Effects of Treatment on the Treated: Identification and Generalization

By Ilya Shpitser, Judea Pearl. 2009

83 citations recorded by the source corpus.

Many applications of causal analysis call for assessing, retrospectively, the effect of with-holding an action that has in fact been implemented. This counterfactual quantity, sometimes called "effect of treatment on the treated," (ETT) have been used to to evaluate educational programs, critic public policies, and justify individual decision making. In this paper we explore the conditions under which ETT can be estimated from (i.e., identified in) experimental and/or observational studies. We show that, when the action invokes a singleton variable, the conditions for ETT identification have simple characterizations in terms of causal diagrams. We further give a graphical characterization of the conditions under which the effects of multiple treatments on the treated can be identified, as well as ways in which the ETT estimand can be constructed from both interventional and observational distributions.

Subfields: Artificial Intelligence, Education, Statistics and Probability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Education

In Statistics and Probability

Citations in the corpus

Improving the accuracy of medical diagnosis with causal machine learning (2020, methodology): “To quantify the likelihood that a disease is causing the patient ’ s symptoms, we employ counter-factual inference 56 – 58 .”
Causal inference in statistics: An overview (2009, background): “Shpitser and Pearl (2009) have derived a complete characterization of those models in which ETT can be identified from either experimental or observa- tional studies.”
CAUSAL INFERENCE WITH A GRAPHICAL HIERARCHY OF INTERVENTIONS (2014, background): “Existing results on mediation analysis [24] and ETT identification [28] would be subsumed as special cases under this framework, but it would entail”
Causes of Effects and Effects of Causes (2015, background): “…the counterfactual Yx(u) in model M is defined as the solution for Y in the ‘‘surgically modified’’ submodel Mx. Galles and Pearl (1998) and Halpern (1998) have given a complete axiomatization of structural counterfactuals, embracing both recursive and nonrecursive models (see also Pearl 2009:chap.”
Efficient Evaluation of Natural Stochastic Policies in Offline Reinforcement Learning (2020, methodology): “In a departure from this, in this paper we consider the evaluation of natural stochastic policies, which may depend on the natural value of the action, that is, the treatment that is observed in the data without intervention (Muñoz and Van Der Laan, 2012; Shpitser and Pearl, 2012; Haneuse and Rotnit”
Generalizing Experimental Findings (2015, background): “On the connection between ATE and ETT, see (Shpitser and Pearl, 2009).”