The DOI matches its Crossref record; 53 of the 53 references it cites were found in a registry or an index.
Estimating average causal effects under general interference, with application to a social network experiment
Open accessAbstract
This paper presents a randomization-based framework for estimating causal effects under interference between units motivated by challenges that arise in analyzing experiments on social networks. The framework integrates three components: (i) an experimental design that defines the probability distribution of treatment assignments, (ii) a mapping that relates experimental treatment assignments to exposures received by units in the experiment, and (iii) estimands that make use of the experiment to answer questions of substantive interest. We develop the case of estimating average unit-level causal effects from a randomized experiment with interference of arbitrary but known form. The resulting estimators are based on inverse probability weighting. We provide randomization-based variance estimators that account for the complex clustering that can occur when interference is present. We also establish consistency and asymptotic normality under local dependence assumptions. We discuss refinements including covariate-adjusted effect estimators and ratio estimation. We evaluate empirical performance in realistic settings with a naturalistic simulation using social network data from American schools. We then present results from a field experiment on the spread of anti-conflict norms and behavior among school students.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Statistics and Causal Inference
Holland, P. W. (1986) · Journal of the American Statistical Association · doi:10.1080/01621459.1986.10478354 · Cited by 5,293
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
A generalization of sampling without replacement from a finite universe
Horvitz, D. G., & Thompson, D. J. (1952) · Journal of the American Statistical Association · doi:10.1080/01621459.1952.10483446 · Cited by 4,293
Causal Inference for Statistics, Social, and Biomedical Sciences: An Introduction
Imbens, G. W., & Rubin, D. B. (2015) · doi:10.1017/cbo9781139025751 · Cited by 4,210
Model Assisted Survey Sampling
Särndal, C., Swensson, B., & Wretman, J. H. (1992) · Springer series in statistics · doi:10.1007/978-1-4612-4378-6 · Cited by 3,417
Bayesian inference for causal effects: The role of randomization
Rubin, D. B. (1978) · The Annals of Statistics · doi:10.1214/aos/1176344064 · Cited by 2,503
MacKinnon, J. G., & White, H. (1985) · Journal of Econometrics · doi:10.1016/0304-4076(85)90158-7 · Cited by 1,879 · cited 2×
“variables for the exposure conditions, adjusting for network degree as a covariate, with MacKinnon and White (1985)’s finite sample adjusted “HC2” heteroskedasticity consistent variance estimator;”
MethodThe role of the propensity score in estimating dose-response functions
Imbens, G. W. (2000) · Biometrika · doi:10.1093/biomet/87.3.706 · Cited by 1,779 · cited 1×
“Invoking Imbens (2000)’s generalized propensity score , we call π i the generalized probability of exposure for i .”
BackgroundOn the Application of Probability Theory to Agricultural Experiments. Essay on Principles. Section 9
Neyman, J., Dabrowska, D. M., & Speed, T. P. (1990) · Statistical Science · doi:10.1214/ss/1177012031 · Cited by 1,680
Empirical Strategies in Labor Economics
Angrist, J. D., & Krueger, A. B. (1999) · Handbook of Labor Economics · doi:10.1016/s1573-4463(99)03004-7 · Cited by 1,628 · cited 1×
“However, OLS is known to be biased in its aggregation of unit level heterogeneity in causal effects (Angrist and Krueger, 1999).”
BackgroundIdentification of Peer Effects through Social Networks
Bramoullé, Y., Djebbari, H., & Fortin, B. (2009) · Social Science Research Network · doi:10.1016/j.jeconom.2008.12.021 · Cited by 1,388 · cited 1×
“9 The richness of these data makes Add Health a canonical dataset for methodological research related to social networks, as with Bramoulle et al. (2009), Chung et al. (2008), Goel and Salganik (2010), Goodreau et al. (2009), Goodreau (2007), Handcock et al. (2007), and Hunter et al. (2008).”
Aral, S., & Walker, D. (2011) · Management Science · doi:10.1287/mnsc.1110.1421 · Cited by 786 · cited 1×
“This simulated experiment resembles various studies of network persuasion campaigns (Chen et al., 2010; Aral and Walker, 2011; Paluck, 2011).”
MethodGoodreau, S. M., Kitts, J. A., & Morris, M. (2009) · Demography · doi:10.1353/dem.0.0045 · Cited by 776 · cited 1×
“9 The richness of these data makes Add Health a canonical dataset for methodological research related to social networks, as with Bramoulle et al. (2009), Chung et al. (2008), Goel and Salganik (2010), Goodreau et al. (2009), Goodreau (2007), Handcock et al. (2007), and Hunter et al. (2008).”
