Estimating causal effects of treatments in randomized and nonrandomized studies
Open accessThe DOI matches its Crossref record; 10 of the 10 references it cites were found in a registry or an index.
Abstract
Presents a discussion of matching, randomization, random sampling, and other methods of controlling extraneous variation. The objective was to specify the benefits of randomization in estimating causal effects of treatments. It is concluded that randomization should be employed whenever possible but that the use of carefully controlled nonrandomized data to estimate causal effects is a reasonable and necessary procedure in many cases.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
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?
In the citing paper
“That work argues models may memorize variable names but lack knowledge of the underlying observational distributions ( P ( X,Y ) ) required to identify traps like confounding (Rubin, 1974).”
From Diagnosing and Mitigating Sycophancy and Skepticism in LLM Causal Judgment (2026) · cites this paper as Rubin, 1974
In the citing paper
“Specifically, the assumption that a subunit’s own output or outcomes are not affected by any other unit’s implementation assignment (what is called SUTVA (99)) is invalid, as interactions and synergistic effects underlie many of our implementation strategies.”
From An Overview of Research and Evaluation Designs for Dissemination and Implementation (2017) · cites this paper as [99]
Citations
Cites
Statistical Methods for Research Workers
(1945) · American Journal of Clinical Pathology · doi:10.1093/ajcp/15.11.552b · Cited by 10,880
(2001) · Empirical Direction in Design and Analysis · doi:10.4324/9781410600004-24 · Cited by 6,808
SCHEFFE, H. (1960) · Soil Science · doi:10.1097/00010694-196006000-00016 · Cited by 5,516
Testing Statistical Hypotheses
(2021) · Wiley Series in Probability and Statistics · doi:10.1002/9781119243830.ch12 · Cited by 5,106
Design and Analysis of Experiments
(2007) · Modern Engineering Statistics · doi:10.1002/9780470187548.ch12 · Cited by 1,424
Controlling Bias in Observational Studies: A Review
Cochran, W. G. (2006) · Matched Sampling for Causal Effects · doi:10.1017/cbo9780511810725.005 · Cited by 958 · cited 1×
“Recent work on methods of reducing bias in nonrandomized studies is summarized in Cochran and Rubin (1974).”
MethodThe University Group Diabetes Program. A further statistical analysis of the mortality findings
Cornfield, J. (1971) · JAMA · doi:10.1001/jama.217.12.1676 · Cited by 119
University Group Diabetes Program
Smelo, L. S. (1971) · JAMA: The Journal of the American Medical Association · doi:10.1001/jama.1971.03180260057027 · Cited by 115
Design and Analysis of Industrial Experiments.
E., R. T., & Davies, O. L. (1954) · OR · doi:10.2307/3007149 · Cited by 111
Statistics: A Guide to the Unknown.
Pine, V. R., Tanur, J. M., Mosteller, F. C., Kruskal, W., Link, R. F., Pieters, R. S., & Rising, G. R. (1974) · Contemporary Sociology A Journal of Reviews · doi:10.2307/2062882 · Cited by 89
Cited by
Items 401–500 of 1,074
From Infrastructure to Culture: A/B Testing Challenges in Large Scale Social Networks
Xu, Y., Chen, N., Fernandez, A., Sinno, O., & Bhasin, A. (2015) · Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining · doi:10.1145/2783258.2788602 · Cited by 205
“In our discussion thus far, we have assumed the Rubin causal model [1], a standard machinery of testing framework, when conducting and analyzing A/B tests.”
MethodThe effects of quasi-random monetary experiments
Jordà, Ò., Schularick, M., & Taylor, A. M. (2020) · Journal of Monetary Economics · doi:10.1016/j.jmoneco.2019.01.021 · Cited by 205
“Using the potential outcomes notation (see, e.g., Rubin 1974) y1 refers to the potential outcome when there is an intervention and y0 is the potential outcome absent the intervention.”
Rubin, D. B. (2006) · Statistical Science · doi:10.1214/088342306000000114 · Cited by 206
“Maybe adding some simplifying assumptions would help? There are some assumptions that are relatively standard in similar settings, in particular, where the intermediate outcome variable indicates compliance with assigned drug (Angrist, Imbens and Rubin, 1996). In Tables 1–4, the four principal strata could be called c…”
InfluentialSelf-citationThe Blessings of Multiple Causes
Wang, Y., & Blei, D. M. (2019) · Journal of the American Statistical Association · doi:10.1080/01621459.2019.1686987 · Cited by 202
“We develop the deconfounder in the potential outcomes framework (Rubin 1974, 2005; Imbens and Rubin 2015).”
