On Measurement Bias in Causal Inference

By Judea Pearl. 2010

118 citations recorded by the source corpus.

This paper addresses the problem of measurement errors in causal inference and highlights several algebraic and graphical methods for eliminating systematic bias induced by such errors. In particulars, the paper discusses the control of partially observable confounders in parametric and non parametric models and the computational problem of obtaining bias-free effect estimates in such models.

Subfields: Artificial Intelligence, 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 Statistics and Probability

Citations in the corpus

Causal Effect Inference with Deep Latent-Variable Models (2017, neutral): “Consider for example the causal graphs in Figure 1: it’s well known [20, 15, 18, 31, 41] that it is often incorrect to treat the proxies X as if they are ordinary confounders, as this would induce bias.”
Deconfounding Reinforcement Learning in Observational Settings (2018, methodology): “However, using proxy variables to correctly recover causal effects requires strict mathematical assumptions (Louizos et al., 2017; Edwards et al., 2015; Kuroki and Pearl, 2014; Miao et al., 2018; Pearl, 2012; Wooldridge, 2009) and, in practice, we do not know whether or not the available data meets ”
Identifying Causal Effects With Proxy Variables of an Unmeasured Confounder (2016, background): “Pearl (2010) also suggested an approach to test H0 in the presence of an unobserved confounder, which rests on aknown error mechanism, i.e., conditional distribution of the proxy given the unobservedconfounder.”
The Causal Foundations of Structural Equation Modeling (2012, background): “3Causal relationships among latent variables are assessed by treating their indicators as noisy measurement of the formers (Bollen, 1989; Pearl, 2010c; Cai and Kuroki, 2008).”
Challenges of Using Text Classifiers for Causal Inference (2018, methodology): “Working from the derivation for matrix adjustment in binary models given by Pearl (2010), we identify the causal effect of A on Y (Figure 3c) as τ ME (Eq 3 in Figure 4.)”
Policy Evaluation with Latent Confounders via Optimal Balance (2019, methodology): “In particular, there is a variety of work that studies sufficient conditions for the identifiability of latent confounder models [6, 34, 37].”
Measurement bias and effect restoration in causal inference (2014, methodology): “For example, if we postulate a noninformative W , pr(w | u) = pr(w), and we find that W strongly depends on X , a contradiction arises and no effect estimate will emerge (Pearl, 2010).”
Estimation of Individual Treatment Effect in Latent Confounder Models via Adversarial Learning (2018, methodology): “where the second equality follows from the rules of do-calculus1 applied to the causal graph in Figure 1(b) [3].”
3. The Foundations of Causal Inference (2010, methodology): “In addition to providing causally sound estimates for mediation effects, the Mediation Formula also enables researchers to evaluate analytically the effectiveness of various parametric specifications relative to any assumed model (Imai et al. 2008; Pearl 2010a).”
De-Confounded Variational Encoder-Decoder for Logical Table-to-Text Generation (2021, methodology): “When cooperating with GPT-2, our model outperforms "GPT-TableGen" by 1.6, 2.2, and 5.2 points of BLEU-3, SP-Acc, and NLI-Acc scores, respectively on the LogicNLG dataset, and 0.2, 1.3, and 5.4 points on the Logic2Text dataset.”