This is the course home page. I will post general announcements and record changes I make to the website here. Use the dates as guidelines of what's been added since the last time you visited. Active links on the calendar indicate that lecture summaries have been posted for those dates.
March 31, 2007
- Additional hints have been added to the first problem in Assignment 9.
March 30, 2007
- Assignment 9 is posted and is due April 13 (2 weeks!)
March 28, 2007
- I made a small change in the boot.confi function that appears near the end of the online notes for lecture 28. The previous version had ci as an argument but then proceeded to assume ci had value "basic" at a later point. This has now been fixed.
March 27, 2007
- Notes for the lab session lecture 28 (March 23) are posted.
March 23, 2007
- Assignment 8 is posted and is due March 30.
- Solutions to the midterm are posted.
March 7, 2007
- Hint on Midterm—Problem 2
- For question 2, try removing the intercept from your model specification when you fit the model. You should then get all of Manly's estimates. Think about the effect that including or not including an intercept in the model has on the parameter estimates you get. If you understand what's going on you should be able to take the parameter estimates from one specification and use them to calculate the estimates obtained from the other specification.
- For question 3 the hint I gave in class is appropriate.
March 6, 2007
- Bonus Question on Midterm
- There's a question I forgot to put on the midterm so I will add it now as an extra credit question. In question 3 of Problem 1, when you calculate the mean household size and land ownership by stratum you should find that one of the subvillages, Madukani, yields zero as the estimate for the standard error of the mean in both cases. Why?
- Hint on Midterm—Problem 1
- To have a chance at getting questions 2–4 correct you have to get question 1 correct. Think hard about how the sampling is being done. You haven't had a design exactly like this one but you've had every component of this design, just not together before.
- You will need to combine the three "amy" files two at a time. This involves two simple function calls in R using the same function each time. See Lecture 4.
- The amy3 file contributes an important ingredient to the design. If you did the previous step correctly you won't have any "rep"ping to do when you make use of it.
March 4, 2007
- Solutions to Assignment 6 are posted.
March 3, 2007
- Notes for lecture 20 (Feb 26), lecture 21 (Feb 28), and the lab session, lecture 22 (Mar 2), are posted.
March 2, 2007
- The midterm is posted and is due Friday, March 9.
March 1, 2007
- Solutions to Assignments 4 and 5 are posted.
- For Problem 1 on Assignment 7 both the Pearson and G2 statistics should be based on the sum of ten terms.
