I have to vectors $\mathbf{y1_{post}}$ and $\mathbf{y2_{post}}$ (post because they were measured after a treatment).
y1=80.564,89.299,89.829,77.371,76.883,87.038,87.83,89.127,89.291,48.011,47.5,55.081,55.358,13.54,12.641,41.106,42.555,2.599,2.607,20.762,20.867,3.088,2.881
y2=48.464,47.571,47.313,43.149,42.868,35.127,34.984,33.347,31.039,11.4,11.47,17.422,16.859,1.596,1.89,16.503,15.749,1.405,1.398,3.82,3.805,1.2,1.17
I want to see how these values have changed relative to some pretreatment (and fixed) measures $y1_{pre}$ and $y2_{pre}$. For $y1_{pre} = 80.112$, and for $y2_{pre} = 48.51$.
This paper says that relative change should be maseured as $\log(y_{post}/y_{pre})$. What I want is a masure of how the two vectores changed from $y_{pre}$ relative to the other, to se which vector changed the most. For example, the ratio of their absolute difference, something like $\frac{\mid y1_{post} - y1_{pre}\mid}{\mid y2_{post} - y2_{pre}\mid}$, which, for the first post-treatement measure will be $\frac{\mid80.564 - 80.112\mid}{\mid48.464 - 48.51\mid} = 9.8$ and, if they change the same amount, it will approach 1.
My question is whether there is a standard measures for this, as I don't know my approach is correct or will introduce some artifacts or bias.
