multiple regression and finding more about it

[bunny ears]

hi all. i need help with my analysis.
i have multiple IVs and one DV. one of the IVs is a political party.
the political party is coded with, for example, 1=CDU, 2=Die Grüne, .... (these are names of political parties in Germany, there is a total 6 of them).
the results suggest that political party has a great influence on my DV (p-value is lower than 0.05). How do I know which party (out of 6 of them) has the highest impact on my DV?

[Nick Cox]

Please cross-reference your thread on Reddit so that people interested in answering can see what has already been said.

[Carlo Lazzaro]

In addition to Nick's wise advice, I would say that https://www.statalist.org/forums/help#realnames applies here.
That said, why describing what concerns you instead of sharing what you typed and what Stata gave you back (as per FAQ again)? Thanks.

[William Salazar]

the political party is coded with, for example, 1=CDU, 2=Die Grüne, .... (these are names of political parties in Germany, there is a total 6 of them).

I suspect from what you say you might not be treating your political party IV as a factor variable (I could be wrong - it was hard to tell). If this is the case your setup might not make sense.

An example of what I think you are doing is below. In my example sector is a number between 15 and 60 and is just used to tell me which sector a company belongs to:

reg y_roa_w cr_w sector


------------------------------------------------------------------------------
     y_roa_w | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        cr_w |    1.20637   .0373656    32.29   0.000     1.133131    1.279609
      sector |   .0061549   .0031689     1.94   0.052    -.0000563    .0123661
       _cons |   2.930498   .1219199    24.04   0.000     2.691527    3.169469
------------------------------------------------------------------------------

This regression doesn't make a great deal of sense as sector is being treated as a continuous variable and the information is categorical.
You can see the difference when treating sector as a factor variable below - think of this as creating a dummy variable for each political party in your case:

 reg y_roa_w cr_w i.sector


------------------------------------------------------------------------------
     y_roa_w | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        cr_w |   1.093426   .0385329    28.38   0.000     1.017899    1.168953
             |
      sector |
         15  |    1.49516   .2177774     6.87   0.000     1.068303    1.922018
         20  |   2.080485   .1970163    10.56   0.000     1.694321     2.46665
         25  |   3.788437   .2049704    18.48   0.000     3.386682    4.190192
         30  |    4.32988   .2312023    18.73   0.000     3.876709    4.783052
         35  |   2.960722   .2396386    12.35   0.000     2.491015    3.430429
         40  |   4.517356   .3367595    13.41   0.000     3.857286    5.177427
         45  |    2.93624   .2221074    13.22   0.000     2.500896    3.371585
         50  |   1.727329    .249175     6.93   0.000      1.23893    2.215727
         55  |   .6960268   .2495895     2.79   0.005     .2068154    1.185238
         60  |   .8601114   .3312336     2.60   0.009     .2108724    1.509351
             |
       _cons |   .8665103   .1844657     4.70   0.000     .5049458    1.228075
------------------------------------------------------------------------------

Here the regression is making a dummy variable for each sector allowing for differences in the constant term. For differences in any slopes you would need to look at interactions.
For more on factor variables see https://www.stata.com/features/overv...tor-variables/