|
BGonline.org Forums
Variance Reduction
Posted By: Bob Koca In Response To: Variance Reduction (rambiz)
Date: Monday, 21 July 2014, at 6:37 p.m.
"Do you possibly have any ideas how to include the effect of the inaccurate sample variance"
i) Just ignore it. It will be accurate enough. ii) I think a t - test with degrees of freedom equaling one less than number of trials could be used to account for that. It is extremely close to normal distribution though for a large number of trials.
It seems you are fine with CI from 0 - 1 sampling despite it having the same problem.
"Yes I need to know how much VR helps"
Is there a reason beyond curiosity? Knowing that might help how to best answer.
|
BGonline.org Forums is maintained by Stick with WebBBS 5.12.