cjvanlissa
commented
1 month ago Theoretically it is possible, but practically I have had real trouble getting such models to converge. The optimizer seems to be ill suited to such cases.
The way to go about it is to use the lavaan syntax for thresholds to specify the model for the ordinal indicators. Maybe it will converge if you use really good starting values...
lewolfra
I want to apologize in advance if this is a stupid question. And I know it's more of a question than an issue, but I would still highly appreciate any answer!
In the research project I am working in, we have been conducting a lca with ordinal and continuous indicators in Latent Gold. But since we want our research to be as easy to replicate as possible, we would like to approach this same lca in R. That's when I have come across the paper you wrote about best practices for lca (https://doi.org/10.1080/10705511.2023.2250920). The examples made were really helpful, but it's still not clear to me if it is possible with mx_mixture() or any other function to conduct an exploratory lca that contains both ordinal and continuous indicators (without "mistreating" the ordinal indicators as continuous). Of course I have tried a couple of things with the function to make this work but failed, so I keep wondering if something like this is even possible, since it was never clearly stated.
I hope my question doesn't take up to much time and effort to answer, and am looking forward to an answer!