interaction effect

[Marie Schroder]

Hello,

I am using stata for a seminar at the university and am considering an interaction effect. But I have trouble interpreting the results, because the coefficient of the independent variable creative is negative regarding the interaction effect, but positive in the normal regression of creative on social_eco_need_adressed. Furthermore I am not sure how to interpret the coefficient 0.117 of the interaction effect. I know that the results are not significant, but still I have to interpret them.
May you please help me?

Interaction effect of age_when_finished education and creativity on social_eco_need_adressed:

normal regression; dependent variable: social_eco_need adressed; independent variables: achievement_oriented & creative

[Carlo Lazzaro]

Marie:
t5he way you share your results does not enourage helpful replies, as screenshots (as usual deprecated on this forum) does not allow a Comprehensive reading of your results.
For the future, please use CODE delimiters to post what you typed and what Stata gave you back. Thanks.
All that said, the best way I 've learnt to check my understanding of the regression outcome, especially when interactions are part of it, is to invoke -predict- and then calculate them by hand for a handful of observations, as you can see from the following toy-example:

use "C:\Program Files (x86)\Stata15\ado\base\a\auto.dta"
. reg price c.mpg##c.turn

      Source |       SS           df       MS      Number of obs   =        74
-------------+----------------------------------   F(3, 70)        =      9.00
       Model |   176805116         3  58935038.5   Prob > F        =    0.0000
    Residual |   458260281        70  6546575.44   R-squared       =    0.2784
-------------+----------------------------------   Adj R-squared   =    0.2475
       Total |   635065396        73  8699525.97   Root MSE        =    2558.6

------------------------------------------------------------------------------
       price |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         mpg |   940.2851   514.5387     1.83   0.072    -85.93003      1966.5
        turn |   559.5406   271.8051     2.06   0.043      17.4424    1101.639
             |
c.mpg#c.turn |  -32.80499    13.9182    -2.36   0.021    -60.56396   -5.046014
             |
       _cons |  -8936.859   10714.57    -0.83   0.407     -30306.4    12432.68
------------------------------------------------------------------------------

. drop pred

. predict pred, xb

. list price mpg turn pred in 1/3

     +-------------------------------+
     | price   mpg   turn       pred |
     |-------------------------------|
  1. | 4,099    22     40    5262.65 |
  2. | 4,749    17     40   7122.222 |
  3. | 3,799    22     35   6073.495 |
     +-------------------------------+

. di -8936.859 + (940.2851*22)+ (559.5406*40) + (-32.80499*22*40)
5262.646

. di -8936.859 + (940.2851*17)+ (559.5406*40) + (-32.80499*17*40)
7122.2185

. di -8936.859 + (940.2851*22)+ (559.5406*35) + (-32.80499*22*35)
6073.4919

.

[Marcos Almeida]

Welcome to the Stata Forum/Statalist,

To interpret the interaction term, you may just use the formula found in any textbook on the matter.

That said, I believe the better approach is using - margins - and - marginsplot - command.