BackgroundMethodCampedelli, G. M., Aziani, A., & Favarin, S. (2021) · American Journal of Criminal Justice · doi:10.1007/s12103-020-09578-6 · Cited by 205
“While Pearl (2009) demonstrates that post-facto observational studies cannot provide evidence of causal inference due to the potential presence of confounding factors, several quasi-experimental alternatives have been proposed to overcome the difficulty of running RCT in certain scientific fields.”
BackgroundGoldberg, L. R. (2019) · Quantitative Finance · doi:10.1080/14697688.2019.1655928 · Cited by 206
“Many researchers rely on the Neyman (or Neyman–Rubin) potential outcomes model, which is discussed in Neyman (1923/1990), Rubin (1974, 2005) and Sekhon (2008).”
MethodEconometric Evaluation of Labour Market Policies
(2001) · ZEW Economic Studies · doi:10.1007/978-3-642-57615-7 · Cited by 205
Related Causal Frameworks for Surrogate Outcomes
Joffe, M. M., & Greene, T. (2009) · Biometrics · doi:10.1111/j.1541-0420.2008.01106.x · Cited by 203
“account of causality, based on potential outcomes (Rubin, 1974).”
BackgroundInfluentialMixing Methods: A Bayesian Approach
HUMPHREYS, M., & JACOBS, A. M. (2015) · American Political Science Review · doi:10.1017/s0003055415000453 · Cited by 198
“These types differ in their “potential outcomes”— that is on what outcomes, Y, they would take on under alternative treatment conditions, X (Rubin 1974).”
BackgroundMethod(2011) · Oxford University Press eBooks · doi:10.1093/acprof:oso/9780199574131.001.0001 · Cited by 199
Some methods for heterogeneous treatment effect estimation in high dimensions
Powers, S., Qian, J., Jung, K., Schuler, A., Shah, N. H., Hastie, T., & Tibshirani, R. (2018) · Statistics in Medicine · doi:10.1002/sim.7623 · Cited by 202
“τl = ∑S s=1 nsl ( Ȳ1sl − Ȳ0sl ) ∑S s=1 nsl , where Ȳtsl = ∑nl i=1 I{Ti=t∧si=s}Yi ntsl (3) is the mean response among the treatment (t = 1) or control (t = 0) group in stratum s and ntsl = ∑nl i=1 I{si=s} is the corresponding number of patients in leaf l for t ∈ {0, 1}, s ∈ {1, .”
BackgroundMethodThe impact of job uncertainty on first-birth postponement
Vignoli, D., Tocchioni, V., & Mattei, A. (2020) · Advances in Life Course Research · doi:10.1016/j.alcr.2019.100308 · Cited by 202
“From the statistical point of view, we adopt the potential outcome approach to causal inference (Balbo & Arpino, 2016; Imbens & Rubin, 2015; Rubin, 1974).”
MethodCausal mediation analysis with double machine learning
Farbmacher, H., Huber, M., Lafférs, L., Langen, H., & Spindler, M. (2022) · The Econometrics Journal · doi:10.1093/ectj/utac003 · Cited by 198
“We use the potential outcome framework, see for instance Rubin (1974), to define the direct and indirect effects of interest, see also Ten Have, Joffe, Lynch, Brown, Maisto, and Beck (2007) We denote the ATE by ∆ = E [ Y (1 , M (1)) − Y (0 , M (0))], which comprises both direct and indirect effects.”
MethodThe Effect of the Medicare Part D Prescription Benefit on Drug Utilization and Expenditures
Yin, W., Basu, A., Zhang, J. X., Rabbani, A., Meltzer, D. O., & Alexander, G. C. (2008) · Annals of Internal Medicine · doi:10.7326/0003-4819-148-3-200802050-00200 · Cited by 202
“Thus, we sought to estimate the effect of Part D on expenditures and utilization netting out contemporaneous changes to the outcome variable due to factors unrelated to Part D (15, 16).”
BackgroundIdentification of Causal Parameters in Randomized Studies With Mediating Variables
Sobel, M. E. (2008) · Journal of Educational and Behavioral Statistics · doi:10.3102/1076998607307239 · Cited by 202
“Building on Rubin’s model for causal inference (Rubin, 1974, 1977, 1978, 1980), Holland (1988) argued that this is incorrect.”
