Your dependent/response/explained/left-hand-side/y-variable should not be standardized, as you want to use an ordinal regression model, which will ignore any information except for the order of the values.
I don't see how question 2 can give you information on "financial distress": A change could be distressing or the opposite, no change at a low level could be very distressing, while no change at a high level can would be the opposite. So any value on that variable could represent both distress and the opposite.
The easiest solution is probably to turn all the ordinal variables for financial distress into indicator/dummy variables, include the indicator variables in your model and use sheafcoef for the entire set of indicator variables. That way you can get a single effect for financial distress. This assumes that the observed variable influence the latent financial distress rather than the other way around.
I don't see how question 2 can give you information on "financial distress": A change could be distressing or the opposite, no change at a low level could be very distressing, while no change at a high level can would be the opposite. So any value on that variable could represent both distress and the opposite.
The easiest solution is probably to turn all the ordinal variables for financial distress into indicator/dummy variables, include the indicator variables in your model and use sheafcoef for the entire set of indicator variables. That way you can get a single effect for financial distress. This assumes that the observed variable influence the latent financial distress rather than the other way around.
Comment