Login  Register

Re: Negative binomial regression analysis: Different results in SPSS and STATA?

Posted by Student073 on Jan 26, 2014; 11:50am
URL: http://spssx-discussion.165.s1.nabble.com/Negative-binomial-regression-analysis-Different-results-in-SPSS-and-STATA-tp5724145p5724152.html

Rich, Ryan, thanks you!

These are the results I got. I tested the moderating effect of "moderator" on the relation between IV and DV. Although none of the analyses showed a significant result, I'm still alarmed the results were so different... Even in SPSS, there's a considerable difference in the significance levels if I use the "model-based estimators" compared to the "robust estimation".

What should I do to get the more accurate outcome??? Thanks again!!

STATA

. nbreg DVcount Controlvar1 Controlvar2 IV Moderator IVxModer

Fitting Poisson model:

Iteration 0:   log likelihood = -55609.736  
Iteration 1:   log likelihood = -45032.466  (backed up)
Iteration 2:   log likelihood =  -27555.08  (backed up)
Iteration 3:   log likelihood = -19627.822  
Iteration 4:   log likelihood = -8845.6562  
Iteration 5:   log likelihood = -8488.7504  
Iteration 6:   log likelihood = -8388.2676  
Iteration 7:   log likelihood = -8388.1451  
Iteration 8:   log likelihood = -8388.1451  

Fitting constant-only model:

Iteration 0:   log likelihood = -5506.8228  
Iteration 1:   log likelihood = -4983.0922  
Iteration 2:   log likelihood = -3158.2798  
Iteration 3:   log likelihood = -3157.9536  
Iteration 4:   log likelihood = -3157.9535  

Fitting full model:

Iteration 0:   log likelihood = -3078.1447  
Iteration 1:   log likelihood = -3027.5247  
Iteration 2:   log likelihood = -3024.3007  
Iteration 3:   log likelihood = -3024.2785  
Iteration 4:   log likelihood = -3024.2785  

Negative binomial regression                      Number of obs   =       5447
                                                  LR chi2(5)      =     267.35
Dispersion     = mean                             Prob > chi2     =     0.0000
Log likelihood = -3024.2785                       Pseudo R2       =     0.0423

------------------------------------------------------------------------------
     DVcount |      Coef.   Std. Err.            z       P>|z|       [95% Conf. Interval]
-------------+----------------------------------------------------------------
 Controlvar1 |   .4006587   .1612887     2.48   0.013     .0845387    .7167787
 Controlvar2 |   .0192326   .0034626     5.55   0.000      .012446    .0260192
          IV      |   .0699464   .0091503     7.64   0.000     .0520122    .0878806
   Moderator |   .2164698   .0293511     7.38   0.000     .1589428    .2739968
    IVxModer |  -.0018834    .003332    -0.57   0.572    -.0084139    .0046471
       _cons   |   -1.74222   .1218459   -14.30   0.000    -1.981034   -1.503407
-------------+----------------------------------------------------------------
    /lnalpha |   2.803212   .0562555                      2.692953    2.913471
-------------+----------------------------------------------------------------
       alpha |   16.49756   .9280785                      14.77525    18.42063
------------------------------------------------------------------------------
Likelihood-ratio test of alpha=0:  chibar2(01) = 1.1e+04 Prob>=chibar2 = 0.000


SPSS

MODEL BASED ESTIMATOR

* Generalized Linear Models.
GENLIN DVcount BY Controlvar1 (ORDER=ASCENDING) WITH Controlvar2 IV Moderator IVxModerator
  /MODEL Controlvar1 Controlvar2 IV Moderator IVxModerator INTERCEPT=YES
 DISTRIBUTION=NEGBIN(1) LINK=LOG
  /CRITERIA METHOD=FISHER(1) SCALE=1 COVB=MODEL MAXITERATIONS=100 MAXSTEPHALVING=5 PCONVERGE=1E-006(ABSOLUTE) SINGULAR=1E-012 ANALYSISTYPE=3(WALD) CILEVEL=95 CITYPE=WALD LIKELIHOOD=FULL
  /MISSING CLASSMISSING=EXCLUDE
  /PRINT CPS DESCRIPTIVES MODELINFO FIT SUMMARY SOLUTION.


ROBUST ESTIMATION

* Generalized Linear Models.
GENLIN DVcount BY Controlvar1 (ORDER=ASCENDING) WITH Controlvar2 IV Moderator IVxModerator
  /MODEL Controlvar1 Controlvar2 IV Moderator IVxModerator INTERCEPT=YES
 DISTRIBUTION=NEGBIN(1) LINK=LOG
  /CRITERIA METHOD=FISHER(1) SCALE=1 COVB=ROBUST MAXITERATIONS=100 MAXSTEPHALVING=5 PCONVERGE=1E-006(ABSOLUTE) SINGULAR=1E-012 ANALYSISTYPE=3(WALD) CILEVEL=95 CITYPE=WALD LIKELIHOOD=FULL
  /MISSING CLASSMISSING=EXCLUDE
  /PRINT CPS DESCRIPTIVES MODELINFO FIT SUMMARY SOLUTION.

NBR_SPSS.doc