PLR m model residuals not normally distributed

[RashidBakirov]

I am using Light GBM for both m and l models. When analysed the residuals of both, I found that the residuals of m are not normally distributed (see the image). l residuals are normal. Would this invalidate the results? I also am interested to learn how exactly is the p-value and the confidence intervals are calculated, if you could refer me to any documentation on this, it would be much appreciated.

[SvenKlaassen]

@RashidBakirov it should not matter if the residuals are not normally distributed. The p-values and CIs are based on asymptotic normality of the estimator (based on an CLT). You can take a look at the DoubleML Documentation or the original Double ML Paper for more information on the CIs.