|
![](http://www.statmodel.com/discussion/icons/csc/1px.gif) |
Best format for 3-cat ordinal vars in... |
![Previous](http://www.statmodel.com/discussion/icons/csc/navgen_arrow_left.gif) ![](http://www.statmodel.com/discussion/icons/csc/1px.gif)
|
Message/Author |
|
Guido Benini posted on Friday, September 06, 2013 - 11:52 am
|
|
|
Hi, We are constructing LCA with dichotomous, ordinal, and interval-level data. One of these could be considered a "count," but the others are three-point scales. From a statistical perspective, is it is better to enter those as counts, categorical data with distances constrained, or as regular interval-level scales? Thanks! |
|
|
I would only put true count variables on the COUNT list. I would put the variables with three-point scales on the CATEGORICAL list if the three points are ordered. If not, I would put them on the NOMINAL list. |
|
|
Thanks! |
|
Back to top |
|
|