Control of Confounding in Causal Inference

金华,方积乾

DOI: https://doi.org/10.3969/j.issn.1001-9626.2001.03.016

2001-01-01

Abstract:Control of confounding is a difficult issue in the analysis of causal effects. In this paper, we focus on the problem of confounding control in estimation of the average effect in the population, and propose a method to identify sufficient subsets for control of confounding.

What problem does this paper attempt to address?

Confounding and Confounding Control

Shaolong Ren,Genming Zhao

DOI: https://doi.org/10.21203/rs.3.rs-2384549/v1

2022-01-01

Abstract:Abstract Currently, the classical, counterfactual, and structural confounding definitions, are simultaneously used in teaching or practice context of epidemiology. Despite the resemblance, the three definitions are different in many aspects. A unified theory of confounding and confounding control has not been established. At the same time, the distinction between several confounding-related concepts is not clear-cut. Inspired by the structural causal model, this paper puts forwards an improved confounding definition to unify previous confounding definitions. The improved definition defines confounding as the association between exposure and the risk factor for the outcome, which cannot be interpreted by the causal effect of exposure on the risk factor. In so doing, confounding control methods is derived naturally, and several confusing confounding-related concepts is distinguished clearly.
A Confounding Bridge Approach for Double Negative Control Inference on Causal Effects

Wang Miao,Xu Shi,Yilin Li,Tchetgen

DOI: https://doi.org/10.1080/24754269.2024.2390748

2024-01-01

Statistical Theory and Related Fields

Abstract:Unmeasured confounding is a key challenge for causal inference. In this paper, we establish a framework for unmeasured confounding adjustment with negative control variables. A negative control outcome is associated with the confounder but not causally affected by the exposure in view, and a negative control exposure is correlated with the primary exposure or the confounder but does not causally affect the outcome of interest. We introduce an outcome confounding bridge function that depicts the relationship between the confounding effects on the primary outcome and the negative control outcome, and we incorporate a negative control exposure to identify the bridge function and the average causal effect. We also consider the extension to the positive control setting by allowing for the nonzero causal effect of the primary exposure on the control outcome. We illustrate our approach with simulations and apply it to a study about the short-term effect of air pollution on mortality. Although a standard analysis shows a significant acute effect of PM2.5 on mortality, our analysis indicates that this effect may be confounded, and after double negative control adjustment, the effect is attenuated toward zero.
Control of confounding through secondary samples.

Li Yin,Rolf Sundberg,Xiaoqin Wang,Donald B Rubin

DOI: https://doi.org/10.1002/sim.2468

2006-01-01

Statistics in Medicine

Abstract:The control of confounding is essential in many statistical problems, especially in those that attempt to estimate exposure effects. In some cases, in addition to the 'primary' sample, there is another 'secondary' sample which, though having no direct information about the exposure effect, contains information about the confounding factors. The purpose of this article is to study the influence of the secondary sample on likelihood inference for the exposure effect. In particular, we investigate the interplay between the efficiency improvement and the possible bias introduced by the secondary sample as a function of the degree of confounding in the primary sample and the sizes of the primary and secondary samples. In the case of weak confounding, the secondary sample can only little improve estimation of the exposure effect, whereas with strong confounding the secondary sample can be much more useful. On the other hand, it will be more important to consider possible biasing effects in the latter case. For illustration, we use a formal example of a generalized linear model and a real example with sparse data from a case-control study of the association between gastric cancer and HM-CAP/Band 120. Copyright (c) 2006 John Wiley & Sons, Ltd.
Distinguishing Between Mediators and Confounders is Important for the Causal Inference in Observational Studies

Zhongheng Zhang

DOI: https://doi.org/10.21037/amj.2019.09.03

2019-01-01

AME Medical Journal

Abstract:Observational studies are common in clinical literature because they are less costly and easy to perform, as compared to the randomized controlled trials (RCT). The limitation of such kind of studies is their inability to draw a causal inference due to potential bias. However, there are many approaches being proposed in recent years for the causal inference in observational studies (1-3). Although these methods cannot replace RCTs for causal inference, analysis in their framework can help to make the causal evidence more reliable. In the following discuss, I would like to discuss some of these points by using a published article.
Identification and estimation of causal effects in the presence of confounded principal strata

