The DOI matches its Crossref record; 136 of the 136 references it cites were found in a registry or an index; 2 of the 72 sources its claims cite were checked for support.
Causal inference in statistics: An overview
Open accessAbstract
This review presents empirical researchers with recent advances in causal inference, and stresses the paradigmatic shifts that must be undertaken in moving from traditional statistical analysis to causal analysis of multivariate data. Special emphasis is placed on the assumptions that underly all causal inferences, the languages used in formulating those assumptions, the conditional nature of all causal and counterfactual claims, and the methods that have been developed for the assessment of such claims. These advances are illustrated using a general theory of causation based on the Structural Causal Model (SCM) described in Pearl (2000a), which subsumes and unifies other approaches to causation, and provides a coherent mathematical foundation for the analysis of causes and counterfactuals. In particular, the paper surveys the development of mathematical tools for inferring (from a combination of data and assumptions) answers to three types of causal queries: (1) queries about the effects of potential interventions, (also called “causal effects” or “policy evaluation”) (2) queries about probabilities of counterfactuals, (including assessment of “regret,” “attribution” or “causes of effects”) and (3) queries about direct and indirect effects (also known as “mediation”). Finally, the paper defines the formal and conceptual relationships between the structural and potential-outcome frameworks and presents tools for a symbiotic analysis that uses the strong features of both.
Overview
CoChat Summary
TL;DR
This survey argues that causal inference requires more than standard statistical association: it needs a formal language for causal assumptions and a mathematical machinery for combining those assumptions with data. Pearl presents the Structural Causal Model as a unifying framework that subsumes graphical, potential-outcome, and structural-equation approaches.
Key findings
- The paper distinguishes associational concepts (e.g., correlation, regression, conditional independence) from causal concepts (e.g., intervention, effect, confounding, attribution), arguing that causal claims cannot be defined from observed distributions alone.
- In Markovian structural models, the post-intervention distribution is identified by truncated factorization; for example, the paper shows that \(E(Y \mid do(x_0)) = E(Y \mid X=x_0)\) under the stated assumptions.
- The back-door criterion provides a graphical rule for choosing admissible adjustment sets; the paper states that if a set \(S\) blocks all back-door paths from \(X\) to \(Y\), then stratifying on \(S\) removes confounding bias.
- For noncompliance in clinical trials, the paper says treatment effects are generally nonidentifiable, but bounds can be derived; Balke and Pearl (1997) obtained closed-form bounds for 16 equivalent classes in the binary case, and the paper notes that these bounds can yield “significant and sometimes accurate” information.
Methods
- This is a conceptual review/survey, not an empirical study; it synthesizes advances in causal analysis “from the unifying perspective” of the Structural Causal Model.
- It develops the formalism using structural equations, causal diagrams, do-operator interventions, and counterfactual notation, and then maps these ideas to the potential-outcome framework.
- The analysis is mainly theoretical and illustrative, using examples such as mediation, instrumental variables, covariate adjustment/back-door identification, and noncompliance in clinical trials.
Limitations
- The paper states that counterfactual probabilities conditioned on outcomes “can in general [not] be identified from observational or even experimental studies” without additional conditions.
- It states that natural direct effects “cannot in general be identified or estimated, even with the help of ideal, controlled experiments” unless extra assumptions hold.
- Evident from the text: this is a survey and framework paper, so it does not report a new dataset, sample size, or empirical effect estimate.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
2 Valid · 1 Not valid · 53 Not checked · 133 Own results and interpretations
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“Note also that the truncated factorization formula is not restricted to interventions on a single variable; it is applicable to simultaneous or sequential interventions such as those invoked in the analysis of time varying treatment with time varying confounders (Robins, 1986; Arjas and Parner, 2004).”
What the sources say
- Supportedcites Robins, 1986 source
Section F describes treatments assigned at successive times, with assignment depending on past measured covariates. The G-Computation Algorithm multiplies conditional probabilities along treatment and covariate-history paths and sums across those paths to calculate outcomes under a hypothetical treatment. This supports the cited point that the calculation extends beyond an intervention on one variable, although the paper calls it the G-computation algorithm rather than the truncated factorization formula. (LLM, conf=0.87)
- Not checked — no full text of the sourcecites Arjas and Parner, 2004 source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“The selection criterion can be applied systematically to diagrams of any size and shape, thus freeing analysts from judging whether ' X is conditionally ignorable given S ,' a formidable mental task required in the potential-response framework (Rosenbaum and Rubin, 1983).”
What the sources say
- Supportedcites Rosenbaum and Rubin, 1983 source
The citation supports the sentence’s reference to conditional ignorability in the potential-response framework. Section 1-2 defines strongly ignorable treatment assignment using the independence of treatment assignment from potential responses given covariates. The diagrammatic selection criterion is the citing authors’ own contrast, not a method attributed to Rosenbaum and Rubin. (LLM, conf=0.88)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“A few but challenging algebraic steps would reveal that one can perform the summation over Z 2 to obtain”
- Also stated in: § 3.3. Coping with unmeasured confounders · plain
“which means that we need only adjust for Z 1 and Z 3 without ever measuring Z 2 .”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § 3.3. Coping with unmeasured confounders · plain
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The set { Z 3 } , however, is not sufficient for adjustment because, as explained above, it does not block the path X ← W 1 ← Z 1 → Z 3 ← Z 2 → W 2 → Y .”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
In the paper
“This expression coincides with Robins' (1987) G -computation formula, which was derived from a more complicated set of (counterfactual) assumptions.”
What the sources say
- Not checked — no full text of the sourcecites Robins, 1987 source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
114 more claims
Create a free account to see all 189 claims and to verify the 7 references still open.
