1. Given the printout in Exercise 16.15, determine the approximate

a. 90% confidence interval for the mean rating on test 4 for units that have been rated at 300, 500, and 400 on the first three tests.

b. 90% prediction interval for the rating a unit will get on test 4 if its ratings are 300, 500, and 400 on the first three tests.

2. The regression equation yˆ 5 5.0 1 1.0x1 1 2.5x2 has been fitted to 20 data points. The means of x1 and x2 are 25 and 40, respectively. The sum of the squared dif- ferences between observed and predicted values of y has been calculated as SSE 5 173.5, and the sum of squared differences between y values and the mean of y is SST 5 520.8. Determine the following:

a. The mean of the y values in the data.

b. The multiple standard error of estimate.

c. The approximate 95% confidence interval for the mean of y whenever x1 5 20 and x2 5 30.

d. The approximate 95% prediction interval for an individual y value whenever x1  5 20 and x2  5 30. ( DATA SET ) Note: Exercises 16.18-16.22 require a computer and statistical software. If possible with your statistical software, determine the exact, rather than the approximate, intervals.