Shanshan Luo,Wei Li,Wang Miao,Yangbo He

DOI: https://doi.org/10.48550/arXiv.2206.08228

2022-06-17

Abstract:The principal stratification has become a popular tool to address a broad class of causal inference questions, particularly in dealing with non-compliance and truncation-by-death problems. The causal effects within principal strata which are determined by joint potential values of the intermediate variable, also known as the principal causal effects, are often of interest in these studies. Analyses of principal causal effects from observed data in the literature mostly rely on ignorability of the treatment assignment, which requires practitioners to accurately measure as many as covariates so that all possible confounding sources are captured. However, collecting all potential confounders in observational studies is often difficult and costly, the ignorability assumption may thus be questionable. In this paper, by leveraging available negative controls that have been increasingly used to deal with uncontrolled confounding, we consider identification and estimation of causal effects when the treatment and principal strata are confounded by unobserved variables. Specifically, we show that the principal causal effects can be nonparametrically identified by invoking a pair of negative controls that are both required not to directly affect the outcome. We then relax this assumption and establish identification of principal causal effects under various semiparametric or parametric models. We also propose an estimation method of principal causal effects. Extensive simulation studies show good performance of the proposed approach and a real data application from the National Longitudinal Survey of Young Men is used for illustration.

Methodology
Causal inference using regression-based statistical control: Confusion in Econometrics

Fan Chao,Guang Yu

DOI: https://doi.org/10.2478/jdis-2023-0006

2023-02-01

Journal of Data and Information Science

Abstract:Abstract Regression is a widely used econometric tool in research. In observational studies, based on a number of assumptions, regression-based statistical control methods attempt to analyze the causation between treatment and outcome by adding control variables. However, this approach may not produce reliable estimates of causal effects. In addition to the shortcomings of the method, this lack of confidence is mainly related to ambiguous formulations in econometrics, such as the definition of selection bias, selection of core control variables, and method of testing for robustness. Within the framework of the causal models, we clarify the assumption of causal inference using regression-based statistical controls, as described in econometrics, and discuss how to select core control variables to satisfy this assumption and conduct robustness tests for regression estimates.

information science & library science
Confounding of three binary-variables counterfactual model

Jingwei Liu,Shuang Hu

DOI: https://doi.org/10.48550/arXiv.1108.1497

2011-08-06

Applications

Abstract:Confounding of three binary-variables counterfactual model is discussed in this paper. According to the effect between the control variable and the covariate variable, we investigate three counterfactual models: the control variable is independent of the covariate variable, the control variable has the effect on the covariate variable and the covariate variable affects the control variable. Using the ancillary information based on conditional independence hypotheses, the sufficient conditions to determine whether the covariate variable is an irrelevant factor or a confounder in each counterfactual model are obtained.
[A Structural Classification of Strategies for Confounding Control in Research Design].

Y N He,L,Q Y Cai,N Q Zhao,Y J Zheng

DOI: https://doi.org/10.3760/cma.j.issn.0254-6450.2018.07.025

2018-01-01

Abstract:Confounding affects the causal relation among the population. Depending on whether the confounders are known, measurable or measured, they can be divided into four categories. Based on Directed Acyclic Graphs, the strategies for confounding control can be classified as (1) the broken-confounding-path method, which can be further divided into single and dual broken paths, corresponding to exposure complete intervention, restriction and stratification, (2) and the reserved-confounding-path method, which can be further divided into incomplete exposure intervention (in instrumental variable design and non-perfect random control test), mediator method and matching method. Among them, random control test, instrumental variable design or Mendelian randomized design, mediator method can meet the requirements for controlling all four types of confounders, while the restriction, stratification and matching methods are only applicable to known, measurable and measured confounders. Identifying the mechanisms of confounding control is a prerequisite for obtaining correct causal effect estimates, which will be helpful in research design.
Causal Inference with High-dimensional Discrete Covariates

