too high CV

#1
Hello friends, when i am analysing my data using ANOVA, i came up with a very large cv(55) is there any means to lower it? my design is 4x5 CRD with factorial arrangement.

the SAS out put is this
R-Square Coeff Var Root MSE LN Mean

0.516573 55.69053 1.842112 3.307765
 

Miner

TS Contributor
#2
Is your data interval or ratio scale? CV should only be calculated for a ratio scale. It has no real meaning for an interval scale, in which zero is an arbitrary value.
 
#3
Is your data interval or ratio scale? CV should only be calculated for a ratio scale. It has no real meaning for an interval scale, in which zero is an arbitrary value.
yeah, my data is percentage data. i have counted and converted to percentage. I think i shall go for arcsine or square root transformation but still my CV is too large
 

Dason

Ambassador to the humans
#4
By percentage data do you mean you have something like a total number of attempts and a total number of 'successes' for each observation? Or was the percentage based on something else?