- Probably not a substantial claim. 39
- It may state two separate claims as one. 39
- The quote only partly states this claim. 1
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source says what the claim attributes to it.
In the citing paper
“which is known as the truncated factorization formula , see for example, Corollary 1 in Pearl (2009).”
From Discrimination-Free Insurance Pricing (2022) · cites this paper as Pearl (2009)
In the citing paper
““observational data” (such as rfMRI) is in general not a robust and safe tool for inferring causality, compared with “interventional” studies (such as task-fMRI experiments) [63].”
From Functional connectomics from resting-state fMRI (2013) · cites this paper as [63]
Citations
Cites
The central role of the propensity score in observational studies for causal effects
ROSENBAUM, P. R., & RUBIN, D. B. (1983) · Biometrika · doi:10.1093/biomet/70.1.41 · Cited by 31,241 · cited 3×
“The selection criterion can be applied systematically to diagrams of any size and shape, thus freeing analysts from judging whether “X is conditionally ignorable given S,” a formidable mental task required in the potential-response framework (Rosenbaum and Rubin, 1983).”
BackgroundMethod(2001) · RePEc: Research Papers in Economics · Cited by 27,577
Structural Equations with Latent Variables.
Clogg, C. C., & Bollen, K. A. (1991) · Contemporary Sociology · doi:10.2307/2072165 · Cited by 17,442
Structural Equations with Latent Variables
Bollen, K. A. (1989) · Wiley series in probability and statistics · doi:10.1002/9781118619179 · Cited by 17,277 · cited 3×
“…identification has received considerable attention in econometrics (Hurwicz, 1950; Marschak, 1950; Koopmans, 1953) and social science (Duncan, 1975; Bollen, 1989), usually in linear parametric settings, were it reduces to asking whether some model parameter, β, has a unique solution in terms of…”
Causality: Models, Reasoning, and Inference
Pearl, J. (2009) · doi:10.1017/cbo9780511803161 · Cited by 11,653
Self-citationMediation in experimental and nonexperimental studies: New procedures and recommendations.
Shrout, P. E., & Bolger, N. (2002) · Psychological Methods · doi:10.1037/1082-989x.7.4.422 · Cited by 11,335 · cited 2×
“…) = ∑ z E(Y |x, z)[P (z|x′)− P (z|x)] (52) This expression deserves the label Mediation Formula, due to its pivotal role in mediation analysis (Imai et al., 2008), which has been a thorny issue in several sciences (Shrout and Bolger, 2002; MacKinnon et al., 2007; Mortensen et al., 2009).”
BackgroundEstimating causal effects of treatments in randomized and nonrandomized studies
Rubin, D. B. (1974) · Journal of Educational Psychology · doi:10.1037/h0037350 · Cited by 9,436 · cited 4×
“…equation models (SEM) used in economics and social science (Goldberger, 1973; Duncan, 1975), the potentialoutcome framework of Neyman (1923) and Rubin (1974), and the graphical models developed for probabilistic reasoning and causal analysis (Pearl, 1988; Lauritzen, 1996; Spirtes et al., 2000;…”
BackgroundMethodInfluentialMacKinnon, D. P., Fairchild, A. J., & Fritz, M. S. (2007) · Annual Review of Psychology · doi:10.1146/annurev.psych.58.110405.085542 · Cited by 5,688 · cited 2×
“…) = ∑ z E(Y |x, z)[P (z|x′)− P (z|x)] (52) This expression deserves the label Mediation Formula, due to its pivotal role in mediation analysis (Imai et al., 2008), which has been a thorny issue in several sciences (Shrout and Bolger, 2002; MacKinnon et al., 2007; Mortensen et al., 2009).”
BackgroundDiscrete Multivariate Analysis: Theory and Practice.
Beniger, J. R., Bishop, Y. M. M., Feinberg, S. E., Holland, P. W., Light, R. J., & Mosteller, F. (1975) · Contemporary Sociology · doi:10.2307/2063625 · Cited by 5,108
Recent Developments in the Econometrics of Program Evaluation
Imbens, G. W., & Wooldridge, J. M. (2009) · Journal of Economic Literature · doi:10.1257/jel.47.1.5 · Cited by 4,926
Research works printed but not linked: 11
- Glymour, M. and Greenland, S. (2008). Causal diagrams. In, Modern Epidemiology (K. Rothman, S. Greenland and T. Lash, eds.), 3rd ed. Lippincott Williams & Wilkins, Philadelphia, PA, 183–209.
- Hurwicz, L. (1950). Generalization of the concept of identification. In, Statistical Inference in Dynamic Economic Models (T. Koopmans, ed.). Cowles Commission, Monograph 10, Wiley, New York, 245–257.
- Koopmans, T. (1953). Identification problems in econometric model construction. In, Studies in Econometric Method (W. Hood and T. Koopmans, eds.). Wiley, New York, 27–48.
- Marschak, J. (1950). Statistical inference in economics. In, Statistical Inference in Dynamic Economic Models (T. Koopmans, ed.). Wiley, New York, 1–50. Cowles Commission for Research in Economics, Monograph 10.
- Pearl, J. (1988)., Probabilistic Reasoning in Intelligent Systems. Morgan Kaufmann, San Mateo, CA.
- Pearl, J. (2001). Direct and indirect effects. In, Proceedings of the Seventeenth Conference on Uncertainty in Artificial Intelligence. Morgan Kaufmann, San Francisco, CA, 411–420.
