Dear Statalist, I would like to know if it is possible to compare two different coefficients from two separate regressions in which the dep variable is the same and the sample of firms from the two regressions are also the same. The change is in the indep variables even though they are standardized. So, I have the first regression (see below) with variables intra and inter which are the expenditures of a group of firms at the sectoral level. Thereafter, I calculate the same variables but this time for other group of firms (second regression below) and name the indep variables as newintra and newinter. Both regressions are done over the same sample of firms. The coefficient of interest are cL.x#cL.intra3 from the first regression and cL.x#cL.newintra3 from the second one. You can see that the coefficient for cL.x#cL.intra3 is half from the coefficient for cL.x#cL.newintra3, and this latter coefficient (cL.x#cL.newintra3) is above the 95% confidence interval from the first coefficient (cL.x#cL.intra3). Can I use this fact to highlight the possibility of these two coefficients being statistically different? Or should I need to do a test for that? In such a case, how can this test be done?
Thanks in advance!
Thanks in advance!
Code:
HDFE Linear regression Number of obs = 37,172
Absorbing 2 HDFE groups F( 13, 31) = 3.93
Statistics robust to heteroskedasticity Prob > F = 0.0009
R-squared = 0.0514
Adj R-squared = -0.0828
Within R-sq. = 0.0004
Number of clusters (sectors) = 32 Root MSE = 3.8428
(Std. Err. adjusted for 32 clusters in sectors)
--------------------------------------------------------------------------------
| Robust
y | Coef. Std. Err. t P>|t| [95% Conf. Interval]
---------------+----------------------------------------------------------------
x |
L1. | .1281176 .2443026 0.52 0.604 -.3701408 .626376
|
intra1 |
L1. | .0511315 .0426808 1.20 0.240 -.0359165 .1381795
|
intra2 |
L1. | .0121387 .0502433 0.24 0.811 -.0903332 .1146106
|
intra3 |
L1. | -.0380906 .043149 -0.88 0.384 -.1260936 .0499124
|
inter1 |
L1. | .3097762 .2404039 1.29 0.207 -.1805308 .8000832
|
inter2 |
L1. | -.1675679 .0938318 -1.79 0.084 -.3589391 .0238032
|
inter3 |
L1. | .0384597 .0354421 1.09 0.286 -.0338249 .1107444
|
cL.x#cL.intra1 | .6758195 .3186679 2.12 0.042 .0258921 1.325747
|
cL.x#cL.intra2 | -.2850499 .2970101 -0.96 0.345 -.8908059 .3207062
|
cL.x#cL.intra3 | .4430368 .2070881 2.14 0.040 .0206778 .8653958
|
cL.x#cL.inter1 | -.6084978 .6207346 -0.98 0.335 -1.874494 .6574987
|
cL.x#cL.inter2 | .36404 .6675775 0.55 0.589 -.9974932 1.725573
|
cL.x#cL.inter3 | -.0302176 .0802173 -0.38 0.709 -.1938218 .1333866
|
_cons | -.1158469 .0083171 -13.93 0.000 -.1328097 -.0988842
--------------------------------------------------------------------------------
Code:
HDFE Linear regression Number of obs = 37,172
Absorbing 2 HDFE groups F( 13, 31) = 2.74
Statistics robust to heteroskedasticity Prob > F = 0.0105
R-squared = 0.0514
Adj R-squared = -0.0829
Within R-sq. = 0.0004
Number of clusters (sectors) = 32 Root MSE = 3.8430
(Std. Err. adjusted for 32 clusters in sectors)
-----------------------------------------------------------------------------------
| Robust
y | Coef. Std. Err. t P>|t| [95% Conf. Interval]
------------------+----------------------------------------------------------------
x |
L1. | .173518 .2929203 0.59 0.558 -.4238969 .770933
|
newintra1 |
L1. | .0240514 .0688895 0.35 0.729 -.1164498 .1645525
|
newintra2 |
L1. | -.0046079 .0469632 -0.10 0.922 -.10039 .0911742
|
newintra3 |
L1. | -.0679491 .0468925 -1.45 0.157 -.163587 .0276888
|
newinter1 |
L1. | -.0541769 .1812328 -0.30 0.767 -.4238035 .3154498
|
newinter2 |
L1. | .0506375 .1695744 0.30 0.767 -.2952118 .3964868
|
newinter3 |
L1. | .0619607 .164531 0.38 0.709 -.2736026 .3975239
|
cL.x#cL.newintra1 | -.1120985 .2421835 -0.46 0.647 -.606035 .3818379
|
cL.x#cL.newintra2 | -.6568034 .3075257 -2.14 0.041 -1.284006 -.0296005
|
cL.x#cL.newintra3 | .9389757 .3915355 2.40 0.023 .1404338 1.737518
|
cL.x#cL.newinter1 | .3128542 .4878765 0.64 0.526 -.6821763 1.307885
|
cL.x#cL.newinter2 | .1486688 .3420751 0.43 0.667 -.5489979 .8463354
|
cL.x#cL.newinter3 | -.6125896 .3287086 -1.86 0.072 -1.282995 .057816
|
_cons | -.1344938 .0116396 -11.55 0.000 -.1582329 -.1107547
-----------------------------------------------------------------------------------
Comment