Directions: Please answer all questions.  All point values are one unless marked.  Partial credit is awarded on a case by case basis.

Your task is to create a multiple linear regression study.  I am not providing you with data, it is your job to determine the variables and collect data. Please answer all of the questions below.

• Create a problem/study. Explain why the you have selected the dependent variable to study. (2 pts)

• Please find at least 4 total variables and one binary variable to use in your study. Describe the variables. (3 pts)

• Describe what you expect the relationship between the dependent and each independent variable. Specifically, explain what sign you expect each variable to be. (2 pts)

• Calculate the multiple linear regression using all variables.

• Interpret the adjusted R-squared. What do you conclude?

• Interpret the coefficient of the intercept.

• Choose a variable and create the 99^th percent confidence interval.

• Interpret the F-test. What do you conclude?

• Calculate DFITTS. Are there outliers?

• Calculate COOKSD. Are there outliers?

• Interpret the QQ Plot. What do you conclude?

• Is there evidence of heteroskedasticty from the residual plot?

• Calculate the Breusch pagan test. What do you conclude?

• Calculate the VIFs. Is there evidence of multicollinearity? How do you know?

• Run forward stepwise regression using AIC criteria. (2 pts)