Ordered vs Unordered Solutions PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
Message/Author
 Willie McBride posted on Wednesday, May 02, 2018 - 10:49 pm
Hello

What is the difference between ordered and unordered solutions for a Latent class analysis with categorical variables?
 Bengt O. Muthen posted on Thursday, May 03, 2018 - 4:48 pm
I don't know about that term. Unordered usually refers to a nominal variable, as opposed to an ordinal variable.
 Willie McBride posted on Friday, May 04, 2018 - 9:27 am
Thank you. I am getting an error of this

*** WARNING in ANALYSIS command
Estimator WLSMV is not allowed with TYPE=MIXTURE.
Default will be used.
1 WARNING(S) FOUND IN THE INPUT INSTRUCTIONS

What would be the change of TYPE to?
 Bengt O. Muthen posted on Friday, May 04, 2018 - 4:37 pm
Mixture requires Estimator = ML(R).
 Willie McBride posted on Friday, May 04, 2018 - 4:38 pm
Thank you. What type would work with WLSMV?
 Bengt O. Muthen posted on Friday, May 04, 2018 - 4:44 pm
General single-level and multilevel.
 Willie McBride posted on Monday, May 07, 2018 - 1:45 pm
I have a data set that I am examining 9 variables that I am attempting to perform an LCA:
1 coded as binary
6 coded as ordinal with 3 levels each
2 coded with as ordinal with 4 levels each

Would this all be considered a categorical LCA?
 Bengt O. Muthen posted on Monday, May 07, 2018 - 5:30 pm
Yes.
 Willie McBride posted on Tuesday, May 08, 2018 - 7:26 am
This has been very helpful.

which type command and which estimator would you recommend using based on the variable description I have provided?
 Willie McBride posted on Tuesday, May 08, 2018 - 12:28 pm
How are we able to view class membership in the output? Is there a Tech command that should be used?

For example, how would I determine this information?

Mplus creates an output file which contains the original data used in the analysis (i.e., item1 to item9) followed by the probability that Mplus estimates that the
observation belongs to Class 1, Class2, and Class 3. Next, the class with the highest probability (the modal class) is shown.
 Bengt O. Muthen posted on Tuesday, May 08, 2018 - 2:50 pm
Use ML.

See the video and handout from our Short Course Topic 5.
 Willie McBride posted on Tuesday, May 22, 2018 - 1:44 pm
Hello

I am performing a LCA and I want to use the DEFINE command to sum the number of criteria met for my various classes. In the topic 5 video it stated that this can be done. I was getting an error and am not sure if I am writing it correctly.

Data:
FILE IS /Users/wfmcbr01/Desktop/Dissertation/Dissertation Work/dissertation_LCA_1156.csv;
Variable:
NAMES are wmt_lca msvt_lca nvms_lca rds_lca f_lca fp_lca fs_lca fbs_lca rbs_lca;
CLASSES = c (4);
CATEGORICAL = wmt_lca msvt_lca nvms_lca rds_lca f_lca fp_lca fs_lca fbs_lca rbs_lca;
DEFINE:
SUM = Sum (wmt_lca msvt_lca nvms_lca rds_lca f_lca fp_lca fs_lca fbs_lca rbs_lca);
Analysis:
TYPE = MIXTURE;

For some reason it is not working properly.

Thank you,
 Bengt O. Muthen posted on Tuesday, May 22, 2018 - 5:42 pm
Use Savedata to save CPROBS. Then look for the column with Most Likely Class. Then for each class, sum up the criteria.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: