Thanks to all who responded
Andy Liaw suggested
wild bootstrap which essentially leave all residuals along, but
bootstrap their signs. (The procedure is a bit more elaborate, e.g.,
some moment matching are involved, but that's the basic idea.)
Jim Stapleton suggested using Wilcoxon rank sum test.
Tim Hesterberg suggested using permutation tests, but he also sent his
message to the list. (BTW, I got the Resample Library beta version which
he mentioned, and I like it a lot).
Chuck Cleland suggested that my concerns might be overstated, and
suggested a Welch t-test
Finally Dimitris Ritzopolous provided an S+ program to avoid the error
I was having
THanks again
Peter
Peter L. Flom, PhD
Assistant Director, Statistics and Data Analysis Core
Center for Drug Use and HIV Research
National Development and Research Institutes
71 W. 23rd St
New York, NY 10010
(212) 845-4485 (voice)
(917) 438-0894 (fax)
|