- Robins, J. (1989). The analysis of randomized and non-randomized aids treatment trials using a new approach to causal inference in longitudinal studies. In, Health Service Research Methodology: A Focus on AIDS (L. Sechrest, H. Freeman and A. Mulley, eds.). NCHSR, U.S. Public Health Service, Washington, D.C., 113–159.
- Wermuth, N. (1992). On block-recursive regression equations., Brazilian Journal of Probability and Statistics (with discussion) 6 1–56.
- Cole, P. (1997). Causality in epidemiology, health policy, and law., Journal of Marketing Research 27 10279–10285.
- Pearson, K., Lee, A. and Bramley-Moore, L. (1899). Genetic (reproductive) selection: Inheritance of fertility in man., Philosophical Transactions of the Royal Society A 73 534–539.
- Robertson, D. (1997). The common sense of cause in fact., Texas Law Review 75 1765–1800.
Cited by
Items 1,001–1,100 of 2,354
Trend Analysis of Civil Aviation Incidents Based on Causal Inference and Statistical Inference
He, P., & Sun, R. (2023) · Aerospace · doi:10.3390/aerospace10090822 · Cited by 7
Causal scientific explanations from machine learning
Buijsman, S. (2023) · Synthese · doi:10.1007/s11229-023-04429-3 · Cited by 7
“( Pearl, 2009, p.99) In order to get to these causal relations assumptions need to be made.”
InfluentialOlaguez-Gonzalez, J. M., Schaeffer, S. E., Bretón‐Deval, L., Alfaro-Ponce, M., & Chaírez, I. (2024) · Neural Computing and Applications · doi:10.1007/s00521-024-09458-8 · Cited by 9
“As pointed out by Pearl [27], achieving causal comprehension of a phenomenon requires assuming that no hidden confounders exist among inputs as well as a sensitivity analysis (that is, the proportionality between the magnitude of changes of the outcomes the inputs are altered and the magnitude of…”
Walker, A. R., Venetis, C. A., Opdahl, S., Chambers, G. M., Jorm, L. R., & Vajdic, C. M. (2024) · Human Reproduction · doi:10.1093/humrep/deae053 · Cited by 10
“( Pearl, 2009) Generally, when trying to answer a causal question, researchers must deal with two sources of error: random error and bias (Hernan and Robins, 2020).”
BackgroundCausal Representation Learning via Counterfactual Intervention
Li, X., Sun, S., & Feng, R. (2024) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v38i4.28108 · Cited by 9
Igoche, B. I., Matthew, O., Bednár, P., & GEGOV, A. E. (2024) · Journal of Computing Theories and Applications · doi:10.62411/jcta.10501 · Cited by 11
“When this happens, the validation process is not considered successful because there is something wrong with the SCM DAG ontological framework designed or the dataset distribution has issues [62] – [64], [93].”
MethodNassir, M., & Benoliel, P. S. (2024) · Educational Management Administration & Leadership · doi:10.1177/17411432241273748 · Cited by 12
“This model reverses the direction of selected structural paths to estimate the extent to which one construct is deemed to "cause" another (V1 → V2) or vice versa (V2 → V1) (Pearl, 2009).”
Through the lens of causal inference: Decisions and pitfalls of covariate selection
Chen, G., Cai, Z., & Taylor, P. A. (2024) · Aperture Neuro · doi:10.52294/001c.124817 · Cited by 12
Current Role of Artificial Intelligence in the Management of Esophageal Cancer
Mela, E., Tsapralis, D., Papakonstantinou, D., Sakarellos, P., Vergadis, C., Klontzas, M. E., Rouvelas, I., Tzortzakakis, A., & Schizas, D. (2025) · Journal of Clinical Medicine · doi:10.3390/jcm14061845 · Cited by 12
“Furthermore, causal inference offers a novel approach for investigating causality as opposed to simple correlation, hence enhancing the reliability of AI prediction in the medical field [61].”
VC-Mamba: Causal Mamba representation consistency for video implicit understanding
Hu, Y., Zhao, J., Qi, C., Qiang, Y., Zhao, J., & Pei, B. (2025) · Knowledge-Based Systems · doi:10.1016/j.knosys.2025.113437 · Cited by 7
Econometría de evaluación de impacto
Núñez, L. G. (2011) · Economía · doi:10.18800/economia.201101.004 · Cited by 10
From forager tracks to prey distributions: an application to tuna vessel monitoring systems (VMS)
Walker, E., Rivoirard, J., Gaspar, P., & Bez, N. (2014) · Ecological Applications · doi:10.1890/14-0193.1 · Cited by 7
“However these techniques exhibit statistical or associational links that cannot easily be turned into causal links (Pearl 2009) or that are not based on biological processes.”
BackgroundOn the non-robustness of essentially conditional information inequalities
Kaced, T., & Romashchenko, A. (2012) · IEEE Information Theory Workshop (ITW) · doi:10.1109/itw.2012.6404672 · Cited by 7
A theory of structural determination
Gallow, J. D. (2015) · Philosophical Studies · doi:10.1007/s11098-015-0474-5 · Cited by 9
On the Interpretation of Instrumental Variables in the Presence of Specification Errors: A Reply
Swamy, P. A. V. B., Hall, S., Tavlas, G. S., et al. (2017) · Econometrics · doi:10.3390/econometrics5030032 · Cited by 7
Event history analysis of longitudinal data: a methodological application to sport sponsorship
Jensen, J. A., & Turner, B. A. (2016) · Sport in Society · doi:10.1080/17430437.2016.1179728 · Cited by 8
“Moreover, such data are oftentimes a necessary condition for the researcher’s ability to make causal inferences (Pearl 2009).”
