The issue is that an empty cell implies a correlation of one. When two variables correlate one, only one of them should be used in the analysis.
Abdel posted on Saturday, April 09, 2011 - 11:48 pm
Thanks! This may be a silly question but: what if an empty cell occurs only in one or two of the many groups in a multigroup CFA? Would it still be okay to use both variables in the analysis or would that violate some assumption?