Jmlr

Are p-values better better for detecting causal links than raw regression coefficients? Not always, we demonstrate in the paper Causal Structure Learning from Time Series, which is now on arXiv. [PDF]