BackgroundPractical and Conceptual Considerations for the Primary Prevention of Asthma
Brunwasser, S. M., & Hartert, T. V. (2018) · Clinics in Chest Medicine · doi:10.1016/j.ccm.2018.10.002 · Cited by 8
Disruptive technologies: Present and future
Novaes, M. D. A., & Basu, A. (2019) · Elsevier eBooks · doi:10.1016/b978-0-12-814309-4.00014-8 · Cited by 9
Quantifying the Impact of Data Sharing on Outbreak Dynamics (QIDSOD)
Mietchen, D., & Li, J. (2020) · Research Ideas and Outcomes · doi:10.3897/rio.6.e54770 · Cited by 7
“For example, outbreak propagation could be better characterized, understood, predicted and communicated if precise demographic information and migration information were available for individuals near the epicenter (McDonald et al. 1992,Pastor-Satorras and Vespignani 2001).”
BackgroundConceptual and Historical Perspectives
Cloninger, S. C. (2020) · Cambridge University Press eBooks · doi:10.1017/9781108264822.004 · Cited by 8
Temporal Causal Modelling on Large Volume Enterprise Data
Mohan, R., Chaudhury, S., & Lall, B. (2021) · IEEE Transactions on Big Data · doi:10.1109/tbdata.2021.3053879 · Cited by 8
Levy, J., Lebeaux, R. M., Christensen, B. C., Tosteson, T. D., & Bryan, Y. F. (2021) · Regional Anesthesia & Pain Medicine · doi:10.1136/rapm-2020-101984 · Cited by 8
White, V. M., Avendano, S. A., Albert, L. A., Zgierska, A. E., Balles, C. J., & Zayas‐Cabán, G. (2021) · Drug and Alcohol Dependence · doi:10.1016/j.drugalcdep.2021.108915 · Cited by 10
Hou, J., Deng, J., Li, C., & Wang, Q. (2022) · Journal of Personalized Medicine · doi:10.3390/jpm12050742 · Cited by 8
Wittenstein, T., Leibovich, N., & Hilfinger, A. (2022) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1010183 · Cited by 6
Han, X., Cramer, S. R., & Zhang, N. (2022) · Journal of Neural Engineering · doi:10.1088/1741-2552/ac9d66 · Cited by 7
“One approach that may overcome the limitations of Pearson’s correlation is partial correlation, which computes the linear correlation between two regions while controlling for the time series of all of the other regions [36, 37].”
Towards Causality-Aware Inferring: A Sequential Discriminative Approach for Medical Diagnosis
Lin, J., Wang, K., Chen, Z., Liang, X., & Lin, L. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2023.3292363 · Cited by 10
“…there are many different methods proposed in recent years, which can be approximately categorized into two branches according to whether the causal relationship is modeled explicitly or implicitly, namely, the structural causal model (SCM) framework [13] and the potential outcome framework [6].”
BackgroundIoT Anomaly Detection Via Device Interaction Graph
Wang, J., Li, Z., Sun, M., Yuan, B., & Lui, J. C. S. (2023) · International Conference on Dependable Systems & Networks/Proceedings · doi:10.1109/dsn58367.2023.00053 · Cited by 15
“[88] to the stochastic functional equations [44], [89] in the last century.”
BackgroundResultInfluentialIntroducing diminutive causal structure into graph representation learning
Gao, H., Qiao, P., Jin, Y., et al. (2024) · Knowledge-Based Systems · doi:10.1016/j.knosys.2024.111592 · Cited by 7
“We also conduct a theoretical analysis of DCSGL based on the Structural Causal Model (SCM) [23, 24, 25], demonstrating its e ff ectiveness.”
MethodBayesian sequential monitoring design for two‐arm randomized clinical trials with noncompliance
Shen, W., Ning, J., & Yuan, Y. (2015) · Statistics in Medicine · doi:10.1002/sim.6474 · Cited by 7
“A comprehensive review has been given by Pearl [18].”
Inferring gene regulatory networks by an order independent algorithm using incomplete data sets
Aghdam, R., Ganjali, M., Niloofar, P., & Eslahchi, C. (2015) · Journal of Applied Statistics · doi:10.1080/02664763.2015.1079307 · Cited by 7
“For the following, suppose that the set of parents of X i is { X i 1 , . . . , X is i } , where s i denotes the number of parents of X i ( | Pa ( X i ) | = s i ) [50,54].”
BackgroundAdoption and Impact of Black Pepper Certification in India
Parvathi, P., & Waibel, H. (2015) · RePEc: Research Papers in Economics · doi:10.22004/ag.econ.210312 · Cited by 7
“But PSM helps to generate valid counterfactuals from a non-random sample (MEZZATESTA et al., 2013).”
MethodInfluentialTowards W2T Foundations: Interactive Granular Computing and Adaptive Judgement
Skowron, A., & Jankowski, A. W. (2016) · Springer eBooks · doi:10.1007/978-3-319-44198-6_3 · Cited by 7
Imam, N., Spelman, T. D., Johnson, S. A. S., et al. (2019) · Quality Management in Health Care · doi:10.1097/qmh.0000000000000201 · Cited by 7
Cognitive Identity Management: Synthetic Data, Risk and Trust
Yanushkevich, S. N., Stoica, A., Shmerko, P., Howells, W. G. J., Crockett, K. A., & Guest, R. (2020) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn48605.2020.9207385 · Cited by 7
“enabling these operations is known as probabilistic inference called machine reasoning [36], [37].”
