New to LCA please help with a methodological question

Hello, I am new to LCA and attempting to design a secondary data study utilizing LCA as to show latent classes (subtypes) of a specific disorder based on symptoms of this condition ie. aggression toward others (13 symptoms total)that are binary (yes/no)(as measured by a behavioral checklist) I am also interested in how emotional intelligence, (another measured construct containing a number of variables that is included within the data set), is associated with the latent classes. Is it best practice to run the LCA first on the symptoms of the disorder and assume emotional intelligence variables are to be treated as covariates, or is there some other way to find (explain methodologically) how to statistically account for finding the association of EI with the latent classes? Any help is greatly appreciated, thank you very much!
Last edited: