Course Information
Syllabus
Great quotes
R Code from class
IMPORTANT NOTES:
Project Progress Reports: They've been moved back to the week of Nov 13.
Like the PIPs, we'll try to keep them to 7 minutes, at most two slides
(emailed it to me ideally as pdf), and I'll try to be tougher about
keeping people on time. I haven't decided on how to determine and
reveal the order of presentations.
HOMEWORK
Problem 1: Due on August 28
Modify the rare disease simulation to include some fraction of
false negatives. Make a graph that illustrates something interesting.
Does your simulation match the mathematical expectation for that case?
Kind of assigned problem 2: "Due" on September 4
Take a look at the data people generated (source the file "readraw.R"
in Rcode/geometric) and then use the ideas in fiendish.R. Can you see
what the other fiends were up to?
Really assigned homework 3 for September 11
Figure out how to create data where the event rate lambda changes
over time. Make a Kaplan-Meier plot of your results. Does it
resemble your original assumption?
Homework 4 for September 25
1. Use the word "heteroscedasticity" in conversation with a non-statistician.
Extra credit if it triggers a serious long-term relationship.
2. Create, find, or steal a data set with one continuous covariate and
one factor (with two or more levels) that has a significant interaction.
Show that recentering the data does not change the interaction, but does
change the "main effects". Create a graph that makes sense of these
results.
