Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | "JVerkuilen (Gmail)" <jvverkuilen@gmail.com> |
| To | statalist@hsphsun2.harvard.edu |
| Subject | Re: st: Hierarchical CFA problem |
| Date | Mon, 22 Apr 2013 08:32:04 -0500 |
On Mon, Apr 22, 2013 at 4:31 AM, W Robert Long <W.R.Long@leeds.ac.uk> wrote: The key is in the lavaan output: > L1 0.192 0.170 > L2 0.709 0.107 > L3 0.272 0.069 > G 0.617 0.183 Notice how small the variance of L1 is, and how large its SE is. This is indicative of a log-likelihood that is close to being irregular, in this case probably having a mode at or close to 0 for L1's variance. L3 has a slightly higher variance but a markedly lower standard error, which means it has a proper peak. While you shouldn't use these standard errors to compute hypothesis tests or confidence intervals, they are highly indicative of problems with a model. So I'm not surprised difficult cleaned things up. (lavaan is a nice piece of kit, by the way.) -- JVVerkuilen, PhD jvverkuilen@gmail.com “He uses statistics as a drunken man uses lamp-posts – for support rather than illumination.”--Andrew Lang * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/faqs/resources/statalist-faq/ * http://www.ats.ucla.edu/stat/stata/