Panel data analysis

[Nour Ben Ouhida]

Hello,
I have panel data which consists of 248 firms and 11 years. I opted for FE estimation while including the vce(robust) command in my analysis since my models have autocorrelation and heteroscedasticity problems. The only problem is that I'm getting non-significant results for my analysis. p-value>0.1
what can I do to fix it? should I transform my independent variables?
I really appreciate any help you can provide.
This is the output that I had:

[Jared Greathouse]

Who cares if your results are significant? This is unimportant.

[Nour Ben Ouhida]

Thank you Jared for your answer
but during our econometric classes, we were told that if we don't have significant results then the model is not reliable

[Carlo Lazzaro]

Nour:
a peculiar teacher indeed!
The regression model and rekated coefficient is unreliable when you do niot take into account nuisances such as heteroskedasticity, autocorrelation, enodogeneity and te correct specification of the functional form of the regressand.
Results are what they are and those coefficients thta do not reach statistical significance are as informative as those significant counterparts.
In addition, you do not report the full (discouraged; please use CODE delimiters) screenshot of your -xtreg- outcome table. Therefore, interested listers can see only a part of the story.

[Jared Greathouse]

Whoever said this deserves their Ph.D stripped from them and belongs nowhere near an econometrics classroom, because they're unreliable.

Sometimes an empirical effect is small or non-existent. Your model doesn't have to find "significant" results, it's quite alright for there to be no effect. In all honesty, it's a terrible approach to research and is p-hacking of the highest order.

Please do, read the second paper I linked here.

[Nour Ben Ouhida]

Thank you carlo and Jared
I appreciate the explanation, it's all clearer now
Jared, I'll check the paper you recommended to understand better the statistical tests
and Carlo I didn't report the full screenshot because I didn't think the rest of the table was relevant to my question