Good evening,
So this isn't about SBI per se, but is with regard to the blog posts about finding
real data (which are great, by the way). I happened upon this article in Lancet that I
like for a couple of reasons. It's from 1998, but I think it's still worth
considering:
Wolkenstein et al. (1998). Randomised comparison of thalidomide versus placebo in toxic
epidermal necrolysis. The Lancet 352, 1586-1589.
It's a small sample size (appropriate for Fisher's Exact Test), and the study had
to be stopped because more people were dying from the active treatment than placebo (10/12
vs. 3/10). Thus, it's a nice opportunity to talk about ethics when studying human
subjects and provides a real (and I think interesting as well) example of Fisher's
Test.
Megan
Megan J. Olson Hunt, PhD
ASSISTANT PROFESSOR, STATISTICS
.............................................................................
Department of Natural and Applied Sciences, LS 465
University of Wisconsin-Green Bay
2420 Nicolet Drive
Green Bay, WI 54311
Office: LS 427 |
blog.uwgb.edu/olsonhunt | olsonhum(a)uwgb.edu