
BGonline.org Forums
Variance Reduction
Posted By: rambiz
Date: Monday, 21 July 2014, at 9:42 p.m.
In Response To: Variance Reduction (Bob Koca)
OK, I'm going to ignore the inaccuracy of sample variance, but then I wouldn't be too sure about the confidence intervals, either. Unless the difference is really small i.e. the final estimate is not sensitive to this inaccuracy.
I need to find out what a ttest is.
I was fine with CI from binary sampling because I could use the p*(1p) formula and didn't have to rely on the sample variance. The VRvariables are almost continuous with unknown pdf, so we have to resort to the sample variance. You had introduced the idea of sample variance in one of your earlier posts, however I unfortunately missed it and resold your idea as my own later, sorry for that.
beside curiosity, it is the disbelief which derives me. I want to reproduce the confidence intervals, so that I also make sure, I understand this stuff to some extent.

BGonline.org Forums is maintained by Stick with WebBBS 5.12.