Re: Differences in latent means

[Edward Rigdon]

Sara--

     These postings don't include syntax for the model itself, which will make it hard to identify an error there. Your two images of results apparently relate to different models, based on the object labels. In general, unless sample size is small or there are substantial differences in the scaling of observed variables, or estimation failed to converge, you can trust lavaan when it says there seems to be an identification problem.

     I believe strong invariance, which would also include equality of intercepts, is necessary to rule out confounds when comparing factor means.

On Sat, Sep 13, 2025 at 6:55 AM Sara <sarahfi...@gmail.com> wrote:

Hi,

I have fitted a partial threshold + loading invariance model to compare latent means relative to the year 2000 (baseline), where by default the latent mean is fixed to 0 (first image)

 Now I would like to compare the latent means between successive years and perform Wald tests on these differences.  However, I’m getting a warning about the variance–covariance matrix. In addition, I don’t understand why the factor mean for the baseline year 2000 is being estimated (it shows a non-zero value in the output) even though I explicitly fixed it to 0 in the syntax (second image)

model.base <- '
FA =~ F + C + E + G
FA ~ c(0, m2002, m2004, m2006, m2008, m2010)*1
'

fit.base <- cfa(model.base,
                                  data = long_data,
                                   group = "time",
                                   ordered = c("C","E","F","G"),
                                 group.equal = c("thresholds","loadings"),
                                   group.partial = c("C|t3","F|t3"))

> lavInspect(fit.base, "converged") [1] TRUE

--
You received this message because you are subscribed to the Google Groups "lavaan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lavaan+un...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/lavaan/c0b9a4e7-ee33-4c2e-8836-0fdfe7533d64n%40googlegroups.com.