Sockett et al. (1987) report data related to patterns of residual insulin secretion in children. A portion of the study was concerned with whether age can be used to predict the logarithm of C-peptide concentrations at diagnosis. The observed values are
Verify that r = 0.4 and that you reject H0: β1 = 0 with α = 0.05 and the R function hc4test.
For the data in Exercise 15, verify that a least squares regression line using only X values (age) less than 7 yields b1 = 0.247 and b0 = 3.51. Verify that when using only the X values greater than 7 you get b1 = 0.009 and b0 = 4.8. What does this suggest about using a linear rule for all of the data.