Zhenghao Zeng,Sivaraman Balakrishnan,Yanjun Han,Edward H. Kennedy

2024-05-05

Abstract:When estimating causal effects from observational studies, researchers often need to adjust for many covariates to deconfound the non-causal relationship between exposure and outcome, among which many covariates are discrete. The behavior of commonly used estimators in the presence of many discrete covariates is not well understood since their properties are often analyzed under structural assumptions including sparsity and smoothness, which do not apply in discrete settings. In this work, we study the estimation of causal effects in a model where the covariates required for confounding adjustment are discrete but high-dimensional, meaning the number of categories $d$ is comparable with or even larger than sample size $n$. Specifically, we show the mean squared error of commonly used regression, weighting and doubly robust estimators is bounded by $\frac{d^2}{n^2}+\frac{1}{n}$. We then prove the minimax lower bound for the average treatment effect is of order $\frac{d^2}{n^2 \log^2 n}+\frac{1}{n}$, which characterizes the fundamental difficulty of causal effect estimation in the high-dimensional discrete setting, and shows the estimators mentioned above are rate-optimal up to log-factors. We further consider additional structures that can be exploited, namely effect homogeneity and prior knowledge of the covariate distribution, and propose new estimators that enjoy faster convergence rates of order $\frac{d}{n^2} + \frac{1}{n}$, which achieve consistency in a broader regime. The results are illustrated empirically via simulation studies.

Statistics Theory,Methodology
A Comparison of Approaches to Control for Confounding Factors by Regression Models

Guan Xing,Chang-Yun Lin,Chao Xing

DOI: https://doi.org/10.1159/000332743

2011-01-01

Human Heredity

Abstract:A common technique to control for confounding factors in practice is by regression adjustment. There are various versions of regression modeling in the literature, and in this paper we considered four approaches often seen in genetic association studies. We carried out both analytical and simulation studies comparing the bias of effect size estimates and examining the test sizes under the null hypothesis of no association between an outcome and an exposure. Further, we compared the methods in a nonsynonymous genome-wide scan for plasma lipoprotein(a) levels using a dataset from the Dallas Heart Study. We found that a widely employed approach that models the covariate-adjusted outcome and the exposure leads to an infranominal test size and underestimation of the exposure effect size. In conclusion, we recommend either using multiple regression models or modeling the covariate-adjusted outcome and the covariate-adjusted exposure to control for confounding factors.
Coarsened confounding for causal effects: a large-sample framework

Debashis Ghosh,Lei Wang

2025-01-07

Abstract:There has been widespread use of causal inference methods for the rigorous analysis of observational studies and to identify policy evaluations. In this article, we consider coarsened exact matching, developed in Iacus et al. (2011). While they developed some statistical properties, in this article, we study the approach using asymptotics based on a superpopulation inferential framework. This methodology is generalized to what we termed as coarsened confounding, for which we propose two new algorithms. We develop asymptotic results for the average causal effect estimator as well as providing conditions for consistency. In addition, we provide an asymptotic justification for the variance formulae in Iacus et al. (2011). A bias correction technique is proposed, and we apply the proposed methodology to data from two well-known observational studi

Methodology,Statistics Theory,Applications,Machine Learning
Causal Inference with Complex Treatments: A Survey

Yingrong Wang,Haoxuan Li,Minqin Zhu,Anpeng Wu,Ruoxuan Xiong,Fei Wu,Kun Kuang

2024-07-19

Abstract:Causal inference plays an important role in explanatory analysis and decision making across various fields like statistics, marketing, health care, and education. Its main task is to estimate treatment effects and make intervention policies. Traditionally, most of the previous works typically focus on the binary treatment setting that there is only one treatment for a unit to adopt or not. However, in practice, the treatment can be much more complex, encompassing multi-valued, continuous, or bundle options. In this paper, we refer to these as complex treatments and systematically and comprehensively review the causal inference methods for addressing them. First, we formally revisit the problem definition, the basic assumptions, and their possible variations under specific conditions. Second, we sequentially review the related methods for multi-valued, continuous, and bundled treatment settings. In each situation, we tentatively divide the methods into two categories: those conforming to the unconfoundedness assumption and those violating it. Subsequently, we discuss the available datasets and open-source codes. Finally, we provide a brief summary of these works and suggest potential directions for future research.

