factors. We will use only age as an independent variable. The data on the 43 patients are given below.