BackgroundCox, L. A., & Popken, D. A. (2020) · Global Epidemiology · doi:10.1016/j.gloepi.2020.100033 · Cited by 8
“5 has not been found to have an unequivocal causal effect – or, in many studies, even a clear association – with changes in all-cause mortality risk [10,27]), which would be the hallmark of a genuine causal relationship between them [21].”
InfluentialWhen causation does not imply correlation
Kennaway, R. (2020) · Elsevier eBooks · doi:10.1016/b978-0-12-818948-1.00004-6 · Cited by 7
“The problem of deducing causal information from correlations in observational data is a substantial research area, the simple maxim that “correlation does not imply causation” having been superceded by methods such as those set out in [9, 14], and in shorter form in [10].”
MethodEvent Mining Driven Context-Aware Personal Food Preference Modelling
Pandey, V., Rostami, A., Nag, N., & Jain, R. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-68821-9_52 · Cited by 10
The Different Types of Contributions to Knowledge (in CER): All Needed, But Not All Recognised
DRAPER, S. W., & Maguire, J. (2022) · ACM Transactions on Computing Education · doi:10.1145/3487053 · Cited by 8
(2019) · Technologien für die intelligente Automation · doi:10.1007/978-3-662-59084-3 · Cited by 9
Causal Confirmation Measures: From Simpson’s Paradox to COVID-19
Lu, C. (2023) · Entropy · doi:10.3390/e25010143 · Cited by 7
“These measures include risk difference [12]: RD = P(y1|x1)− P(y1|x0), (1) relative risk difference or the risk ratio (like the likelihood ratio for medical tests): RR = P(y1|x1)/P(y1|x0), (2) and the probability of causation Pd (used by Rubin and Greenland [13]) or the probability of necessity PN (used by Pearl [3]).”
BackgroundMethodInfluentialÖzdamar, Ö., & Gıovanıs, E. (2016) · Economic Analysis and Policy · doi:10.1016/j.eap.2016.11.001 · Cited by 8
“DAGs were primarily developed in computer science by Judea Pearl (1988; 2000; 2009) and Spirtes et al. (2000).”
Towards a New Science of a Clinical Data IntelligencePreprint
Tresp, V., Zillner, S., Costa, M. J., et al. (2013) · arXiv (Cornell University) · doi:10.48550/arxiv.1311.4180 · Cited by 7
“Unfortunately the process is not fool proof; including only a subset of confounders might in rare cases increase bias [2]”
BackgroundChangpetch, P., Haughton, D., Le, M., Ly, S., Nguyen, P., & Thach, T. T. (2016) · SSRN Electronic Journal · doi:10.2139/ssrn.2732844 · Cited by 7
Gıovanıs, E. (2016) · Preprints.org · doi:10.20944/preprints201607.0007.v1 · Cited by 7
“Moreover, BN can be applied also to randomised experiments (Pearl, 2000, 2009; Spirtes et al., 2000).”
BackgroundThe Efficacy of Propensity Score Matching in Bias Reduction with Limited Sample Sizes
Howarter, S. (2015) · KU ScholarWorks (The University of Kansas) · Cited by 14
Bias in OLAP Queries: Detection, Explanation, and RemovalPreprint
Salimi, B., Gehrke, J. E., & Suciu, D. M. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1803.04562 · Cited by 8
“An algorithmic approach for finding the covariates was developed by Judea Pearl [7, 43], who assumes that we are given a causal DAG over the set of attributes, where an edge represents a potential cause-effect between attributes, then gives a formal criterion, called the back-door criterion, to identify the covariates.”
BackgroundMethodCausal network inference using biochemical kineticsPreprint
Oates, C. J., Dondelinger, F., Bayani, N., et al. (2014) · arXiv (Cornell University) · doi:10.48550/arxiv.1406.0063 · Cited by 7
Conterfactual Generative Zero-Shot Semantic SegmentationPreprint
Shen, F., Liu, J., & Hu, P. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.06360 · Cited by 13
“Previous causal inference work has demonstrated the shortcoming of statistical relationship [16].”
BackgroundMethodOmitted variable bias in machine learned causal models
Chernozhukov, V., Cinelli, C., Newey, W. K., Sharma, A. K., & Syrgkanis, V. (2021) · doi:10.47004/wp.cem.2021.4521 · Cited by 6
“…outcomes) conditional on a set of observed variables; or, equivalently, that the set of observed covariates satisfy the “backdoor” (or, more generally, adjustment) criterion [Rosenbaum and Rubin, 1983a, Pearl, 1995, 2009a, Angrist and Pischke, 2009, Shpitser et al., 2012, Imbens and Rubin, 2015].”
BackgroundMethodAlhumaidan, O. A., Alkhunein, S., Alakeel, S. A., Fallata, G., Alsukait, R. F., Binsheehah, R. H., Alnassar, R. F., & Alqaseer, H. M. (2024) · Discover Food · doi:10.1007/s44187-024-00192-7 · Cited by 9
“Adjusting this effect is essential when evaluating the relationship between independent and dependent variables or cause-and-effect connections [45, 46].”
Ma, S., & Jia, N. (2024) · Psychology Research and Behavior Management · doi:10.2147/prbm.s483231 · Cited by 8
“70,71 Initially, the DAG was constructed using the hill-climbing algorithm via the “hc()” function in the “bnlearn” package.”
MethodWhen no answer is better than a wrong answer: A causal perspective on batch effects
Bridgeford, E., Powell, M., Kiar, G., Noble, S., Chung, J., Panda, S., Lawrence, R., Xu, T., Milham, M. P., Caffo, B. S., et al. (2025) · Imaging Neuroscience · doi:10.1162/imag_a_00458 · Cited by 14
“Both of these characteristics are im-5 mediately problematic for subsequently deriving causal conclusions, as failure to account for the batch effect in some may limit the identifiability of other potential causal estimands we may wish to subsequently investigate [23–25].”
