Lin posted on Sunday, December 18, 2011 - 11:48 am
Hi, I've run an EFA with 26 variables, 8 out of them were categorical. The model suggested 5 factor but I've got factor loadings who were greater than 1 or negative. For example, AROAD. Also some variables held small factor loading, such as LUMHHI, BROAD. Would you have any suggestion? Shall I think about get rid of them? Thanks.