consider moving consistency/variability analysis to correlations

[mcfrank]

I want to see if the bug in CoV analysis can be solved by trying to force all the measures onto the same scale (namely correlation).

should work for most chapters, not sure about noun bias though, how do we compute that as a correlation?

[mikabr]

decided this was a dead end