Sorry to jump into this but I dont see whay Dan cannot
use z-scores as a dependent variable, even if z-scores
are standarized. Wouldnt it depend on which are the
independent variables used?
No, no reason why not, and it actually doesn't depend on the
regressors. If you run
z(y) = a + b X + u
b is the number of standard deviations of y in response to a one unit
change in X. If that makes sense as a way to express dY/dX, fine. If
you run
z(y) = b z(X) + u
you are calculating the 'beta coefficient', i.e. regress y X, beta.
