Association, Causation, And Marginal Structural Models

By James M. Robins. 1999

How researchers see this paper

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

Citations in the corpus

Estimating Counterfactual Treatment Outcomes over Time Through Adversarially Balanced Representations (2020, methodology): “Robins (1999) shows that in this case, the estimation of counterfactual treatment outcomes is unbiased.”
Automated versus Do-It-Yourself Methods for Causal Inference: Lessons Learned from a Data Analysis Competition (2017, methodology): “Similar to matching, the goal of weighting is to create a pseudo-population of controls that have a joint distribution of covariates that is similar to the joint distribution of covariates for the inferential group (Rosenbaum, 1987; Robins, 1999).”
A Distributional Approach for Causal Inference Using Propensity Scores (2006, methodology): “However, the model itself is susceptible to misspecification (see Rotnitzky, Scharfstein, Su, and Robins 2001, sec.”
Estimating the optimal dynamic antipsychotic treatment regime: Evidence from the sequential multiple assignment randomized CATIE Schizophrenia Study (2012, background): “Thus, owing to the familiarity of MSMs, the dynamic regime MSM is an appealing approach for estimating dynamic treatment regimes.”
Handling drop‐out in longitudinal studies (2004, background): “Methods described herein are also related to non-compliance and causal inference, in the sense that causal inference based on counterfactuals or potential outcomes can be viewed as a missing data problem [20–24].”
Imprisonment and Reoffending (2009, methodology): “However, when they apply methods recently developed by Robins and colleagues (Robins 1999; Hernán, Brumback, and Robins 2000; Robins, Hernán, and Brumback 2000) for causal inference with nonexperimental data, they find that the evidence of a seeming criminogenic effect of imprisonment is quite…”
DOES MARRIAGE REDUCE CRIME? A COUNTERFACTUAL APPROACH TO WITHIN‐INDIVIDUAL CAUSAL EFFECTS* (2006, background): “Controlling for endogenous time-varying confounders can also induce a correlation between the treatment and response even when no causal association exists (Robins, 1999).”
A Class of Unbiased Estimators of the Average Treatment Effect in Randomized Experiments (2013, background): “Formally, the estimator will generally be biased because E dYT 1h i ¼ E XN i¼1 Ti Y1i π1i XN i¼1 Ti f Xi; θið Þ π1i þ XN i¼1 f Xi; θið Þ " ¼ YT1 XN i¼1 Cov Ti; f Xi; θið Þ π1i ; ½17 and, likewise, 4 Interestingly (and perhaps unsurprisingly), cΔG is quite similar to the double robust (DR) estimator ”
Inference on breakdown frontiers (2017, background): “For discussions on this point for different measures of deviations or relaxations from independence in various settings, see Rotnitzky, Robins, and Scharfstein (1998), Robins (1999), Imbens (2003), Altonji, Elder, and Taber (2005, 2008), and Oster (2019).”
Sensitivity analysis for inverse probability weighting estimators via the percentile bootstrap (2017, background): “We refer the reader to Robins (1999), Scharfstein et al. (1999), Imbens (2003), Altonji et al. (2005), Hudgens and Halloran (2006), Vansteelandt et al. (2006), McCandless et al. (2007), VanderWeele and Arah (2011), Richardson et al. (2014) and Ding and VanderWeele (2016) for alternative sensitivity…”
Strategic networks and entrepreneurial ventures (2007, methodology): “Another approach attempts to minimize the problem of unobserved heterogeneity through selection on observables econometric techniques, such as propensity score matching (Dehejia and Wahba, 2003) and inverse probability of treatment weighting ( Robins, 1999 ) procedures.”
Q-LEARNING WITH CENSORED DATA (2012, unknown): “Orellana et al. considered dynamic regime marginal structural mean models [Robins (1999)].”
Bayesian Nonparametric Modeling for Causal Inference (2011, background): “The second propensity-score-based estimator is a straightforward extension (Imbens 2004; Kurth et al. 2006) of standard inverse-probability-of-treatment-weighted (IPTW) estimators (Rosenbaum 1987; Robins 1999).”
Bayesian Regression Tree Models for Causal Inference: Regularization, Confounding, and Heterogeneous Effects (with Discussion) (2020, unknown): “An unbiased effect estimate can be obtained by “correcting” the observed outcome for unmeasured confounding, namely, Y − [E(Y (z)|Z = z, x) − E(Y (z)|x)] (Robins, 1999; Brumback et al., 2004).”