BackgroundMethodWhen Natural Experiments Are Neither Natural nor Experiments
SEKHON, J. S., & TITIUNIK, R. (2012) · American Political Science Review · doi:10.1017/s0003055411000542 · Cited by 201
The Impact of Gurus: Parker Grades and en Primeur Wine Prices
Ali, H. H., Lecocq, S., & Visser, M. (2008) · The Economic Journal · doi:10.1111/j.1468-0297.2008.02147.x · Cited by 200
Shimonovich, M., Pearce, A., Thomson, H., Keyes, K., & Katikireddi, S. V. (2021) · European Journal of Epidemiology · doi:10.1007/s10654-020-00703-7 · Cited by 195
Causation with a Human Face: Normative Theory and Descriptive Psychology
Woodward, J. (2021) · doi:10.1093/oso/9780197585412.001.0001 · Cited by 188
Li, F., Hughes, J. P., Hemming, K., Taljaard, M., Melnick, E. R., & Heagerty, P. J. (2021) · Statistical Methods in Medical Research · doi:10.1177/0962280220932962 · Cited by 196
Methods and tools for causal discovery and causal inference
Nogueira, A. R., Pugnana, A., Ruggieri, S., Pedreschi, D., & Gama, J. (2022) · WIREs Data Mining and Knowledge Discovery · doi:10.1002/widm.1449 · Cited by 196
“Introduced in Rubin (1974), it is very popular in economics and social sciences, and logically equivalent to SCM (Pearl, 2009).”
BackgroundDevelopments in Post-marketing Comparative Effectiveness Research
Schneeweiss, S. (2007) · Clinical Pharmacology & Therapeutics · doi:10.1038/sj.clpt.6100249 · Cited by 197
“The causal effect of a drug would ideally be assessed by administering the drug to a person and comparing this person’s experience with the counterfactual experience of what would have happened to the same person at the same time had the drug not been taken.(30) As such an experiment is not practical, research seeks t…”
BackgroundRemarks on the analysis of causal relationships in population research
Moffitt, R. (2005) · Demography · doi:10.1353/dem.2005.0006 · Cited by 197
“In statistics, this model is often associated with the work of Rubin (1974) and is generally called the Rubin causal model in that literature, but it is often said to have originated with the work of Neyman (1923, 1935).”
MethodThe Role of a Second Control Group in an Observational Study
Rosenbaum, P. R. (1987) · Statistical Science · doi:10.1214/ss/1177013232 · Cited by 195
ANCOVA Versus CHANGE From Baseline in Nonrandomized Studies: The Difference
van Breukelen, G. J. P. (2013) · Multivariate Behavioral Research · doi:10.1080/00273171.2013.831743 · Cited by 196
“…1990; D. T. Campbell & Kenny, 1999; Holland & Rubin, 1983; Kenny, 1975; Lord, 1967, 1969; Maris, 1998; Porter & Raudenbush, 1987; Rausch, Maxwell, & Kelley, 2003; Reichardt, 1979; Rubin, 1974, 1977; Senn, 2006; Van Breukelen, 2006; Wainer, 1991; Wainer & Brown, 2007; Weisberg, 1979; Wright, 2006).”
Voluntary environmental regulation in developing countries: Mexico’s Clean Industry Program
Blackman, A., Lahiri, B., Pizer, W., Rivera Planter, M., & Muñoz Piña, C. (2010) · Journal of Environmental Economics and Management · doi:10.1016/j.jeem.2010.05.006 · Cited by 196
Social influence in social advertising: evidence from field experiments
Bakshy, E., Eckles, D., Yan, R., & Rosenn, I. (2012) · Proceedings of the 13th ACM Conference on Electronic Commerce · doi:10.1145/2229012.2229027 · Cited by 196
Conscientious Classification: A Data Scientist's Guide to Discrimination-Aware Classification
d'Alessandro, B., O'Neil, C., & LaGatta, T. (2017) · Big Data · doi:10.1089/big.2016.0048 · Cited by 195
Hesser, H. (2015) · Internet Interventions · doi:10.1016/j.invent.2015.02.003 · Cited by 192
“, experimental manipulation), constant across conditions, the average difference between participants treated and controls is an unbiased estimate of the average casual effect of the treated (Rubin, 1974).”
BackgroundInferring the conservative causal core of gene regulatory networks
Altay, G., & Emmert-Streib, F. (2010) · BMC Systems Biology · doi:10.1186/1752-0509-4-132 · Cited by 192
“This problem has already received considerable interest and is, aside from biology, of the utmost interest in a series of fields like economics, epidemiology, medicine, sociology and statistics [12-15].”
BackgroundBrand, J. E., & Halaby, C. N. (2006) · Social Science Research · doi:10.1016/j.ssresearch.2005.06.006 · Cited by 193
“We parameterize the ‘‘effect’’ of attending an elite college by formally locating it within the potential outcome, counterfactual approach to causal inference pioneered by Rubin (Rosenbaum and Rubin, 1983; Rubin, 1974) and developed by others (e.g., Heckman, 1997).2 Within this framework there is more than one paramet…”
Difference-in-differences techniques for spatial data: Local autocorrelation and spatial interaction
Delgado, M. S., & Florax, R. J. G. M. (2015) · Economics Letters · doi:10.1016/j.econlet.2015.10.035 · Cited by 189
“At its core, a treatment effect is the difference between two potential outcomes, with potential outcomes being a function of treatment status (Rubin, 1974).”
