Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: multilevel analysis


From   Andi Kopf <andik79@web.de>
To   statalist@hsphsun2.harvard.edu
Subject   st: multilevel analysis
Date   Sat, 25 Apr 2009 16:04:21 +0200

Hey everybody, as discribed above, I'm working on my final master thesis 
including multilevel analysis with 3 metric group-level variables. I've got 6 ML-models
1.empty model
2. random intercept with individual level fixed effects
3. Random intercept model with individual- und group-level fixed effects
4 Random slope model mit individual-level fixed effects
5. Random slope model with individual- und group level fixed effects und cross-level interaction
6. Random intercept model mit individual- und group-level fixed effects und cross-level interactions

So this is my first multilevel analysis and the results rised some questions:
Model 3 produced only not significant group level effects.
Model 5 produced one significant and two not significant (p>|z|=>0.05) group variables, the cross-level-interactions
showed: two are significant, one isn't. One of the significant cross-level-interactions is also the only significant group level variable.
So, how to interpret these findings?
And what does it mean that a group-level variable is not significant but the cross-level interaction still is?

Hope someone can help me
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index