Re: st: Stata post test (simple question)

[Richard Williams <Richard.A.Williams.5@nd.edu>]

At 11:47 PM 1/29/2004 +0000, Joao Pedro W. de Azevedo wrote:
> Dear Stata users,
>
> I have a very simple question.
>
> I would like to test the hypothesis of equality of all coefficients (except
> the intercept) using two or more different sub samples of my dataset.
>
> For example, I would like to compare the results of the following models:
>
>         Lnwage = cons + educ + exp + expsq + error (if inside==1)
>         Lnwage = cons + educ + exp + expsq + error (if inside==0)
Here are two UCLA FAQs on making comparisons across groups if they help any:

http://www.ats.ucla.edu/stat/stata/faq/compreg2.htm

http://www.ats.ucla.edu/stat/stata/faq/compreg3.htm

I know you said you did not want to compute new variables, but it is really 
not hard to do and life will be much simpler if you do.  Here is how you 
could do it with the xi command (I just made up some data with your 
variable names to make sure I got it right:)

. xi: reg Lnwage i.inside*educ i.inside*exp i.inside*expsq

i.inside          _Iinside_0-1        (naturally coded; _Iinside_0 omitted)
i.inside*educ     _IinsXeduc_#        (coded as above)
i.inside*exp      _IinsXexp_#         (coded as above)
i.inside*expsq    _IinsXexpsq_#       (coded as above)

      Source |       SS       df       MS              Number of obs =     100
-------------+------------------------------           F(  7,    92) =    0.34
       Model |  2.06640451     7  .295200644           Prob > F      =  0.9353
    Residual |  80.7333601    92  .877536523           R-squared     =  0.0250
-------------+------------------------------           Adj R-squared = -0.0492
       Total |  82.7997646    99  .836361259           Root MSE      =  .93677

------------------------------------------------------------------------------
      Lnwage |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
  _Iinside_1 |  (dropped)
        educ |   .0718497   .1402154     0.51   0.610    -.2066302    .3503296
_IinsXeduc_1 |  -.0836037   .1942447    -0.43   0.668    -.4693904    .3021831
  _Iinside_1 |   .0222896   .2546331     0.09   0.930    -.4834337    .5280129
         exp |  -.1128918   .1267362    -0.89   0.375    -.3646009    .1388172
 _IinsXexp_1 |   .2634762   .2252115     1.17   0.245    -.1838132    .7107656
  _Iinside_1 |  (dropped)
       expsq |    .025165    .096337     0.26   0.795    -.1661686    .2164986
_IinsXexps~1 |  -.0140251   .2183539    -0.06   0.949    -.4476947    .4196445
       _cons |  -.1139492   .1711175    -0.67   0.507    -.4538033    .2259048
------------------------------------------------------------------------------

. test  _IinsXeduc_1 _IinsXexp_1 _IinsXexpsq_1

 ( 1)  _IinsXeduc_1 = 0
 ( 2)  _IinsXexp_1 = 0
 ( 3)  _IinsXexpsq_1 = 0

       F(  3,    92) =    0.66
            Prob > F =    0.5815

Note that I am only testing the three interaction terms, so that allows the 
intercept to differ across groups.  You would conclude that the effects of 
educ, exp and expsq do not differ between groups.

If you had more than two groups, xi would compute more vars, and (I think) 
you would want to test all the ones that started with 
_IinsX.  Equivalently, you would not test  _Iinside_1 _Iinside_2 etc.

If the weird variable names created by xi drive you blind (they do me!) 
this is equivalent to (assuming inside is coded 0-1 and nothing else)

. gen inseduc = inside * educ

. gen insexp = inside * exp

. gen insexpsq = inside * expsq

. reg  Lnwage educ exp expsq inside inseduc insexp insexpsq

      Source |       SS       df       MS              Number of obs =     100
-------------+------------------------------           F(  7,    92) =    0.34
       Model |  2.06640451     7  .295200644           Prob > F      =  0.9353
    Residual |  80.7333601    92  .877536523           R-squared     =  0.0250
-------------+------------------------------           Adj R-squared = -0.0492
       Total |  82.7997646    99  .836361259           Root MSE      =  .93677

------------------------------------------------------------------------------
      Lnwage |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        educ |   .0718497   .1402154     0.51   0.610    -.2066302    .3503296
         exp |  -.1128918   .1267362    -0.89   0.375    -.3646009    .1388172
       expsq |    .025165    .096337     0.26   0.795    -.1661686    .2164986
      inside |   .0222896   .2546331     0.09   0.930    -.4834337    .5280129
     inseduc |  -.0836037   .1942447    -0.43   0.668    -.4693904    .3021831
      insexp |   .2634762   .2252115     1.17   0.245    -.1838132    .7107656
    insexpsq |  -.0140251   .2183539    -0.06   0.949    -.4476947    .4196445
       _cons |  -.1139492   .1711175    -0.67   0.507    -.4538033    .2259048
------------------------------------------------------------------------------

. test  inseduc insexp insexpsq

 ( 1)  inseduc = 0
 ( 2)  insexp = 0
 ( 3)  insexpsq = 0

       F(  3,    92) =    0.66
            Prob > F =    0.5815


*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/