Difference between groups

[Theofanis Nikolopoulos]

I study tax and accounting practices in business. The dependent variable is var1 , the independent ones are var2, var3 and a number of control variables such as var4 and var5

But the whole analysis is based on how this dependent var1 differs between the 2 groups. I have devided the obs in 2 groups. Therefore all commands should be related between group 1 and group 2. So I have to split all the statistics analysis into 2 groups each time, because that's what I want to show.

Oh, and don't forget. I would like you to tell me how to do operations ( subtraction or addition ) on values of variables belonging to 2 or more time periods. For example, I want to examine whether there is a positive or negative change of var2 and var3 between 2000 and 2001 and whether this ( the difference )has an effect on in 2001.

[Carlo Lazzaro]

Theofanis:
there's no way to split your regression in two, as you can add one categorical predictor in the right-hand side of your regression equation to take the two groups in to account:

. use "https://www.stata-press.com/data/r17/nlswork.dta"
(National Longitudinal Survey of Young Women, 14-24 years old in 1968)

. xtset idcode year

Panel variable: idcode (unbalanced)
 Time variable: year, 68 to 88, but with gaps
         Delta: 1 unit

. xtreg ln_wage i.union c.age##c.age, fe vce(cluster idcode)

Fixed-effects (within) regression               Number of obs     =     19,229
Group variable: idcode                          Number of groups  =      4,150

R-squared:                                      Obs per group:
     Within  = 0.0978                                         min =          1
     Between = 0.0461                                         avg =        4.6
     Overall = 0.0587                                         max =         12

                                                F(3,4149)         =     225.93
corr(u_i, Xb) = 0.0183                          Prob > F          =     0.0000

                             (Std. err. adjusted for 4,150 clusters in idcode)
------------------------------------------------------------------------------
             |               Robust
     ln_wage | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
     1.union |   .1054165   .0098217    10.73   0.000     .0861607    .1246723
         age |   .0324229   .0051058     6.35   0.000     .0224129     .042433
             |
 c.age#c.age |  -.0002752   .0000818    -3.37   0.001    -.0004355   -.0001149
             |
       _cons |   .9942539   .0782432    12.71   0.000     .8408553    1.147652
-------------+----------------------------------------------------------------
     sigma_u |  .42293457
     sigma_e |  .26192667
         rho |  .72278248   (fraction of variance due to u_i)
------------------------------------------------------------------------------

As the regression above is in log-linear mode, others things being equal being member of the union increases -ln_wage- of about 11%:

. di exp(.1054165)-1
.11117332

The difference for -age- for each -idcode- across years can be calculated as follows:

. bysort idcode (year): gen wanted=age[_n+1]-age[_n]


. list idcode year age wanted if idcode==1

       +------------------------------+
       | idcode   year   age   wanted |
       |------------------------------|
    1. |      1     70    18        1 |
    2. |      1     71    19        1 |
    3. |      1     72    20        1 |
    4. |      1     73    21        2 |
    5. |      1     75    23        2 |
       |------------------------------|
    6. |      1     77    25        1 |
    7. |      1     78    26        2 |
    8. |      1     80    28        3 |
    9. |      1     83    31        2 |
   10. |      1     85    33        2 |
       |------------------------------|
   11. |      1     87    35        2 |
   12. |      1     88    37        . |
       +------------------------------+

.

[Theofanis Nikolopoulos]

Thank you Carlo!