Performing CFA with WLSMV in JASP's SEM module (ordinal manifest variables)
Hi! I performed a series of CFA's using JASP's SEM modules to test the factor structure of a 36 item emotion regulation questionnaire (DERS). I love the way JASP gives us beautifully formatted tables that are ready to use.
But I realized that some of the values i get with JASP are not quite consistent with what I get using lavaan directly.
It seems that JASP doesn't use the scaled fit indices and chisquare estimates that which are computed with lavaan when using WLSMV on CFAs with ordinal items.
Is that the case? If so, why could anyone elightnen me to why that choice was made, and where I can find literature to support this decision?
Any help will be welcome!
Thanks in advance!