BackgroundCommentary: Mediation Analysis, Causal Process, and Cross-Sectional Data
Shrout, P. E. (2011) · Multivariate Behavioral Research · doi:10.1080/00273171.2011.606718 · Cited by 190
“Holland (1986) and Rubin (1974) showed how betweenindividual information in randomized studies allows scientists to learn about within-individual causal process, and how causal processes can also be inferred from a careful analysis of within-individual variation.”
An axiomatic characterization of causal counterfactuals
Galles, D., & Pearl, J. (1998) · Foundations of Science · doi:10.1023/a:1009602825894 · Cited by 192
“…the formal equivalence of structural equation modeling,popular in economics and the social sciences [Goldberger, 1992], and the potential-responseframework, as used in statistics [Rubin, 1974; Holland, 1986; Robins, 1986].11 In nonrecur-sive models, however, this is not necessarily the case.”
BackgroundMethodInfluentialA Selective Review of Negative Control Methods in Epidemiology
Shi, X., Miao, W., & Tchetgen, E. T. (2020) · Current Epidemiology Reports · doi:10.1007/s40471-020-00243-4 · Cited by 188
“defined pair of potential outcomes (Y (0), Y (1)) corresponding to their outcome under active treatment a = 1 and control treatment a = 0, respectively [15, 16].”
BackgroundSocial Influence in Social Advertising: Evidence from Field Experiments
Bakshy, E., Eckles, D., Yan, R., & Rosenn, I. (2012) · arXiv (Cornell University) · doi:10.48550/arxiv.1206.4327 · Cited by 182
Chernozhukov, V., Demirer, M., Duflo, E., & Fernández-Val, I. (2018) · National Bureau of Economic Research · doi:10.3386/w24678 · Cited by 190
“Let Y (1) and Y (0) be the potential outcomes in the treatment state 1 and the non-treatment state 0 (see Neyman (1923), Rubin (1974)).”
InfluentialLi, Y. P., Propert, K. J., & Rosenbaum, P. R. (2001) · Journal of the American Statistical Association · doi:10.1198/016214501753208573 · Cited by 191
“Following Neyman (1923) and Rubin (1974, 1977), each patient m has a potential response rtvm that would be observed from patient m at time t if the patient received the treatment at time v 2 601dm7 [ 8c9.”
MethodEstimation of Average Treatment Effects with Misclassification
Lewbel, A. (2007) · Econometrica · doi:10.1111/j.1468-0262.2006.00756.x · Cited by 191
Goncharov, A., Haase, R. F., Santiago-Rivera, A., Morse, G., McCaffrey, R. J., Rej, R., & Carpenter, D. O. (2008) · Environmental Research · doi:10.1016/j.envres.2007.10.006 · Cited by 192
“This adjustment unintentionally removes much of the variance that is desirable for study as part of the exposureendpoint process (Greenland, 2000; Rosenbaum, 1995; Rothman and Greenland, 1998, 2005; Rubin, 1974; Sobel, 1994; Susser, 1973).”
BackgroundDo Voters Polarize When Radical Parties Enter Parliament?
Bischof, D., & Wagner, M. (2019) · American Journal of Political Science · doi:10.1111/ajps.12449 · Cited by 190
Predicting the efficacy of future training programs using past experiences at other locations
Joseph Hotz, V., Imbens, G. W., & Mortimer, J. H. (2005) · Journal of Econometrics · doi:10.1016/j.jeconom.2004.04.009 · Cited by 190
Imbens, G. W. (2020) · Journal of Economic Literature · doi:10.1257/jel.20191597 · Cited by 186
Time-uniform, nonparametric, nonasymptotic confidence sequences
Howard, S. R., Ramdas, A., McAuliffe, J., & Sekhon, J. (2021) · The Annals of Statistics · doi:10.1214/20-aos1991 · Cited by 190
“As one illustration of Theorem 4, we consider the sequential estimation of average treatment effect under the Neyman–Rubin potential outcomes model [27, 61, 67].”
MethodRobustness Tests for Quantitative Research
Neumayer, E., & Plümper, T. (2017) · Cambridge University Press eBooks · doi:10.1017/9781108233590 · Cited by 185
Causal Mediation Programs in R, M plus, SAS, SPSS, and Stata
Valente, M. J., Rijnhart, J. J. M., Smyth, H. L., Muniz, F. B., & MacKinnon, D. P. (2020) · Structural Equation Modeling: A Multidisciplinary Journal · doi:10.1080/10705511.2020.1777133 · Cited by 186
“Causal mediation analysis is based on the potential outcomes framework and the more general counterfactual framework (Holland, 1986; Rubin, 1974).”
