Announcements

Sunday, August 19

I've added some comments about Ch. 11 (for August 24 meeting). I'll add comments about Ch. 12 (also to be discussed on 8/24) later.

Monday, August 6

I've added some comments about Ch. 10 (for August 17 meeting --- recall that we aren't going to meet on August 10).

Sunday, July 29

I've added some comments about Ch. 9 (for August 3 meeting), and I've also added a few more comments about previous chapters that we've discussed. I've added links to related web sites of J. Gentle, Y. Li, and A. Keesee, and these can be found on the home page for this web site. Finally, and this is important, there will be no meeting on August 10 --- I've updated the schedule to indicate new plan for finishing book during August.

Tuesday, July 24

I've added some more comments about Ch. 8 (for July 27 meeting), and I've also updated the notes in some earlier sections. To make it easier to spot the changes in the earlier parts, I've put the changes in different colors (as explained at the top of the Comments page).

Monday, July 23

I've added some more comments about Ch. 8 (for July 27 meeting), but still want to add a few more.

Sunday, July 22

I've added some comments about Ch. 8 (for July 27 meeting), but need to add more (and need to check over what I have typed so far). I plan to add more comments either Monday or Tuesday. Also I hope to read over some of the articles that Wilcox refers to in Ch. 8.

Tuesday, July 17

I've added some comments about the material to be discussed this coming Friday. (Note: I don't have a lot of comments this time, so please read the material in the book and let me know if you have any concerns or comments about it. Also, I didn't find any mistakes, so you might be on the lookout for any that I missed.) Since there may be less to discuss from the book this week, perhaps we can revisit the Monte Carlo project that some of us worked on during the spring semester. Yaru Li's web page shows some results. (Let's set up the projector this week and try to determine what modifications may need to be done and/or determine what results to obtain next.) Yaru gives us some estimated powers, and for the nonnormal distribution considered, the alternative procedures exhibit greater power than the OLS t test. But since this is also the case with some of the alternative tests when the error term distribution is normal, one wonders about the accuracy of the alternative procedures (perhaps they routinely reject too much when the null hypothesis is true). I think a good next step would be to obtain some results on rejection rates under the null hypothesis (and so we need to set the slope to 0). Finally, I'll mention that a possibility for a future seminar is The Elements of Statistical Learning: Data Mining, Inference, and Prediction, by Hastie, Tibshirani, and Friedman. (This book isn't available yet --- August 2001 is the target date given by Springer.)

Sunday, July 8

I've added some comments about the material to be discussed this coming Friday.

Monday, July 2

I've added some comments about the material to be discussed this coming Friday. I challenge you to confirm the values I report for the expected percentage of outliers for the lognormal distribution (and also confirm the value I give for the kurtosis).

Monday, June 25

Since some people may want to print the comments prior to the meeting, I changed the text in that section to black (based on Arndt's suggestion). To partially compensate, I brightened up the blue a bit on the home page.

Sunday, June 24

I added some comments about the material to be discussed on June 24. (Please try to check some of the values in the last 4 rows on my table pertaining to the proportions of outliers.) I thought that we had a very nice discussion last time, and I hope that everyone comes prepared with questions and comments this time. I've found several errors in the book, and so you might want to make a note of them so that you won't waste time trying to make sense out of things that are incorrect.

Sunday, June 17

I added some comments about the material for our first meeting on June 22. Also, in response to numerous complaints (all by the same person, btw), I eased up on the color brightness a bit. (I think this web site looks best when viewed in a dark room using Netscape Navigator.)

Thursday, June 7

Please be aware that if you want to take the course for credit, due to the schedule of meeting dates, the course cannot be completed until after the last day to submit grades for the summer session. All this means is that I would submit a grade of Incomplete in mid August, and then change your grade as appropriate when you complete your project (hopefully by the end of August or very early September).