Statistics 550 Spring 2019 Exam 2 Take-home questions
1. Exercise 12.4 in the text.
2. The NFL data file has data on National
Football League teams and their offensive performance from last year.
There is a variable called 'Diff' in the data file. Create a new variable by
adding 28 to the Diff variable. Now fit a model to predict this variable from the
RushYds, PassYds, TakeInt, and GiveInt
variables. Check the studentized residuals, hat
values, and Cook's D statistic and report on which observations (if any)
exceeded the 'cutoff' values for each statistic.
3. Using the same data as in problem 2, obtain VIF and PCA information (no intercept) and interpret collinearity from this information. Also use these data to verify the identity on page 356 relating VIF values to PCA results for the 'PassYds' variable.
4. Use the Box-Cox procedure to find the optimal transformation for the model in Problem 2. Also recommend a transformation from a nearby better known value. Use the constructed BoxCox variable to create a partial regression plot. What does the plot tell you about the need for a transformation?
5. Name one concept or result from Chapter 10 that you enjoyed learning. Briefly
describe (in your own words) how this idea helps you better understand
regression analysis. Aim your explanation to be understandable to a student
taking an applied regression course like Stat 431.