Closed lf-araujo closed 3 years ago
thanks!
confint
gives SE-based confidence intervals (WALD type CIs).
umxCI()
gives likelihood-based CIs, which are superior, but usually similar to the SE-based CIs. They can be asymmetric, as the umxCI
actually explores how parameter changes drop the model fit. This can also make them time consuming to compute.
umxConfint
I think you can ignore - it's an solution from early in umx's development which I haven't yet rolled into umxCI and deprecated.
usability suggestions welcome t
Thanks Tim for this, I think in terms of usability there is nothing I would change.
Dear Tim,
Thanks for the package!
I am having difficulties understanding which confidence interval to report in my analyses. Currently working on an ACE,
Thank you.