Interpreting output for predictors PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
 Sophie Barthel posted on Wednesday, January 14, 2009 - 1:54 am

I have run the following code:

DATA: FILE IS comb1log.dat;
VARIABLE: NAMES ARE age t tdur liv x1-x9;
CLASSES = c (4);
Missing are all (-999) ;
STARTS = 20 2;
algorithm = integration;
i s | x1@0 x2@1 x3@2 x4@3 x5@4 x6@5 x7@6 x8@7 x9@8;
s on t;
c#1 on t;
c#2 on t;
c#3 on t;

But I have got difficulty in understanding the output - in particular, MPLUS outputs that:

Categorical Latent Variables

C#1 ON
T -1.516 -0.985 0.709 2.402 2.934

C#2 ON
T 9.235 9.706 11.205 12.704 13.175

C#3 ON
T -2.165 -1.991 -1.439 -0.886 -0.713

The above does no make sense to me, as classes 2 and 3 are virtually identical when looking at the graph (except that the intercept for 3 is slightly higher than for 2). Could you help me with the interpretation please?
 linda beck posted on Wednesday, January 14, 2009 - 3:39 am
I would say, if two of your classes look nearly identical you should prefer a 3-class solution instead of 4 classes independently of what fit- or test criteria say. May be that would help to get more plausible effects of c on t.
 Sophie Barthel posted on Wednesday, January 14, 2009 - 6:39 am
yes, I did think that but even then I still have those two classes separating. It does make sense conceptually, but I am stuck on how to interpret the different effect of T on them.
 Linda K. Muthen posted on Wednesday, January 14, 2009 - 8:10 am
The results you show look odd. Please send your input, data, output, and license number to
 Sophie Barthel posted on Monday, January 19, 2009 - 8:16 am
Linda - thank you for the offer but unfortunately I am unable to send the data. I did, however, realise over the weekend that I should have specified T in dummy variables as it is a nominal categorical variable. That has solved the strange output problem.

Still, I have got the following questions:

1) I am assuming that the model specification above does not estimate the effect of T on variability within classes? How would I specify that?

2) How do I specify that I would like the effect of s on T to be different across different classes?

Thank you!
 Linda K. Muthen posted on Monday, January 19, 2009 - 10:38 am
1. i s ON t;
2. t ON s;
 Sophie Barthel posted on Tuesday, January 20, 2009 - 1:20 am
thank you!

unfortunately, when I use

t on s

I get the following fatal error:
reciprocal interaction problem

Is there anything else I need to change in the model?
 Linda K. Muthen posted on Tuesday, January 20, 2009 - 6:28 am
Please send your full output and license number to
 Michael Spaeth posted on Tuesday, January 20, 2009 - 6:39 am
That's an easy one I guess!
I had the same problem some time ago.
you should mention i on t in the %obverall%-statement (if defined class invariant) and the class variant s on t in the class-specific statements (i think you have 4) for each class.
it goes like this:

s on t;

s on t;

The same procedure with other model parameters.
Furthermore, I would increase the starts (you try to extract 4 classes!!!) up to at least 500 20, to avoid local maxima.
stiterations = 20 also helps a lot to get more trustworthy solutions, especially when you have class variant effects.
In addition, class variant effects can often lead (that's my experience) to LL's not replicated. Then stscale = 1 sometimes helps to replicate a LL-value.
your coverage seems pretty low, which can also lead to estimates that are not trustworthy. Is this what you intended?

Tschuess, Michael!
 Sophie Barthel posted on Wednesday, January 21, 2009 - 1:53 am
Michael - thank you very much for your help! Also, your pointers about the starts are greatly appreciated.

 Gregory M. Dams posted on Saturday, January 20, 2018 - 3:54 pm
Good evening Dr. Muthen,
I've run an LPA with 3 predictor variables and selected a 2-cluster skew normal distribution model (unrestricted covariance matrix) as the optimal model. I'm having trouble understanding how to interpret the skew and df parameter output. A portion of my output looks something like this:

Skew and Df Parameters

Latent Class 1

Predictor 1 16.991 12.200 1.393 0.164
Predictor 2 14.354 5.769 2.488 0.013
Predictor 3 27.466 7.234 3.797 0.000

How do I know what the skew value is for each variable?
 Bengt O. Muthen posted on Saturday, January 20, 2018 - 4:39 pm
Send your full output to Support along with your license number.
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