送交者: wasguru 于 2006-6-27, 13:59:11:
回答: Sample size is not the problem here as the study detected the difference, but 由 blackbox 于 2006-6-27, 13:30:01:
In an observational study like this, it is quite possible that there are large background covariate inbalances. I'm more interested in how they eliminate the possible bias raised from such background covariate inbalances.