Dear members,
I am conducting a study where the coefficient of interest is the one of an interaction term (dummy##continuous).
One of the control variables is "experience" (the years of experience of an individual in a certain industry, integers).
A version for "experience" is "no_experience", a dummy taking the value 1 if the individual has zero years of experience, and zero otherwise.
As you can see from the regressions below, the p-value of the coefficient of the interaction term goes up from 0.020 when I use the "c.experience" to 0.073 when I use "i.no_experience".
I was told (by someone much more experienced than I) that the change in the p-value was indicating an omitted variable.
I am taking this comment very seriously, and I have been thinking about this for days now, but I am puzzled - I cannot think of a variable that would be related to "no_experience" and not to "experience".
I would really use any comment or reference you could have on this one.
What am I missing?
Thank you very much in advance!
MM
I am conducting a study where the coefficient of interest is the one of an interaction term (dummy##continuous).
One of the control variables is "experience" (the years of experience of an individual in a certain industry, integers).
A version for "experience" is "no_experience", a dummy taking the value 1 if the individual has zero years of experience, and zero otherwise.
As you can see from the regressions below, the p-value of the coefficient of the interaction term goes up from 0.020 when I use the "c.experience" to 0.073 when I use "i.no_experience".
I was told (by someone much more experienced than I) that the change in the p-value was indicating an omitted variable.
I am taking this comment very seriously, and I have been thinking about this for days now, but I am puzzled - I cannot think of a variable that would be related to "no_experience" and not to "experience".
I would really use any comment or reference you could have on this one.
What am I missing?
Thank you very much in advance!
MM
Code:
probit part i.dummy##c.continuous c.experience time_elapsed popularity type
Iteration 0: log likelihood = -472.92546
Iteration 1: log likelihood = -375.12789
Iteration 2: log likelihood = -373.9209
Iteration 3: log likelihood = -373.91832
Iteration 4: log likelihood = -373.91832
Probit regression Number of obs = 741
LR chi2(13) = 198.01
Prob chi2 = 0.0000
Log likelihood = -373.91832 Pseudo R2 = 0.2094
-------------------------------------------------------------------------------
part | Coef. Std. Err. z P>|z| [95% Conf. Interval]
--------------+----------------------------------------------------------------
|
dummy | -3.329878 .9732082 -3.42 0.001 -5.237331 -1.422424
continous | -.1627043 .1041651 -1.56 0.118 -.3668642 .0414555
|
dummy# |
continuous | .2587861 .110819 2.34 0.020 .041585 .4759873
|
c.experience | .0937605 .0312805 3.00 0.003 .0324519 .1550691
time_elapsed | -.0532365 .0177352 -3.00 0.003 -.0879969 -.0184761
popularity | .0524593 .0395867 1.33 0.185 -.0251293 .1300479
|
type |
2 | -.8001533 .2560044 -3.13 0.002 -1.301913 -.2983939
4 | -.0078864 .1471351 -0.05 0.957 -.296266 .2804931
5 | -.3953362 .1735647 -2.28 0.023 -.7355167 -.0551556
6 | -.3139073 .3798108 -0.83 0.409 -1.058323 .4305081
7 | -.008509 .1919858 -0.04 0.965 -.3847943 .3677763
8 | -.2624637 .3092251 -0.85 0.396 -.8685337 .3436063
9 | .2764111 .2174467 1.27 0.204 -.1497766 .7025989
|
_cons | 1.903462 .8897393 2.14 0.032 .1596047 3.647319
-------------------------------------------------------------------------------
probit part i.dummy##c.continuous i.no_experience time_elapsed popularity type
Iteration 0: log likelihood = -472.92546
Iteration 1: log likelihood = -376.54574
Iteration 2: log likelihood = -375.98521
Iteration 3: log likelihood = -375.98374
Iteration 4: log likelihood = -375.98374
Probit regression Number of obs = 741
LR chi2(13) = 193.88
Prob chi2 = 0.0000
Log likelihood = -375.98374 Pseudo R2 = 0.2050
-------------------------------------------------------------------------------
part | Coef. Std. Err. z P>|z| [95% Conf. Interval]
--------------+----------------------------------------------------------------
dummy | -2.70438 .9308157 -2.91 0.004 -4.528745 -.8800147
continuous | -.1006177 .0993523 -1.01 0.311 -.2953445 .0941092
|
dummy# |
continuous | .1901668 .1062418 1.79 0.073 -.0180632 .3983968
|
i.no_exper | -.2857927 .1133559 -2.52 0.012 -.5079663 -.0636192
time elapsed | -.0519214 .0175974 -2.95 0.003 -.0864117 -.0174312
popularity | .057135 .0395783 1.44 0.149 -.0204371 .1347071
|
type |
2 | -.7549467 .2562922 -2.95 0.003 -1.25727 -.2526233
4 | -.0360474 .1470612 -0.25 0.806 -.324282 .2521873
5 | -.3904853 .1736325 -2.25 0.025 -.7307987 -.0501718
6 | -.3563329 .3817795 -0.93 0.351 -1.104607 .3919411
7 | -.0131869 .1910058 -0.07 0.945 -.3875513 .3611775
8 | -.2367377 .3053641 -0.78 0.438 -.8352404 .361765
9 | .2734384 .2158454 1.27 0.205 -.1496108 .6964876
|
_cons | 1.55703 .8543099 1.82 0.068 -.1173863 3.231447
-------------------------------------------------------------------------------