Moderator and Independent Variable using Dummy (Binary Variable)

[Canthy Meilanda]

Hi Stata Experts Team,

I would like to know whether there is a different procedure in regression if there is independent variable dummy and moderator variable dummy. The current model I used for the model is fixed effect and I have checked the normality, multicolinearity, heteroscedasticity, and also the autocolinearity. Is the any additional procedure that I should perform for the model below?

xtreg roa relateddummy unrelateddummy efisiensidummy relateddummy*efisiensidummy unrelateddummy*efisiensidummy size growth leverage age, fe

Thank you for your attention.

Best Regards,

Canthy Meilanda

[Carlo Lazzaro]

Canthy:
welcome to this forum.
Some comments about your query:
- why creating interactions and categorical variables yourself when -fvvarlist- can do it for you?
- one of the most useful postestimation test concerns its correct specification:

. use "http://www.stata-press.com/data/r15/nlswork.dta"
(National Longitudinal Survey.  Young Women 14-26 years of age in 1968)

. xtreg ln_wage age grade, fe
note: grade omitted because of collinearity

Fixed-effects (within) regression               Number of obs     =     28,508
Group variable: idcode                          Number of groups  =      4,708

R-sq:                                           Obs per group:
     within  = 0.1026                                         min =          1
     between = 0.0879                                         avg =        6.1
     overall = 0.0774                                         max =         15

                                                F(1,23799)        =    2720.20
corr(u_i, Xb)  = 0.0314                         Prob > F          =     0.0000

------------------------------------------------------------------------------
     ln_wage |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         age |   .0181349   .0003477    52.16   0.000     .0174534    .0188164
       grade |          0  (omitted)
       _cons |   1.148225   .0102577   111.94   0.000      1.12812    1.168331
-------------+----------------------------------------------------------------
     sigma_u |  .40637763
     sigma_e |  .30349389
         rho |  .64195114   (fraction of variance due to u_i)
------------------------------------------------------------------------------
F test that all u_i=0: F(4707, 23799) = 8.81                 Prob > F = 0.0000

. predict fitted, xb
(26 missing values generated)

. g sq_fitted=fitted^2
(26 missing values generated)

. xtreg ln_wage fitted sq_fitted , fe

Fixed-effects (within) regression               Number of obs     =     28,508
Group variable: idcode                          Number of groups  =      4,708

R-sq:                                           Obs per group:
     within  = 0.1087                                         min =          1
     between = 0.1007                                         avg =        6.1
     overall = 0.0865                                         max =         15

                                                F(2,23798)        =    1451.88
corr(u_i, Xb)  = 0.0441                         Prob > F          =     0.0000

------------------------------------------------------------------------------
     ln_wage |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      fitted |   7.143482   .4789041    14.92   0.000     6.204799    8.082164
   sq_fitted |  -1.816235   .1414686   -12.84   0.000    -2.093523   -1.538948
       _cons |  -5.167839   .4038031   -12.80   0.000    -5.959319    -4.37636
-------------+----------------------------------------------------------------
     sigma_u |  .40395152
     sigma_e |  .30245467
         rho |  .64077441   (fraction of variance due to u_i)
------------------------------------------------------------------------------
F test that all u_i=0: F(4707, 23798) = 8.74                 Prob > F = 0.0000

. test sq_fitted=0

 ( 1)  sq_fitted = 0

       F(  1, 23798) =  164.83
            Prob > F =    0.0000
*As -test- reaches statistical significance there's evidence of model misspecification*
.

[Canthy Meilanda]

Dear Carlo,

Thanks for your response. There are further question I would confirm about the model on previous question:
1. Is it okay to multiply dummy with dummy?
2. Could you provide me example that compatible with state 12.0?

Best regards,

Canthy Meilanda

[Carlo Lazzaro]

Canthy:
1) yes, it is.
2) -fvvarlist- is not available in Stata 12 (please note that FAQ reccommends to declare which Stata release supports your analysis just to avoid providing the poster with unhelpful advice); see -help xi- instead.

[Canthy Meilanda]

Dear Carlo,

Thank you very much for the guidance ^^

Best regards,

Canthy Meilanda