BackgroundMethodClicking on Heaven’s Door: The Effect of Immigrant Legalization on Crime
Pinotti, P. (2017) · American Economic Review · doi:10.1257/aer.20150355 · Cited by 187
“For each individual, we can also define the potential outcomes C 1 and C 0 when L = 1 and L = 0, respectively (Rubin, 1974).”
Causal impact of masks, policies, behavior on early COVID-19 pandemic in the U.S.
Schrimpf, P., Kasahara, H., & Chernozhukov, V. (2020) · doi:10.1920/wp.cem.2020.2420 · Cited by 187
Causal Inference from Graphical Models
Lauritzen, S. (2000) · C&H/CRC Monographs on Statistics & Applied Probability · doi:10.1201/9781420035988.ch2 · Cited by 175
From Violence to Voting: War and Political Participation in Uganda
Blattman, C. (2008) · SSRN Electronic Journal · doi:10.2139/ssrn.1100110 · Cited by 185
Detecting Spillover Effects: Design and Analysis of Multilevel Experiments
Sinclair, B., McConnell, M., & Green, D. P. (2012) · American Journal of Political Science · doi:10.1111/j.1540-5907.2012.00592.x · Cited by 184
“SUTVA and Related Literature The Rubin Causal Model (Rubin 1974, 1978, 1980) de fines the effect of a treatment on individual i as the dif ference between two potential outcomes: the outcome Yu that manifests when this individual receives the treatment (f = 1) and the corresponding outcome Yj),…”
BackgroundTo prosper, organizational psychology should … overcome methodological barriers to progress
Edwards, J. R. (2008) · Journal of Organizational Behavior · doi:10.1002/job.529 · Cited by 185
“Control variables are useful in empirical research, as they can reduce bias due to omitted variables (James, 1980; Mauro, 1990) and address alternative explanations that undermine causal inference (Cook & Campbell, 1979; Rubin, 1974).”
Perils and Potentials of Self-Selected Entry to Epidemiological Studies and Surveys
Keiding, N., & Louis, T. A. (2016) · Journal of the Royal Statistical Society Series A: Statistics in Society · doi:10.1111/rssa.12136 · Cited by 184
Regression Discontinuity in Time: Considerations for Empirical Applications
Hausman, C., & Rapson, D. (2017) · National Bureau of Economic Research · doi:10.3386/w23602 · Cited by 186
Chou, R., Aronson, N., Atkins, D., Ismaila, A. S., Santaguida, P., Smith, D. H., Whitlock, E., Wilt, T. J., & Moher, D. (2010) · Journal of Clinical Epidemiology · doi:10.1016/j.jclinepi.2008.06.007 · Cited by 184
Women Politicians, Institutions, and Perceptions of Corruption
Barnes, T. D., & Beaulieu, E. (2019) · Comparative Political Studies · doi:10.1177/0010414018774355 · Cited by 184
“Random assignment of treatments is one essential element of experimental research designs (Rubin, 1974).”
BackgroundRemediation in the Community College: An Evaluator's Perspective
Levin, H. M., & Calcagno, J. C. (2008) · Community College Review · doi:10.1177/0091552107310118 · Cited by 182
“A control group generated through random assignment provides the best counterfactual to describe what would have happened to students in a treatment group if they had not been exposed to the treatment (Rubin, 1974; Holland, 1986).”
Graph cluster randomization: network exposure to multiple universes
Ugander, J., Karrer, B., Backstrom, L., & Kleinberg, J. (2013) · Proceedings of the 19th ACM SIGKDD international conference on Knowledge discovery and data mining · doi:10.1145/2487575.2487695 · Cited by 183
“Yet random-Permission ized experiments (or ‘A/B tests’), the standard machinery of testing frameworks including the Rubin causal model [14], critically assume what is known as the ‘stable unit treatment value assump-tion’ (SUTVA), that each individual’s response is affected only by their own…”
MethodA comparison of different methods to adjust survival curves for confounders
Denz, R., Klaaßen‐Mielke, R., & Timmesfeld, N. (2023) · Statistics in Medicine · doi:10.1002/sim.9681 · Cited by 181
“13,50-52 We roughly follow the method described in Austin13 and use the following three steps: (1) estimation of the propensity score, (2) matching patients with similar scores, and (3) using a”
MethodPropensity score matching and complex surveys
Austin, P. C., Jembere, N., & Chiu, M. (2018) · Statistical Methods in Medical Research · doi:10.1177/0962280216658920 · Cited by 181
“Prior to presenting alternate estimators, we briefly introduce the potential outcomes framework.(12) Let Y(1) and Y(0) denote the potential outcomes observed under the active treatment (Z1⁄4 1) and the control treatment (Z1⁄4 0), respectively.”
