I have a large number (~1000) of number sets containing 8 numerical elements. Here's an example:
A = (5, 10, 9, 11, 4, 8, 8, 9)
B = (14, 10, 12, 19, 20, 13, 8, 13)
C = (8, 8, 8, 9, 8, 8, 9, 10)
I want to determine which sets vary the least. I have already calculated mean, standard deviation and variance for each set but I am not sure how to apply these calculated values to my question. Can I simply group the sets in order of increasing variance and say that the lower the variance, the more consistent the numbers within that set are?
In case you are interested: the practical application is that each set represents a measurement of a biological system across 8 control samples; the goal is to prioritize which measurements are least likely to give wildly different values unless altered by experimental conditions.
Variance is invariant under addition. That is $\{9,11\}$ and $\{90,92\}$ will have the same variance although they have different means. [if you subtract the mean from each they will be $\{-1,1\}$]
However, variance is not invariant under multiplication. Consider $\{9,11\}$ and $\{90,110\}$. These do not have the same variance, but, if you divide by the mean both sets become $\{0.9,1.1\}$ which do have the same variance.
You need to decide, based on your context, which measure is more relevant for you.