X = {
{1, 1250, 6},
{1, 1300, 7},
{1, 1350, 6},
{1, 1250, 7},
{1, 1300, 6},
{1, 1250, 8},
{1, 1300, 8},
{1, 1350, 7},
{1, 1350, 8},
{1, 1250, 6},
{1, 1300, 7},
{1, 1350, 6},
{1, 1250, 7},
{1, 1300, 6},
{1, 1250, 8},
{1, 1300, 8},
{1, 1350, 7},
{1, 1350, 8},
{1, 1250, 6},
{1, 1300, 7},
{1, 1350, 6},
{1, 1250, 7},
{1, 1300, 6},
{1, 1250, 8},
{1, 1300, 8},
{1, 1350, 7},
{1, 1350, 8}
}
y = {80, 95, 101, 85, 92, 87, 96, 106, 108, 80, 95, 101, 85, 92, 87, 96, 106, 108, 80, 95, 101, 85, 92, 87, 96, 106, 108},
except it has to be your own data.
2. Your design matrix X must be of dimensions n by (k+1), where n is no less than 20 and k+1 is no less than 3.
3. The due date is May 3, 2002, in class. Your project must be written in full English sentences and typed. Any relevant computer printouts must be attached. How the printout is used to obtain your conclusion must be shown clearly and in sufficient (but not excessive) detail.
4. You must state the objectives, indicate the source of your data, give the model statement, describe the relevant theoretical background, and indicate the kind of software you are using. Your conclusion must contain the following clearly indicated elements:
* estimates of all parameters of the model: the beta's and sigma-squared
* R-squared
* the Studentized residuals
* a general statement about the quality of the fit
5. A project done correctly and according to all of these rules will be given 10 extra credit points. However, failure to follow these rules, especially those stated above in parts 1--3, may result in summary rejection of your paper.