|
|
| Message/Author |
|
|
| Sunny Shin posted on Thursday, February 19, 2009 - 2:39 pm
|
|
|
| My project is to run LCA of adolescent substance use. I am trying to constrain one class to have 0 probability of use across all 6 illicit substances. This is to reflect an "abstainer" class, which is based on previous findings and exploratory analyses of my data. Could you let me know if you have any example of this? Thanks. |
|
|
|
|
Assuming that your outcomes are binary, you simply say (assuming class 1 is the zero class and that for 10 binary u's illicit use is u=1 as opposed to 0):
%c#1%
[u1$1-u10$1@15]; |
|
|
| Back to top |
|
|
