Plan for Friday, Oct. 18 (3 PM meeting)

We'll start by finishing Ch. 3.  All that we have left is Dr. Gantz's
presentation of Sec. 3.8.  (But since we crammed in 3.7 quickly at the end
last time, you can feel free to ask Dr. Gentle some questions about it.)

I propose that we plan to spend most of the next 3 meetings (incl. 10/18)
on Ch. 4.  Despite some early difficulties with Harrell's book, Ch. 4
really gets us into some good material, and I think it will be best to not
rush through it. Plus, by allowing 3 weeks for Ch. 4, maybe we can take
some time to explore some data sets using various statistical software
packages.

For this coming Friday, I suggest that we only plan to cover through the
end of Sec. 4.4 on p. 61 (so basically, pp. 53-61 ... just 9 pages).  This
week I think longer prepared presentations will be appropriate --- but
keep in mind that it's nearly always a good idea to identify the main
idea(s) before addressing the major points one-by-one ... addressing some
of the details as well. 


             pp. 53-54   M.L.  

      top 1/3 of p. 55   C.S.

paragraph on p. 55       
starting w/              A.L.
"An open question ..."   

   bottom 1/2 of p. 55   J.W  (Can you find/make an example with mildish
                               "nonlinearity" and illustrate the F test
                               and the graphical method described?)
            
              Sec. 4.2   C.S        
 
      part of Sec. 4.3   J.G.  (Perhaps inserting a brief description  
          on pp. 56-58          of stepwise regression will be good.)

                 p. 59   J.S.
  
       rest of Sec 4.3   A.K.  (Can you find a small economic or 
                                medical data set that has an estimated
                                coefficient with the "wrong sign"?)

              Sec. 4.4   C.P. 

If others not referred to above plan to attend this week, identify a
small segment of the text and do an example related to it.

In addition to Problem 5 of Ch. 2 that I previously suggested, I think
it would be good to do something with the two problems at the end of
Ch. 3.  Strangely, Ch. 4 doesn't have any problems at the end of it, 
but I think we will want to do some regression analysis in order to get
a better feel for the techniques described in the text.


CDS