Dear Martin,
Many thanks. I'll take a look at it. However, I raised a few other issues that are still outstanding (statalist 3223):
1. there is a question about what is the best way of reporting the Standard Deviations (SD) over the bootstrap samples. I hope that someone on the list might volunteer to (slightly) edit Patrick Royston's code, if needed. In particular, I think that Patrick's suggestion might be correct that "It may be preferable to report the SD over the bootstrap samples, this being the bootstrap-based SE of the optimism". That part of the code would then need changing.
2. It would be wonderful of course if someone took up Patrick's suggestions of generalization of Tibshirani's enhanced bootstrap procedure to other commands than just <logit>.
Kind regards,
Gerben
begin:vcard
n:ter Riet;Gerben
fn:Gerben ter Riet
tel;fax:+31 20 5669186
tel;work:+31 20 5664640
org:Academic Medical Center;General Practice
adr:;;Meibergdreef 15, Room J2-117;Amsterdam;;1105 AZ;The Netherlands
version:2.1
email;internet:[email protected]
title:Associate Professor
end:vcard