Methodology,Machine Learning
Multiply robust causal inference with double-negative control adjustment for categorical unmeasured confounding

Xu Shi,Wang Miao,Jennifer C Nelson,Eric J Tchetgen Tchetgen

DOI: https://doi.org/10.1111/rssb.12361

Abstract:Unmeasured confounding is a threat to causal inference in observational studies. In recent years, the use of negative controls to mitigate unmeasured confounding has gained increasing recognition and popularity. Negative controls have a long-standing tradition in laboratory sciences and epidemiology to rule out non-causal explanations, although they have been used primarily for bias detection. Recently, Miao and colleagues have described sufficient conditions under which a pair of negative control exposure and outcome variables can be used to identify non-parametrically the average treatment effect (ATE) from observational data subject to uncontrolled confounding. We establish non-parametric identification of the ATE under weaker conditions in the case of categorical unmeasured confounding and negative control variables. We also provide a general semiparametric framework for obtaining inferences about the ATE while leveraging information about a possibly large number of measured covariates. In particular, we derive the semiparametric efficiency bound in the non-parametric model, and we propose multiply robust and locally efficient estimators when non-parametric estimation may not be feasible. We assess the finite sample performance of our methods in extensive simulation studies. Finally, we illustrate our methods with an application to the post-licensure surveillance of vaccine safety among children.
Confounding, Homogeneity and Collapsibility for Causal Effects in Epidemiologic Studies

Z Geng,JH Guo,TS Lau,WK Fung

IF: 1.4

2001-01-01

Statistica Sinica

Abstract:Detection of confounding and confounders is important for observational studies, and especially so for epidemiologic studies. Miettinen and Cook (1981) derived two criteria for detecting confounders. Using a model, Wickramaratne and Holford (1987) proved that the two criteria are necessary but not sufficient conditions for confounders. We take uniform nonconfounding to mean there is no confounding at a coarse-subpopulation-level obtained by pooling any number of subpopulations. We discuss the necessity and sufficiency of the two criteria for uniform nonconfounding. The concepts of homogeneity and collapsibility for causal effects are also defined, and the relation among confounding, homogeneity and collapsibility is discussed. We show that the common causal effect over all fine subpopulations is just the causal effect of the whole population. The presence of confounding in an epidemiologic study hinders the estimation of causal effects. Miettinen and Cook (1981) derived two criteria of a confounder: (a) it must be predictive of risk in the unexposed population; and (b) it must have different distributions between the exposed and unexposed pop- ulations. Greenland and Robins (1986) discussed a logical connection among identifia- bility, exchangeability and confounding. Using a model, Wickramaratne and Holford (1987) showed that the above criteria are necessary but not sufficient conditions for confounding. Based on the counterfactual model (Rubin (1974) and Holland (1986)), Holland (1989) discussed causal effects and confounding in epidemiologic studies. Holland and Rubin (1988) defined causal effects for three different levels: unit-level, subpopulation-level and population-level. The population-level causal effect is measured by comparing the proportion of dis- eased in the exposed population with the proportion of diseased in the same
Detecting and Measuring Confounding Using Causal Mechanism Shifts

Abbavaram Gowtham Reddy,Vineeth N Balasubramanian

2024-09-26

Abstract:Detecting and measuring confounding effects from data is a key challenge in causal inference. Existing methods frequently assume causal sufficiency, disregarding the presence of unobserved confounding variables. Causal sufficiency is both unrealistic and empirically untestable. Additionally, existing methods make strong parametric assumptions about the underlying causal generative process to guarantee the identifiability of confounding variables. Relaxing the causal sufficiency and parametric assumptions and leveraging recent advancements in causal discovery and confounding analysis with non-i.i.d. data, we propose a comprehensive approach for detecting and measuring confounding. We consider various definitions of confounding and introduce tailored methodologies to achieve three objectives: (i) detecting and measuring confounding among a set of variables, (ii) separating observed and unobserved confounding effects, and (iii) understanding the relative strengths of confounding bias between different sets of variables. We present useful properties of a confounding measure and present measures that satisfy those properties. Empirical results support the theoretical analysis.

