I just took Bengt's class in SF last week and learned about residual covariances within latent class analysis. I'm conducting an EFA with items that were measured with Likert scales. Can I use the latent class analysis syntax with the added residual covariances? The wording of the items were similar across many of my items so the residual covariances seem applicable.
No, this new feature does not apply to EFA. If you do EFA via ESEM, you can include residual covariances in the model. See Example 5.24. This example is the same as EFA if you exclude the covariates and direct effects. You can include residual covariances using the WITH option.