1 part E. Few people realized that you should still
check the resuduals. they just got the expected result and said "we
believe this one since it is what we expected." Are the p-values here
any better? Only after checking do we know this.
3 part 5 you need to guess! How small a p-value will it take you to
be impressed. If you truely won't believe anything than you can go
with "unknown = infinity." Otherwise, you need to specifiy what
p-value you want to go with and hence what n you are using.
Ponzironi issues
discuss McCall, a fund manager (handout)
discuss statistical forecasting paper (how do you count n here?)
Homework 4 (you may turn it in on Nov 26th before 10:00 and be in the first batch)
Homework 5 / Presentations
form your groups and email your title and prefered day
Time series using JMP
hit "analyse time series" button--the rest is obvious!