Ancestor regression in linear structural equation models

Abstract

We present a new method for causal discovery in linear structural equation models. We propose a simple technique based on statistical testing in linear models that can distinguish between ancestors and non-ancestors of any given variable. Naturally, this approach can then be extended to estimating the causal order among all variables. We provide explicit error control for false causal discovery, at least asymptotically. This holds true even under Gaussianity, where other methods fail due to non-identifiable structures. These Type I error guarantees come at the cost of reduced power. Additionally, we provide an asymptotically valid goodness-of-fit p -value for assessing whether multivariate data stem from a linear structural equation model.