Running an LCA on ordinal data using tidySEM, issue with thresholds

35 Views Asked by At

This is the first time I am running a Latent Class Analysis and using tidySEM.

I have attached my descriptives table below for the 8 indicators I want to enter into the model. Seven of the variables are Likert scale responses either 1-4, 1-5 or 1-7. One variable is a binary no/yes response coded as 0/1 respectively.

enter image description here

I have cleaned the data and there are no missing variables (I have a big enough dataset to only include full responses). When I try and run the model I am getting errors that I don't know how to address.

Is there anyone well-versed in tidySEM LCA modelling that could help me troubleshoot if I am doing something wrong? I'm finding the learning curve and entry-level R specific information on running LCA models very challenging.

When I use the following sytax:

set.seed(123)
tidySEM::mx_lca(data = LCA1_noNA, classes= 1:6, run = TRUE)

I get the following error:

Error in update_thresholds(zscore) : Could not complete thresholds; either specify all thresholds by hand, or remove constraints.

0

There are 0 best solutions below