I just wanted to get an expert second opinion. I've done a CFA in Mplus and the fit was such that I felt I could not diagnose the misfit in CFA. I then went ahead and did an EFA in Mplus on the same sample. I personally can't see any problems doing it this way, but I have several co-authors questioning the methodology. And when I ask them why, they simply say "well it's never done that way." I can't find anything saying this method is wrong. Thanks!
I think EFA is an excellent way to understand the measurement structure. And it can help support the notion of having "m" factors despite pure, simple structure not existing in the usual CFA sense. See also our ESEM page on the web site.
Thank you Bengt! I always feel better when I have the opinion of an expert in the field. And you and Linda definitely fit that description. Doing an applied paper using your ESEM methodology is on my list of next projects to do. I think it is a great idea!
BTW, I'm a big fan of Mplus and what you and Linda do here. I use Mplus in teaching my class and all my research work. I really hope you and Linda keep at this for a long time because what you offer here and with Mplus is second to none in my opinion!