Ordinal DV- only 1 set parameter esti... PreviousNext
Mplus Discussion > Categorical Data Modeling >
 Kandi Felmet posted on Monday, June 29, 2015 - 7:15 am
Hi, I am running multinomial logistic regressions in mplus and my DV is ordinal with 4 levels (match persistence: 0 = 3-6mos; 1= 6-9 mos; 2 = 9-12mos; and 3 = 12 months +). However, my output only has one set of parameter estimates, despite the 4 ordinal categories. I am confused b/c I expected 4 sets...??? Any ideas?

Here is my model syntax for reference:

usevariables = persist2 childage volsex ACrel_2
ACgd_2 exp3m frus3m white bach author;

missing = all (999);
categorical = persist2;
analysis: estimator = MLR;
ITERATIONS = 1000000;

model: persist2 on ACrel_2 ACgd_2 childage volsex exp3m frus3m white bach author;

output: sampstat stdyx;
 Bengt O. Muthen posted on Monday, June 29, 2015 - 5:17 pm
You talk about multinomial regression which is for a nominal DV, but then you say your DV is ordinal. The DV is specified as ordinal if you use Categorical= and it is specified as nominal if you use Nominal=.
 Kandi Felmet posted on Monday, June 29, 2015 - 5:26 pm
Yes, I'm sorry, to be clear- I have an ordinal DV and I am using the specification Categorical. Since there are 4 levels, why am I not getting 4 sets of parameter estimates..?
 Bengt O. Muthen posted on Wednesday, July 01, 2015 - 5:38 pm
With a 4-category ordinal DV regressed on an x variable, you get 1 slope and 3 thresholds. See our video and handout for Topic 2 of our short courses on our website.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message