# FW: Re: st: multilevel analysis

 From Andi Kopf To statalist@hsphsun2.harvard.edu Subject FW: Re: st: multilevel analysis Date Sat, 25 Apr 2009 23:46:04 +0200

```Hm...I worked the hole day on the model, tested different theoretical assumptions,
but...I don't understand the output with respect to the gini-var and the cross-level-interaction
EJC_gini. Am i right if i say that the overall effect of income inequality to the dependent variable is negative but not signifikant
-->no explanation through gini. And in regard to the c-l-i, the higher the inequality of a country, the more the extrinsic JC influence the
dependent job satisfaction. Is that what the model says?

Mixed-effects ML regression                     Number of obs      =     17737
Group variable: laender                         Number of groups   =        31

Obs per group: min =       299
avg =     572.2
max =       956

Wald chi2(10)      =   9436.08
Log likelihood = -23805.018                     Prob > chi2        =    0.0000

------------------------------------------------------------------------------
JS |           Coef.       Std. Err.      z       P>|z|         [95% Conf. Interval]
-------------+----------------------------------------------------------------
intrinsicJC      |    .369663   .0084099    43.96   0.000     .3531799    .3861462
extrinsicJC     |   .1572645   .0396179     3.97   0.000     .0796148    .2349141
relatedness   |   .5013823   .0102363    48.98   0.000     .4813195    .5214451
altersgrup~n  |   .0652755   .0065998     9.89   0.000     .0523401    .0782109
fullpart       |   -.054204   .0238984    -2.27   0.023     -.101044    -.007364
SEX       |  -.0420816    .014417    -2.92   0.004    -.0703384   -.0138249
Bildung     |  -.0297484   .0063537    -4.68   0.000    -.0422014   -.0172955
ISEI       |  -.0007213   .0005112    -1.41   0.158    -.0017232    .0002806
gini        |  -.0050667    .004624    -1.10   0.273    -.0141295    .0039962
EJC_gini    |   .0031164   .0010966     2.84   0.004     .0009671    .0052657
_cons     |   .7606644    .171826     4.43   0.000     .4238917    1.097437
------------------------------------------------------------------------------

------------------------------------------------------------------------------
Random-effects Parameters  |   Estimate   Std. Err.     [95% Conf. Interval]
-----------------------------+------------------------------------------------
laender: Identity            |
var(_cons) |   .0156366   .0044273      .0089771    .0272365
-----------------------------+------------------------------------------------
var(Residual) |   .8539798   .0090763      .8363746    .8719556
------------------------------------------------------------------------------
LR test vs. linear regression: chibar2(01) =   211.44 Prob >= chibar2 = 0.0000

Hope someone can help,
Andy

*
*   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/
```