MethodDoes Three Strikes Deter?: A Nonparametric Estimation
Helland, E., & Tabarrok, A. (2007) · Journal of Human Resources · doi:10.3368/jhr.xlii.2.309 · Cited by 181
Bingenheimer, J. B., & Raudenbush, S. W. (2004) · Annual Review of Public Health · doi:10.1146/annurev.publhealth.25.050503.153925 · Cited by 181
“Statistical thinking about these inferential problems draws upon a counterfactual framework sometimes known as the Rubin causal model (86, 129, 133).”
BackgroundGhazouani, A., Xia, W., Ben Jebli, M., & Shahzad, U. (2020) · Sustainability · doi:10.3390/su12208680 · Cited by 179
“To investigate whether the adoption of carbon-tax policy in European countries promotes the reduction of CO2, we implement the nearest-neighbor matching (hereafter NNM) methodology initiated by Rubin [29] and developed by Rosenbaum and Rubin [30].”
MethodUnderstanding user perception toward artificial intelligence (AI) enabled e-learning
Kashive, N., Powale, L., & Kashive, K. (2020) · The International Journal of Information and Learning Technology · doi:10.1108/ijilt-05-2020-0090 · Cited by 181
Wordofa, M. G., Hassen, J. Y., Endris, G. S., Aweke, C. S., Moges, D. K., & Rorisa, D. T. (2021) · Agriculture & Food Security · doi:10.1186/s40066-020-00278-2 · Cited by 177
Ignoring the matching variables in cohort studies – when is it valid and why?
Sjölander, A., & Greenland, S. (2013) · Statistics in Medicine · doi:10.1002/sim.5879 · Cited by 176
“In standard potential-outcome notation [9, 10]: p.”
Non-parametric Methods for Doubly Robust Estimation of Continuous Treatment Effects
Kennedy, E. H., Ma, Z., McHugh, M. D., & Small, D. S. (2017) · Journal of the Royal Statistical Society Series B: Statistical Methodology · doi:10.1111/rssb.12212 · Cited by 174
“;Y) has support Z= (LAY ). Here L denotes a vector of covariates, Aa continuous treatment or exposure, and Y some outcome of interest. We characterize causal effects using potential outcome notation (Rubin, 1974), and so let Yadenote the potential outcome that would have been observed under treatment level a. We denot…”
Dahabreh, I. J., Robertson, S. E., Tchetgen, E. J., Stuart, E. A., & Hernán, M. A. (2019) · Biometrics · doi:10.1111/biom.13009 · Cited by 161
“For each treatment a ∈ A, we use the random variable Y a to denote the potential (counterfactual) outcome under intervention to receive treatment a (Splawa-Neyman, 1990; Rubin, 1974; Robins and Greenland, 2000).”
MethodBrown, C. H., Wang, W., Kellam, S. G., Muthén, B. O., Petras, H., Toyinbo, P., Poduska, J., Ialongo, N., Wyman, P. A., Chamberlain, P., et al. (2008) · Drug and Alcohol Dependence · doi:10.1016/j.drugalcdep.2007.11.013 · Cited by 177
Theory and Practice in Nonprobability Surveys
Mercer, A. W., Kreuter, F., Keeter, S., & Stuart, E. A. (2017) · Public Opinion Quarterly · doi:10.1093/poq/nfw060 · Cited by 176
Usami, S. (2021) · Structural Equation Modeling: A Multidisciplinary Journal · doi:10.1080/10705511.2020.1821690 · Cited by 175
“…can be interpreted in each model, and showed from the viewpoint of a potential outcome (or counterfactual) approach (the Rubin causal model; Rubin, 1974) that including stable trait factors as in the RI-CLPM is mathematically equivalent to controlling for latent (unobserved) time-invariant…”
BackgroundAn Exact and Robust Conformal Inference Method for Counterfactual and Synthetic Controls
Chernozhukov, V., Wüthrich, K., & Zhu, Y. (2021) · Journal of the American Statistical Association · doi:10.1080/01621459.2021.1920957 · Cited by 176
“Our analysis is developed within the potential (latent) outcome framework (Neyman, 1923; Rubin, 1974).”
MethodAssessing the Effect of Race Bias in Post-traffic Stop Outcomes Using Propensity Scores
Ridgeway, G. (2006) · Journal of Quantitative Criminology · doi:10.1007/s10940-005-9000-9 · Cited by 176
Common Sense and Sociological Explanations
Watts, D. J. (2014) · American Journal of Sociology · doi:10.1086/678271 · Cited by 176
The Last Instance: Are Institutions the Primary Cause of Economic Development?
