Omitted Labels in Causality: A Study of Paradoxes

We explore what we call ``omitted label contexts,'' in which training data is limited to a subset of the possible labels. This setting is common among specialized human experts or specific focused studies. We lean on well-studied paradoxes (Simpson's and Condorcet) to illustrate the more general difficulties of causal inference in omitted label contexts. Contrary to the fundamental principles on which much of causal inference is built, we show that ``correct'' adjustments sometimes require non-exchangeable treatment and control groups. These …

call causal inference causality cs.ai cs.it cs.lg cs.si data experts explore general human inference labels lean math.it stat.me studies study training training data