Hello everybody,

I am a completely new user of STATA. I want to estimate the effects of pandemic prevention policy on international trade. ExWork1 is the extent of workplacement closure policies in exporting countries while ImStay1 is stay-at-home requirements in importing countries. These two variables lie in a unit interval [0, 1].

. ppmlhdfe Exportvalue ExWork1 ImStay1 , absorb(imp Month exp) cluster(imp#exp)
> nolog
warning: dependent variable takes very low values after standardizing (2.4417e-
> 08)
Converged in 8 iterations and 18 HDFE sub-iterations (tol = 1.0e-08)

HDFE PPML regression No. of obs = 6,864
Absorbing 3 HDFE groups Residual df = 571
Statistics robust to heteroskedasticity Wald chi2(2) = 14.37
Deviance = 2.63668e+11 Prob > chi2 = 0.0008
Log pseudolikelihood = -1.31834e+11 Pseudo R2 = 0.9423

Number of clusters (imp#exp)= 572
(Std. Err. adjusted for 572 clusters in imp#exp)
------------------------------------------------------------------------------
| Robust
Exportvalue | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ExWork1 | -.1307411 .0367221 -3.56 0.000 -.2027152 -.0587671
ImStay1 | -.0447394 .0262254 -1.71 0.088 -.0961402 .0066614
_cons | 21.09845 .058649 359.74 0.000 20.9835 21.2134
------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
Absorbed FE | Categories - Redundant = Num. Coefs |
-------------+---------------------------------------|
imp | 44 0 44 |
Month | 12 1 11 |
exp | 13 1 12 ?|
-----------------------------------------------------+
? = number of redundant parameters may be higher


Could anyone plese tell me can I use this result in my undergraduate dissertation? This result shows that the workplacement closure policies in exporting countries are negative with bilateral trade and stay-at-home requirements don't have any siginificant results. My supervisor didn't give me any idea so I don't know whether this result is feasible.

Does "Prob > chi2 = 0.0008" mean this result is not useful? How about value of 'z'? Why are they negative and what's the influence of them?

Thanks very much!