|
|
Z scores as factor indicators |
|
Message/Author |
|
Gareth posted on Friday, August 07, 2009 - 2:16 am
|
|
|
I have a latent variable defined by four indicators. I have taken Z scores because they are measured on quite different scales. Using scores helps convergence. However, one of the Z scored variables is categorical, which means I get an error message "MOST LIKELY DUE TO VARIABLE Z1 BEING DICHOTOMOUS BUT DECLARED AS CONTINUOUS." Would you recommend using Z scores for the other three and 1/0 for this variable, or something else? |
|
|
I would not use z-scores. I would rescale the continuous variables by dividing by a constant. I would do nothing to the binary variable. |
|
Back to top |
|
|