1
$\begingroup$

I fit a CFA model in R with the lavaan package using the "MLM" estimator, as the data follows a non-normal multivariate distribution. All items are scaled on a six-point scale.

fit_E <- lavaan::cfa(E_model, data = x, std.lv = FALSE, orthogonal = TRUE, estimator = "MLM")
summary(fit_E, fit.measures = TRUE, modindices = TRUE, standardized = T)

All factor loadings exceed 1 in the standardized solution. Did I do something wrong?

parameterEstimates(fit_E, standardized=T) %>% 
  filter(op == "=~") %>% 
  select('Latent Factor'=lhs, Indicator=rhs, B=est, SE=se, Z=z, 'p-value'=pvalue, Beta=std.all) %>% 
  kable(digits = 3, format="pandoc", caption="Factor Loadings")

enter image description here enter image description here

$\endgroup$

1 Answer 1

2
$\begingroup$

In lavaan specifying standardized=T returns standardized loadings in a separate column in your output. The entire solution is not standardized and the first column still includes your unstandardized loadings, which will vary in magnitude as a function of the indicator variables' scales.

For standardized loadings look for std.all in your output, what you have stored as Beta in your table.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.