1
$\begingroup$

I am new to statistics, and have a fairly basic question. I have two sets of data that I want to compare with a third dataset. I am only interested in comparing the means and variances.

Turns out that the mean of dataset3 is close to that of dataset1. However, when it comes to variances, dataset2 and dataset3 show close agreement. So, I am not sure which dataset is actually similar to dataset3.

So, I would like to simultaneously compare means and variances. Is there a single metric using which I can compare both together? I am not aware of the distribution of both datasets.

Any help or hints would be greatly appreciated.

-ryan

$\endgroup$
  • $\begingroup$ Question - what is that you want to achieve by comparing them simultaneously. One is measure of "center" of all data, and one is measure of how far the data is from "center". More context will be useful $\endgroup$ – Gaurav Singhal Oct 11 '16 at 6:03
  • $\begingroup$ Why not qualitatively report your findings? You can't really use statistics to convey how "close" datasets are, anyway. $\endgroup$ – AdamO Oct 11 '16 at 14:09
  • $\begingroup$ How about the CV--coefficient of variation, calculated as the std dev divided by the mean--a scale invariant measure of variability? It makes comparable objects that are widely different in mean and variance. $\endgroup$ – Mike Hunter Oct 11 '16 at 14:10
  • $\begingroup$ @AdamO: thanks, but there is a lot of such comparisons in my work. I can't be qualitative all the time. $\endgroup$ – ryan80 Oct 11 '16 at 14:17
  • $\begingroup$ @DJohnson: CV sounds like a very good idea! I will try it out. Thanks! $\endgroup$ – ryan80 Oct 11 '16 at 14:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.