PRZEWORSKI, A. (2004) · European Journal of Sociology · doi:10.1017/s0003975604001419 · Cited by 176
Aid as a Tool against Insurgency: Evidence from Contested and Controlled Territory in Afghanistan
SEXTON, R. (2016) · American Political Science Review · doi:10.1017/s0003055416000356 · Cited by 176
Causal Inference for Complex Longitudinal Data: The Continuous Case
Gill, R. D., & Robins, J. M. (2001) · The Annals of Statistics · doi:10.1214/aos/1015345962 · Cited by 176
Counterfactual Probabilities: Computational Methods, Bounds and Applications
Balke, A., & Pearl, J. (1994) · Uncertainty Proceedings 1994 · doi:10.1016/b978-1-55860-332-5.50011-0 · Cited by 168
A CONCEPTUAL FRAMEWORK FOR STUDYING THE SOURCES OF VARIATION IN PROGRAM EFFECTS
Weiss, M. J., Bloom, H. S., & Brock, T. (2014) · Journal of Policy Analysis and Management · doi:10.1002/pam.21760 · Cited by 174
On Pearl’s Hierarchy and the Foundations of Causal Inference
Bareinboim, E., Correa, J. D., Ibeling, D., & Icard, T. (2022) · Probabilistic and Causal Inference · doi:10.1145/3501714.3501743 · Cited by 174
Nonparametric Methods for Doubly Robust Estimation of Continuous Treatment Effects
Kennedy, E. H., Ma, Z., McHugh, M. D., & Small, D. S. (2015) · Collection of Biostatistics Research Archive · Cited by 183
A review of causal inference for biomedical informatics
Kleinberg, S., & Hripcsak, G. (2011) · Journal of Biomedical Informatics · doi:10.1016/j.jbi.2011.07.001 · Cited by 171
“6 This approach is similar to Rubin’s potential outcomes framework [21], which is”
Conti, G., & Heckman, J. J. (2010) · Perspectives on Psychological Science · doi:10.1177/1745691610383502 · Cited by 173
“Equations 2–4 are from the Neyman (1923), Fisher (1935), Cox (1958), and Rubin (1974) model of potential outcomes.”
A generalized probabilistic theory of causal relevance
Hitchcock, C. R. (1993) · Synthese · doi:10.1007/bf01064073 · Cited by 172
“If one were to draw a line of moderate slope, about 2/3, outward from the origin, and consider the restriction of g to this line, the values of g would display an inverted 'U' shape, increasing monotonically to 120/80, and decreasing monotonically beyond that.”
BackgroundMethodInfluentialField Experiments in Organizations
Eden, D. (2017) · Annual Review of Organizational Psychology and Organizational Behavior · doi:10.1146/annurev-orgpsych-041015-062400 · Cited by 167
The Role of Real‐World Evidence in FDA‐Approved New Drug and Biologics License Applications
Purpura, C. A., Garry, E. M., Honig, N., Case, A., & Rassen, J. A. (2022) · Clinical Pharmacology & Therapeutics · doi:10.1002/cpt.2474 · Cited by 171
Learning from Attempts to Improve Schooling: The Contribution of Methodological Diversity
Raudenbush, S. W. (2005) · Educational Researcher · doi:10.3102/0013189x034005025 · Cited by 171
“Statisticians refer to this as the assumption of ignorable treatment assignment (Holland, 1986; Rubin, 1974, 1978).”
BackgroundTipton, E. (2013) · Journal of Educational and Behavioral Statistics · doi:10.3102/1076998612441947 · Cited by 170
“This dual randomization process, however, rarely occurs in practice and is often infeasible (Bloom, 2005; Rubin, 1974; Shadish et al., 2002).”
BackgroundInfluentialPromotions and the Peter Principle*
Benson, A., Li, D., & Shue, K. (2019) · The Quarterly Journal of Economics · doi:10.1093/qje/qjz022 · Cited by 160
Lasso adjustments of treatment effect estimates in randomized experiments
Bloniarz, A., Liu, H., Zhang, C.-H., Sekhon, J. S., & Yu, B. (2016) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1510506113 · Cited by 169
“We give a brief outline of the Neyman-Rubin model for a randomized experiment; the reader is urged to consult [1], [9], and [14] for more details.”
MethodBridging the gap between research and practice: Predicting what will work locally
Joyce, K. E., & Cartwright, N. (2020) · American Educational Research Journal · doi:10.3102/0002831219866687 · Cited by 169
Variable selection for optimal treatment decision
Lu, W., Zhang, H. H., & Zeng, D. (2013) · Statistical Methods in Medical Research · doi:10.1177/0962280211428383 · Cited by 168
The Statistics of Causal Inference: A View from Political Methodology
Keele, L. (2015) · Political Analysis · doi:10.1093/pan/mpv007 · Cited by 168
“The potential outcomes framework (see, e.g., Rubin 1974), often referred to as the Rubin Causal Model (RCM) (Holland 1986), is one way to formalize the causal inference identification problem.”