Predicting accrual success for better clinical trial resource allocation
Ma, S., Wang, Y., Wagner, J. E., Johnson, S., Pakhomov, S., & Aliferis, C. (2025) · Scientific Reports · doi:10.1038/s41598-025-88400-x · Cited by 7
Smith, H. S., Buchanan, J. P., Goranitis, I., IJzerman, M. J., Lavelle, T. A., Marshall, D. A., Regier, D. A., Ungar, W. J., Weymann, D., Wordsworth, S., et al. (2025) · Value in Health · doi:10.1016/j.jval.2025.04.2162 · Cited by 9
Enhancer: A Distribution-Aware Framework with Temporal-Relational Meta-Learning for Stock Prediction
Chen, W., Li, S., Wang, H., & Wang, T. (2025) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3711896.3736934 · Cited by 9
“We assume that the output is O , 𝑑𝑜 denotes do-calculas to intervene the distribution [39], the invariant part provides a sufficient predictive capacity, and our goal is to minimize the following objective: Additional loss terms will be designed to cut off the backdoor path from the variant…”
Overview of Challenges in Brain-Based Predictive Modeling: Toward Meaningful Predictive Insights
Komeyer, V., Nieto, N., Eickhoff, S. B., Raimondo, F., & Patil, K. R. (2025) · Biological Psychiatry · doi:10.1016/j.biopsych.2025.09.003 · Cited by 12
Causality from bottom to top: a survey
Weinberg, A. I., Premebida, C., & Faria, D. R. (2025) · Machine Learning · doi:10.1007/s10994-025-06855-5 · Cited by 24
“Through experiments and - under certain circumstances - natural experiments, researchers can isolate and manipulate proposed causal factors while controlling for other variables [85].”
Machine learning and statistical inference in microbial population genomics
Sheppard, S. K., Arning, N., Eyre, D. W., & Wilson, D. J. (2025) · Genome biology · doi:10.1186/s13059-025-03775-4 · Cited by 11
“GWAS is motivated by a desire to learn about the causal process underlying the data, and pains are taken to avoid artefactual signals of association, while recognizing that observational studies cannot prove causality (see, e.g., [93, 94]).”
Adaptive stream processing on edge devices through active inference
Sedlak, B., Pujol, V. C., Morichetta, A., Donta, P. K., & Dustdar, S. (2025) · Evolving Systems · doi:10.1007/s12530-025-09753-2 · Cited by 18
“This process of using existing beliefs (widely known as priors) to calculate the probability of related events is commonly known as Bayesian inference [18] and it allows agents to improve their understanding of the environment.”
Nassir, M., & Benoliel, P. S. (2025) · Social Psychology of Education · doi:10.1007/s11218-025-10162-1 · Cited by 18
“…model, in which the direct path from TL to CTE was constrained to zero; and (4) Model 4 that tested reverse causal - ity, which reversed the direction of selected structural paths to assess whether one construct is associated with changes in another (V1→V2) or vice versa (V2→V1) (Pearl, 2009).”
InfluentialNegative Control Falsification Tests for Instrumental Variable Designs
Danieli, O., Nevo, D., Walk, I., Weinstein, B., & Zeltzer, D. (2026) · American Economic Review · doi:10.1257/aer.20240636 · Cited by 28
Prior Information in Frequentist Research Designs: The Case of Neyman’s Sampling Theory
Kubiak, A. P., & Kawalec, P. (2022) · Journal for General Philosophy of Science · doi:10.1007/s10838-022-09600-x · Cited by 10
A Critical Review of Propensity Score Matching in Causal Inference
Yu, J., & Lee, W. (2022) · Journal of Health Informatics and Statistics · doi:10.21032/jhis.2022.47.s1.s9 · Cited by 6
Shi, J., Wu, Z., & Dempsey, W. (2022) · Biometrika · doi:10.1093/biomet/asac065 · Cited by 11
Causal Inference in Data Analysis with Applications to Fairness and Explanations
Roy, S., & Salimi, B. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-31414-8_3 · Cited by 11
Rahul, M. R., & Chiddarwar, S. S. (2023) · International Journal of Machine Learning and Cybernetics · doi:10.1007/s13042-023-01891-w · Cited by 6
“In these models, the operator do(⋅) is generally denoted as the fixing operator [36], the interventional distribution is defined as p(G|do(M = camel)) , the conditional distribution p(G|M = camel) reflects correlation in the observed data, whereas p(G|do(M = camel)) reflects causation.”
BackgroundKim, D. H., Kim, S., & Song, B. C. (2024) · IEEE Access · doi:10.1109/access.2024.3400150 · Cited by 7
Objective causal predictions from observational data
Cox, L. A. (2024) · Critical Reviews in Toxicology · doi:10.1080/10408444.2024.2399856 · Cited by 8
“( The notation “ P ( y j do ( x ), z )” is often used in causal artificial intelligence and other fields (Pearl 2009) to denote the conditional probability that Y ¼ y if an intervention sets X to x in a case with covariates Z ¼ z .)”
InfluentialCausal Inference in AI Based Decision Support: Beyond Correlation to Causation
Mannava, M. K. (2024) · International Conference on Ubiquitous Computing and Intelligent Information Systems (ICUIS) · doi:10.1109/icuis64676.2024.10866582 · Cited by 6
“Training on a balanced dataset adjusted for confounding factors increased accuracy from 78% to 85%, demonstrating the impact of addressing confounding variables [1].”
