[R] Latent variable manual construction

Taman Powell powellth at Cardiff.ac.uk
Mon Apr 15 17:41:33 CEST 2013

Dear all,

I had a question around how to manually construct latent variables. I'll provide a little background….

Basically I have a SEM with 17 items, 6 latent variables and 615 rows of data.

I have all my model data (loadings, t values etc.) from lavaan, as well as correlation matrices.

I wanted to get p-values for my correlation matrix (items and latent variables) and understand I can't get this from lavaan. So I've done this for the items manually in r, but not sure how to construct/calculate the latent variables so that I can then determine the p-values.

Alternatively maybe there'a much easier way. As constructing the correlation matrices again when already have them from lavaan seems a little unnecessary, however, need to get these p-values somehow!

Any help much appreciated!

Best, Taman

More information about the R-help mailing list