Emil O W Kirkegaard

Is there a pkg for calculating cross-validated (k fold) fit measures for #SEM in #rstats? Seems possible with Mplus http://www.statmodel.com/discussion/messages/9/12127.html?1389902731 …

there's ecva in lavaan as a fit statistic - but as I think it's asymptoticallt equivalent to aic might not be worth hassle

I've learned not to trust analytic solutions to fit when one can just as easily use k fold cross validation. :)

well running k folds manually in a loop is pretty simple right?

I run sem powe analyses by hand, easier to specify models than canned power solutions and more flexible..