Semiparametric posterior corrections
Yiu, A., Fong, E., Holmes, C. C., & Rousseau, J. (2025) · Journal of the Royal Statistical Society Series B (Statistical Methodology) · doi:10.1093/jrsssb/qkaf005 · Cited by 25
A Semantic Generalization of Shannon’s Information Theory and Applications
Lu, C. (2025) · Entropy · doi:10.3390/e27050461 · Cited by 9
Nature Makes No Leaps: Building Continuous Location Embeddings with Satellite Imagery from the Web
Hao, X., Chen, W., Zou, X., & Liang, Y. (2025) · International World Wide Web Conference (WWW) · doi:10.1145/3696410.3714629 · Cited by 10
“Grounded in causal theory [44], we utilize the robust technique of backdoor adjustment, which enables us to ob-struct the backdoor path by estimating 𝑃 ( 𝐻 | 𝑑𝑜 ( 𝑉 )) , where 𝑑𝑜 (·) denotes the do-calculus operation. where 𝑒 𝑖 ∈ 𝐸 denotes the patches containing environmental information.”
Holmes, T. H., & Lewis, D. B. (2014) · Journal of Immunological Methods · doi:10.1016/j.jim.2014.08.001 · Cited by 6
“The sometimes difficult problem of demonstrating causality is central to model development, as defined above, and can require a combination of manipulative experimental design and/or specialized statistical methods of causal inference (Pearl, 2009) that will not be discussed further here.”
BackgroundYu, Y., Li, H., Sun, X., Su, P., Wang, T., Liu, Y., Yuan, Z., Liu, Y., & Xue, F. (2017) · BMC Medical Research Methodology · doi:10.1186/s12874-017-0449-7 · Cited by 7
“Tests for c-equivalence are fairly easy to perform through a necessary and sufficient condition [15, 16], and they can also be implemented by propensity score methods [17].”
MethodMorphine in the Setting of Acute Heart Failure: Do the Risks Outweigh the Benefits?
Caspi, O., & Aronson, D. (2020) · Cardiac failure review · doi:10.15420/cfr.2019.22 · Cited by 6
Mediation Analysis of Microbiome Data and Detection of Causality in Microbiome Studies
Xia, Y. (2021) · Physiology in health and disease · doi:10.1007/978-3-030-67951-4_16 · Cited by 9
COSY: COunterfactual SYntax for Cross-Lingual Understanding
Yu, S., Zhang, H., Niu, Y., Sun, Q., & Jiang, J. (2021) · Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2021.acl-long.48 · Cited by 6
“To this end, we propose a novel COunterfac-tual SYntax (COSY) method, inspired by causal inference (Roese, 1997; Pearl et al., 2009) and contrastive learning (He et al., 2020).”
MethodInfluentialCoulombe, J., Moodie, E. E. M., Platt, R. W., & Renoux, C. (2022) · The Annals of Applied Statistics · doi:10.1214/21-aoas1570 · Cited by 8
“We rely on the causal Markov condition (Pearl, 2009, pp. 110) to compute the joint intensity of a given monitoring path.”
MethodPatil, S. R., Kenia, N., & Gunatilake, H. (2021) · Energy Policy · doi:10.1016/j.enpol.2021.112427 · Cited by 6
A Qualitative Strategy for Fusion of Physics into Empirical Models for Process Anomaly Detection
Rashdan, A. Y. A., Abdel‐Khalik, H. S., Giraud, K. M., Cole, D. G., Farber, J. A., Clark, W. W., Alemu, A., Allen, M. C., Spangler, R. M., & Varuttamaseni, A. (2022) · Energies · doi:10.3390/en15155640 · Cited by 5
“Standard statistical techniques do not provide a sense of direction on the relationship between cause and effect [20,21].”
BackgroundYessimbet, K., Shepherd, T. G., Ossó, A., & Steiner, A. K. (2022) · Geophysical Research Letters · doi:10.1029/2022gl100895 · Cited by 6
“To test our hypothesis about the underlying causal mechanisms and quantify the relationships from the data, we use a causal network approach proposed by Kretchmer et al. (2021) and based on causal inference theory (Pearl, 2009a, 2009b).”
BackgroundMethodDeterminants of online professor reviews: an elaboration likelihood model perspective
Li, Y., Wang, X., & Slyke, C. V. (2022) · Internet Research · doi:10.1108/intr-11-2020-0627 · Cited by 9
Loi, M., Nappo, F., & Viganò, E. (2023) · Res Publica · doi:10.1007/s11158-023-09586-3 · Cited by 8
Evaluating the effect of SARS-CoV-2 spike mutations with a linear doubly robust learner
Wang, X., Hu, M., Liu, B., Xu, H., Jin, Y., Wang, B., Zhao, Y., Wu, J., Yue, J., & Ren, H. (2023) · Frontiers in Cellular and Infection Microbiology · doi:10.3389/fcimb.2023.1161445 · Cited by 7
“As a novel computing method, causal inference model enjoys broad prospects for applications (Pearl, 2009; Yao et al., 2021).”
BackgroundMethodScene Graph to Image Synthesis via Knowledge Consensus
Wu, Y., Wei, P., & Lin, L. (2023) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v37i3.25387 · Cited by 14
“Causal Inference Causal inference reasoning between cause and effect based on their dependencies has been comprehensively studied in (Spirtes et al. 2000; Pearl 2009).”