Artificial Intelligence
Toward Controlling Hidden Confounding for Causal Effect Estimation with Networked Observational Data.

Ruocheng Guo,Jundong Li,Huan Liu

2019-01-01

Abstract:The convenient access to observational data enables learning causal effects draws without randomized experiments, which draws increasing attention in many research areas such as economics, healthcare, and education. For example, we can study how a medicine (treatment) would affect the health condition (outcome) of a certain patient with electronic health records. To validate causal inference from such data, we need to control confounding bias - the influence of variables which causally influence both the treatment and the outcome. Existing work overwhelmingly relies on the unconfoundedness assumption that there are all confounders can be measured. However, this assumption can be untenable. In fact, an important fact ignored by the majority of existing work is that observational data can come with network information that can be utilized to infer hidden confounders. For example, in an observational study of the individual treatment effect of a medicine, instead of randomized experiments, the medicine is assigned to individuals based on a series of factors. Some factors (e.g., socioeconomic status) are challenging to measure and therefore become hidden confounders. Fortunately, the socioeconomic status of an individual can be reflected by whom she is connected in social networks. With this fact in mind, we aim to exploit the network structure to recognize patterns of hidden confounders for the sake of learning causal effects from observational data. In this work, we propose a novel causal inference framework, the network deconfounder, which learns representations to unravel patterns of hidden confounders from the network structure. Empirically, we perform extensive experiments to validate the effectiveness of the network deconfounder on various datasets.
Confounding caused by causal-effect covariability

Anders Ledberg

DOI: https://doi.org/10.48550/arXiv.1805.06035

2018-05-16

Abstract:Confounding seriously impairs our ability to learn about causal relations from observational data. Confounding can be defined as a statistical association between two variables due to inputs from a common source (the confounder). For example, if $Z\rightarrow Y$ and $Z\rightarrow X$, then $X$ and $Y$ will be statistically dependent, even if there are no causal connections between the two. There are several approaches available to adjust for confounding, i.e. to remove, or reduce, the association between two variables due to the confounder. Common adjustment techniques include stratifying the analysis on the confounder, and including confounders as covariates in regression models. Most adjustments rely on the assumption that the causal effects of confounders, on different variables, do not co-vary. For example, if the causal effect of $Z$ on $X$ and the causal effect of $Z$ on $Y$ co-vary between observational units, a confounding effect remains after adjustment for $Z$. This causal-effect covariability and its consequences is the topic of this paper. Causal-effect covariability is first explicated using the framework of structural causal models. Using this framework it is easy to show that causal-effect covariability generally leads to confounding that cannot be adjusted for by standard methods. Evidence from data indicates that the confounding introduced by causal-effect covariability might be a real concern in applied work.

Methodology
Identification of Causal Effects with Latent Confounding and Classical Additive Errors in Treatment

Wei Li,Zhichao Jiang,Zhi Geng,Xiao-Hua Zhou

DOI: https://doi.org/10.1002/bimj.201700048

IF: 1.715

2018-01-01

Biometrical Journal

Abstract:In this paper, we discuss the identifiability and estimation of causal effects of a continuous treatment on a binary response when the treatment is measured with errors and there exists a latent categorical confounder associated with both treatment and response. Under some widely used parametric models, we first discuss the identifiability of the causal effects and then propose an approach for estimation and inference. Our approach can eliminate the biases induced by latent confounding and measurement errors by using only a single instrumental variable. Based on the identification results, we give guidelines for determining the existence of a latent categorical confounder and for selecting the number of levels of the latent confounder. We apply the proposed approach to a data set from the Framingham Heart Study to evaluate the effect of the systolic blood pressure on the coronary heart disease.
Treatment Effect Estimation with Collider Bias and Confounding Bias