Fu, X., & Akter, S. (2016) · The Journal of Development Studies · doi:10.1080/00220388.2016.1146700 · Cited by 169
“Following Rubin (1974) and Ravallion (2008), we measure impact as follows: where I is “impact”, O is the value of the interpretable impact indicator, T and C represent treatment group and control group respectively, i represents the sample units (households) and n is the sample size.”
MethodRobust post-matching inference
Abadie, A., & Spiess, J. (2022) · Journal of the American Statistical Association · doi:10.1080/01621459.2020.1840383 · Cited by 169
“Consider a standard binary treatment setting along the lines of Rubin (1974) with potential outcomes Y(1) and Y(0), of which we only observe Y = Y(W) for treatment W ∈ {0, 1}.”
MethodWhen Does Regulation Distort Costs? Lessons from Fuel Procurement in US Electricity Generation
Cicala, S. (2015) · American Economic Review · doi:10.1257/aer.20131377 · Cited by 168
Heterogeneous impacts in PROGRESA
Djebbari, H., & Smith, J. (2008) · Journal of Econometrics · doi:10.1016/j.jeconom.2008.05.012 · Cited by 161
“A debate regarding proper citations for the potential outcomes framework enlivens the footnotes of many papers; popular choices include Neyman (1923), Fisher (1935), Roy (1951), Quandt (1972) and Rubin (1974).”
Causal inference from cross-sectional earth system data with geographical convergent cross mapping
Gao, B., Yang, J., Chen, Z., Sugihara, G., Li, M., Stein, A., Kwan, M.-P., & Wang, J. (2023) · Nature Communications · doi:10.1038/s41467-023-41619-6 · Cited by 166
“…Wright 9 and was developed by Pearl 10,11 , and the potential outcome frameworks, which originated from Neyman and Fisher and was developed by Rubin 12,13 , are the two main frameworks to identify the existence of causal relationships and estimate causal effects (the strength of the causal…”
BackgroundEnhancing Rigor in Quantitative Entrepreneurship Research
Maula, M., & Stam, W. (2020) · Entrepreneurship Theory and Practice · doi:10.1177/1042258719891388 · Cited by 167
“Furthermore, when advocating for the PO framework (Rubin, 1974) for causal inference, structural equation modeling (SEM) has previously been falsely portrayed as an incompatible (and as an inferior approach for causal inference) approach without recognizing their equivalency (Bollen & Pearl, 2013; Pearl, 2012).”
BackgroundMethodInfluentialRevisiting the dedifferentiation hypothesis with longitudinal multi-cohort data
de Frias, C. M., Lövdén, M., Lindenberger, U., & Nilsson, L.-G. (2007) · Intelligence · doi:10.1016/j.intell.2006.07.011 · Cited by 168
“The FIML algorithm assumes missing-at-random (MAR: Rubin, 1974; see Schafer & Graham, 2002, for a nontechnical treatment), which means that the probability that a score on X is missing may depend on other variables but not on X itself.”
Method
Metadata
- Authors
- Rubin, D. B.
- Affiliations
- Educational Testing Service (US)
- Corresponding author
- Rubin, Donald B.
- ORCID
- Rubin, Donald B.
- Venue
- Journal of Educational Psychology
- Published
- 1974-10
- Volume
- 66
- Issue
- 5
- Pages
- 688-701
- ISSN
- 0022-0663 (ISSN-L)
- 1939-2176
- OpenAlex
- W2132917208
- Cited by
- 9,436
- References
- 10
- Citations by year
- 2026: 490 · 2025: 658 · 2024: 612 · 2023: 587 · 2022: 589 · 2021: 636 · 2020: 625 · 2019: 554 · 2018: 493 · 2017: 389 · 2016: 370 · 2015: 382 · 2014: 406 · 2013: 343 · 2012: 325
- FWCI
- 2.83
- Type
- journal article
- Language
- en
- Open access
- green
- nrs.harvard.edu
- Licence
- cc-by
- Locations
- Journal of Educational Psychology (publishedVersion)
- Digital Access to Scholarship at Harvard (DASH) (Harvard University) (submittedVersion, cc-by)
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Advanced Causal Inference Techniques
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Bayesian Inference
- Social Sciences › Education › Educational Research and Methods
- Keywords
- causal effect estimation, randomization, non-randomized studies, matching, random sampling, treatment effects, observational studies
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.
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.