11 HW FOR STAT601, Spring, 2008 1) For the colon cancer data on the website: a) Compute the correlation coefficient between meat eaten and colon cancer rates for the 23 given countries. b) Form the "natural" 95% CI for the true corr. coeff. for all countries. c) Using Fisher's transformation, form a better 95% C.I. 2) The fully observed data of Schmee and Hahn (1979) have 4 motorettes tested at each of 3 temperatures, 170, 190 and 220 degrees. The respective failure times were (1764, 2772, 3444, 3542) , (408, 408, 1344, 1440), and (408, 408, 504, 504). a) Give the linear fit to failure time as a function of temperature. Is the linear model useful in explaining failure time as a function of b) temperature? Is the linear fit in a) an adequate fit? Test at = .05. What is your conclusion on the adequacy of the linear model? 3) 11.94, answer in general and under what condition is it unbiased? 4) 12.1 5) Show that the least squares estimators in the GLM correspond with those in simple linear regression, i.e., when there is only one covariate, x. 6) Analyze the home energy data from the website. Answer the follo...