Baohong Li,Kun Kuang,Ruoxuan Xiong,Fei Wu

2023-01-01

Abstract:To answer causal questions from observational data, it is important to consider the mechanisms that determine which data values are observed and which are missing. Prior work has considered the treatment assignment mechanism and proposed methods to remove the confounding bias from the common causes of treatment and outcome. However, there are other issues in sample selection, commonly overlooked in prior work, that can bias the treatment effect estimation, such as the issue of censored outcome as a form of collider bias. In this paper, we propose the novel Selection Controlled CounterFactual Regression (SC-CFR) to simultaneously address confounding and collider bias. Specifically, we first calculate the magnitude of the collider bias of different instances by estimating the selection model and then add a control term to remove the collider bias while learning a balanced representation to remove the confounding bias when estimating the outcome model. Our method is shown to provide unbiased treatment effect estimates from observational data with confounding and collider bias. Extensive empirical results on both synthetic and real-world datasets show that our method consistently outperforms benchmarks when both types of biases exist.
A robust and efficient approach to causal inference based on sparse sufficient dimension reduction

Shujie Ma,Liping Zhu,Zhiwei Zhang,Chih-Ling Tsai,Raymond J. Carroll

DOI: https://doi.org/10.1214/18-AOS1722

2019-01-01

Abstract:A fundamental assumption used in causal inference with observational data is that treatment assignment is ignorable given measured confounding variables. This assumption of no missing confounders is plausible if a large number of baseline covariates are included in the analysis, as we often have no prior knowledge of which variables can be important confounders. Thus, estimation of treatment effects with a large number of covariates has received considerable attention in recent years. Most existing methods require specifying certain parametric models involving the outcome, treatment and confounding variables, and employ a variable selection procedure to identify confounders. However, selection of a proper set of confounders depends on correct specification of the working models. The bias due to model mis-specification and incorrect selection of confounding variables can yield misleading results. We propose a robust and efficient approach for inference about the average treatment effect via a flexible modeling strategy incorporating penalized variable selection. Specifically, we consider an estimator constructed based on an efficient influence function that involves a propensity score and an outcome regression. We then propose a new sparse sufficient dimension reduction method to estimate these two functions without making restrictive parametric modeling assumptions. The proposed estimator of the average treatment effect is asymptotically normal and semiparametrically efficient without the need for variable selection consistency. The proposed methods are illustrated via simulation studies and a biomedical application.

Control of Confounding in Causal Inference

Confounding and Confounding Control

A Confounding Bridge Approach for Double Negative Control Inference on Causal Effects

Control of confounding through secondary samples.

Distinguishing Between Mediators and Confounders is Important for the Causal Inference in Observational Studies

Identification and estimation of causal effects in the presence of confounded principal strata

Causal inference using regression-based statistical control: Confusion in Econometrics

Confounding of three binary-variables counterfactual model

[A Structural Classification of Strategies for Confounding Control in Research Design].

Causal Inference with High-dimensional Discrete Covariates

A Comparison of Approaches to Control for Confounding Factors by Regression Models

Coarsened confounding for causal effects: a large-sample framework

Causal Inference with Complex Treatments: A Survey

Multiply robust causal inference with double-negative control adjustment for categorical unmeasured confounding

Confounding, Homogeneity and Collapsibility for Causal Effects in Epidemiologic Studies

Detecting and Measuring Confounding Using Causal Mechanism Shifts

Toward Controlling Hidden Confounding for Causal Effect Estimation with Networked Observational Data.

Confounding caused by causal-effect covariability

Identification of Causal Effects with Latent Confounding and Classical Additive Errors in Treatment

Treatment Effect Estimation with Collider Bias and Confounding Bias

A robust and efficient approach to causal inference based on sparse sufficient dimension reduction