Hi all,
After reading many many topics I have a question on the possibilities of any test to check whether the coefficients differ significantly or not.
I know about the -test- or -testparm- command but unfortunately can't find what I am looking for.
I test the effect of CEO pay scheme dependence on shareholder return (1=no, 2=yes) on the number of relative layoffs of the firm (percentage). Using paneldata, 5500 firms over 12 year:.
I want to split the effect of my main dependent for firms that made more or less profit than the year before, therefore a binairy variable (1= decreasing, 0=increasing).
xtreg RLO_per CeoPay#D_prof logsize logemployee tobinsq lev prof Domestic BoardSize i.n_ShhlType i.year i.n_Industry i.n_Country, vce(robust)
Now I want to test whether the coefficient of CeoPay=yes given decreasing profits (=0.2976) differs from the coefficient CeoPay=no given decreasing profits (=0.0683).
Anyone knows a proper method and stata command to do so?
Kinds regards,
Stan
Random-effects GLS regression Number of obs = 39,140
Group variable: ID Number of groups = 5,531
R-sq: Obs per group:
within = 0.0169 min = 1
between = 0.1029 avg = 7.1
overall = 0.0399 max = 14
Wald chi2(148) = .
corr(u_i, X) = 0 (assumed) Prob > chi2 = .
(Std. Err. adjusted for 5,531 clusters in ID)
---------------------------------------------------------------------------------------------------------------------------
| Robust
RLO_per | Coef. Std. Err. z P>|z| [95% Conf. Interval]
----------------------------------------------------------+----------------------------------------------------------------
CeoPay#D_prof |
1 1 | .0683347 .0302017 2.26 0.024 .0091404 .1275289
2 0 | .0298435 .0413767 0.72 0.471 -.0512534 .1109404
2 1 | .2976375 .0570798 5.21 0.000 .1857632 .4095118
|
logsize | .1262604 .0180784 6.98 0.000 .0908274 .1616933
logemployee | .0596924 .0114406 5.22 0.000 .0372692 .0821157
tobinsq | .0300837 .0107114 2.81 0.005 .0090897 .0510778
lev | .0175412 .1009887 0.17 0.862 -.180393 .2154755
prof | -2.063631 .2746226 -7.51 0.000 -2.601882 -1.525381
Domestic | -.023963 .0411957 -0.58 0.561 -.1047052 .0567792
BoardSize | .0036602 .0058319 0.63 0.530 -.0077701 .0150906
Related Posts with Wald test or any related for regression with interaction term
xi: xtivreg . Too many dummiesI am doing some work and have the following command for an instrumental varial regression with panel…
hbar and detection limitsHi all, Can someone assist with plotting an hbar for detection limits of different concentrations (…
Storing regression outputs / panel data / issue with system limits / percentage of positive constantsDear Stata users, I am working with panel data. I am running regressions for each id with: bysort …
summarize a variable in panel dataHello, I want to make a descriptive analysis on variables of my panel data. I have attached the dat…
table1_mc can now also report the value of the test statisticThe latest version of table1_mc is now available to download on SSC. Code: . search table1_mc The m…
Subscribe to:
Post Comments (Atom)
0 Response to Wald test or any related for regression with interaction term
Post a Comment