BackgroundVarieties of perceptual truth and their possible evolutionary roots
Edelman, S. (2014) · Psychonomic Bulletin & Review · doi:10.3758/s13423-014-0741-z · Cited by 6
Discriminatory Decision Policy Aware Classification
Mancuhan, K., & Clifton, C. (2012) · IEEE ... International Conference on Data Mining workshops · doi:10.1109/icdmw.2012.96 · Cited by 10
On the impact of trust on consumer willingness to purchase GM food: Evidence from a European survey
Rousselière, D., & Rousselière, S. (2010) · Revue d’études en Agriculture et Environnement · doi:10.3406/reae.2010.1936 · Cited by 8
“If we agree with Pearl (2009), a structural approach to estimating causality typically features stronger and untested assumptions.”
BackgroundHierarchical Graphical Bayesian Models in Psychology
Campitelli, G., & Macbeth, G. (2014) · Revista Colombiana de Estadística · doi:10.15446/rce.v37n2spe.47940 · Cited by 6
“Graphical models are powerful formal tools to represent dependencies and in-dependence among variables, to make probabilistic inferences of conditional probabilities, and to estimate the values of parameters (Jordan 2004, Koller, Friedman, Getoor & Taskar 2007, Pearl 2009).”
Causal Inference Based on the Analysis of Events of Relations for Non-stationary Variables
Yin, Y., & Yao, D. (2016) · Scientific Reports · doi:10.1038/srep29192 · Cited by 6
An Introduction to the Econometrics of Program Evaluation
Cerulli, G. (2015) · Advanced studies in theoretical and applied econometrics · doi:10.1007/978-3-662-46405-2_1 · Cited by 6
Inference network building and movements prediction based on analysis of induced dependencies
Guo, L., Zuo, W., & Peng, T. (2016) · IET Software · doi:10.1049/iet-sen.2015.0033 · Cited by 6
“Causal analysis [8] is aimed to infer not only the likelihood of events under static conditions, but also the dynamics of events under changing conditions.”
BackgroundMethodSocial Influence: From Contagion to a Richer Causal Understanding
Liotsiou, D., Moreau, L., & Halford, S. (2016) · Lecture notes in computer science · doi:10.1007/978-3-319-47874-6_9 · Cited by 8
“…bias to be removed, thus rendering the causal effect X on Y identifiable from non-experimental data, using the back-door adjustment formula [37,38]: Since the right-hand side of Equation 1 contains only probabilities which are estimable (e.g. by regression) from our observational,…”
BackgroundMethodImproving causal determination
Cox, L. A. (2019) · Global Epidemiology · doi:10.1016/j.gloepi.2019.100004 · Cited by 6
Regression analysis of unmeasured confounding
Knaeble, B., Osting, B., & Abramson, M. A. (2020) · Epidemiologic Methods · doi:10.1515/em-2019-0028 · Cited by 9
“Confounding bias can be eliminated, theoretically, using causal graphs to select an admissible set of covariates for adjustment (Pearl 2009), but this relies on a “web of assumptions” (Rosenbaum 1995).”
BackgroundMethodInfluentialLin, J., Lin, K.-H., Lee, W., & Lin, S. (2019) · Statistics in Medicine · doi:10.1002/sim.8401 · Cited by 6
An Invitation to System-wide Algorithmic Fairness
Cortés, E. C., & Ghosh, D. (2020) · Proceedings of the AAAI/ACM Conference on AI Ethics and Society · doi:10.1145/3375627.3375860 · Cited by 10
“The resulting model is shown in figure 3b, which provides us with the distribution of interest: P(τ |do(θ = θ ),do(Q0 = q0)), see [31], and [30].”
MethodAn examination of the effects of outsourcing ticket sales force management
Popp, N. K., Jensen, J. A., McEvoy, C. D., & Weiner, J. F. (2020) · International Journal of Sports Marketing and Sponsorship · doi:10.1108/ijsms-04-2019-0046 · Cited by 8
Consistent Second-Order Conic Integer Programming for Learning Bayesian Networks
Küçükyavuz, S., Shojaie, A., Manzour, H., et al. (2020) · PubMed · doi:10.48550/arxiv.2005.14346 · Cited by 6
“They also play a vital part in causal inference [37].”
BackgroundInfluential
Metadata
- Authors
- Pearl, J.
- Affiliations
- University of California, Los Angeles (US)
- Corresponding author
- Pearl, Judea
- Venue
- Statistics Surveys
- Publisher
- Institute of Mathematical Statistics
- Published
- 2009-01-01
- Volume
- 3
- Issue
- none
- Pages
- 96-146
- ISSN
- 1935-7516 (ISSN-L)
- OpenAlex
- W2143117649
- Cited by
- 2,336
- References
- 135
- Citations by year
- 2026: 184 · 2025: 273 · 2024: 271 · 2023: 245 · 2022: 215 · 2021: 208 · 2020: 243 · 2019: 140 · 2018: 96 · 2017: 77 · 2016: 82 · 2015: 82 · 2014: 63 · 2013: 65 · 2012: 46
- FWCI
- 25.47
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- Locations
- Statistics Surveys (publishedVersion, cc-by)
- Project Euclid (Cornell University) (submittedVersion)
- eScholarship (California Digital Library) (submittedVersion)
- Department of Statistics, UCLA
- Funders
- Office of Naval Research
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Advanced Causal Inference Techniques
- Physical Sciences › Computer Science › Artificial Intelligence › Bayesian Modeling and Causal Inference
- Social Sciences › Arts and Humanities › History and Philosophy of Science › Philosophy of Biology and Scientific Explanation
- Keywords
- causal inference, structural causal models, counterfactual reasoning, causal assumptions, causal effects, policy evaluation, intervention effects, regret, causal attribution, direct and indirect effects, mediation analysis
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.