Since neither WSO nor Williams cares that much about alumni/student interaction, I am unable to participate in this thread about a recent study in the Lancet on Iraqi mortality rates. If I could participate, I would point out that the Wikipedia article is a good place to start and includes a link to the actual report. I also spend absurd amounts of time on this topic at places like this and this.

But the interesting Eph-specific comment is from Ken Flax:

we read the 2004 study (lancet) for psyc stats as just an exercise in statistics….take-home point, the results are very robust and the study was peer-reviewed by other statisticians that belived it was sound. the guy from johns hopkins (les roberts?) is well-known for calculating war-time civillian deaths in places like the congo, etc.

1) Count me among the statisticians with serious doubts about both Lancet studies.

2) Interesting how this article was selected for an introductory stats in psychology. Would that be PSYC 201? Who was the professor?

3) Say what you will about the article, but it has nothing to do with psychology? It is the most famous Bush-bashing statistics article published in the last few years. Perhaps it was selected randomly! Did the professor offer any critiques of the article? Just asking!

4) For the record, I assigned/cajoled one of my summer interns into using the article/data for his STAT 201 project, so using the article is not, ipso facto, evidence of anti-Republican bias. If I can get his permission, I’ll post it here.

UPDATE: For those interested, here is an R package by Kyle Campbell ’08 created for STAT 201 with some of the underlying data and some nice graphics.

Print  •  Email