One thing I've wondered about w/respect to combining chi-squares using the approaches discussed recently in this forum: Which types of chi-squares are legitimate to combine in this way? Only normal theory, maximum likelihood-based chi-square test statistics, or would it be possible to combine any "flavor" of chi-square in this manner? (e.g., Satorra-Bentler kurtosis-adjusted chi-squares, weighted least-squares-based model fit chi-square test statistics, etc.).
Thanks in advance for any input on this issue, Tor Neilands UCSF Tor Neilands Center for AIDS Prevention Studies (CAPS) Prevention Sciences Group (PSG) University of California, San Francisco 74 New Montgomery St., Suite 600 San Francisco, CA 94105 Voice: (415) 597-9236 FAX: (415) 597-9194 E-mail: [EMAIL PROTECTED]