Monday, December 5, 2011

What should be the ideal relation between the sample sizes of the experimental and the control group?

The experimental group should be smaller than the control group.


The experimental group should be larger than the control group.


Both groups should be the same size.


The relation between the sizes of both groups doesn鈥檛 matter.|||I'm gonna guess that it's best when they're both the same size, because that makes mathematical comparisons easier ...... but I could be wrong.|||depends on what test you're running and what you're trying to accomplish


if it's a t-test, you probably want the sets to be the same size


if you're running a z-test, you use real process data to estimate mean and standard deviation of the "control" set so the actual number of control samples = 0

No comments:

Post a Comment