How do I calculate the coefficient of variation, CV (or Relative Standard Deviation, RSD) for a subset? e.g.

group =    A  A  A    B  B  B
X     =    1, 2, 3,   4, 5, 6

Do I use the mean of the subset

CVA = sdA / meanA = 1 / 2 = 50%

...or the grand mean of the data

CVA = sdA / meanX = 1 / 3.5 = 29%

I found only one paper suggesting it is the grand mean here, but most sources simply don't mention subsets...

Many thanks.

  • $\begingroup$ Either could be defensible. Whatever you do, make it clear. $\endgroup$ – Nick Cox Apr 12 '15 at 21:54

If there is homogeneity within groups and the cause of variation between them is isolated, you could estimate their effects, remove from the observations and then calculate the CV. Otherwise you should do the calculations with the whole set of data.

| cite | improve this answer | |

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.