Toward Causal Inference With Interference
Hudgens, M. G., & Halloran, M. E. (2008) · Journal of the American Statistical Association · doi:10.1198/016214508000000292 · Cited by 768 · cited 1×
“9 The richness of these data makes Add Health a canonical dataset for methodological research related to social networks, as with Bramoulle et al. (2009), Chung et al. (2008), Goel and Salganik (2010), Goodreau et al. (2009), Goodreau (2007), Handcock et al. (2007), and Hunter et al. (2008).”
Model-Based Clustering for Social Networks
Handcock, M. S., Raftery, A. E., & Tantrum, J. M. (2007) · Journal of the Royal Statistical Society Series A (Statistics in Society) · doi:10.1111/j.1467-985x.2007.00471.x · Cited by 759
Goodness of Fit of Social Network Models
Hunter, D. R., Goodreau, S. M., & Handcock, M. S. (2008) · Journal of the American Statistical Association · doi:10.1198/016214507000000446 · Cited by 689
Agnostic notes on regression adjustments to experimental data: Reexamining Freedman’s critique
Lin, W. T. (2013) · The Annals of Applied Statistics · doi:10.1214/12-aoas583 · Cited by 686
Changing climates of conflict: A social network experiment in 56 schools
Paluck, E. L., Shepherd, H., & Aronow, P. M. (2016) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1514483113 · Cited by 610 · cited 1×
“A first method of covariance adjustment is based on the so-called “difference estimator” [Raj (1965), Särndal, Swensson and Wretman (1992), Chapter 6].”
MethodSelf-citationFormal mode of statistical inference for causal effects
Rubin, D. B. (1990) · Journal of Statistical Planning and Inference · doi:10.1016/0378-3758(90)90077-8 · Cited by 483
Sobel, M. E. (2006) · Journal of the American Statistical Association · doi:10.1198/016214506000000636 · Cited by 468 · cited 1×
“Consistency for the variance estimators, asymptotic normality and therefore confidence intervals follow straightforwardly when the growth process involves designs and exposure mappings that are independent across subpopulations, implying partial interference (Sobel, 2006).”
BackgroundThe Consistency Statement in Causal Inference
Cole, S. R., & Frangakis, C. (2008) · Epidemiology · doi:10.1097/ede.0b013e31818ef366 · Cited by 451 · cited 2×
“Interference represents a departure from the traditional causal inference scenario wherein units are assigned directly to treatment or control, and the potential outcomes that would be observed for a unit in either the treatment or control condition are fixed (Cole and Frangakis, 2009) and do not depend on the overall…”
BackgroundOn causal inference in the presence of interference
Tchetgen, E. J. T., & VanderWeele, T. J. (2010) · Statistical Methods in Medical Research · doi:10.1177/0962280210386779 · Cited by 423
Survey Design under the Regression Superpopulation Model
Isaki, C. T., & Fuller, W. A. (1982) · Journal of the American Statistical Association · doi:10.1080/01621459.1982.10477770 · Cited by 423 · cited 1×
“To define a notion of asymptotic growth, we let N tend to infinity, allowing for the experimental design to be reapplied anew to each UN , subject to the conditions defined below [Brewer (1979), Isaki and Fuller (1982)].”
MethodAssessing respondent-driven sampling
Goel, S., & Salganik, M. (2010) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1000261107 · Cited by 387
Interference Between Units in Randomized Experiments
Rosenbaum, P. R. (2007) · Journal of the American Statistical Association · doi:10.1198/016214506000001112 · Cited by 371
Concerning the Consistency Assumption in Causal Inference
VanderWeele, T. J. (2009) · Epidemiology · doi:10.1097/ede.0b013e3181bd5638 · Cited by 368 · cited 1×
“We assume the following consistency condition that relates the observed data to potential outcomes under the exposure model (VanderWeele, 2009):”
MethodAdvances in exponential random graph (p*) models applied to a large social network
Goodreau, S. M. (2006) · Social Networks · doi:10.1016/j.socnet.2006.08.001 · Cited by 342 · cited 1×
“9 The richness of these data makes Add Health a canonical dataset for methodological research related to social networks, as with Bramoulle et al. (2009), Chung et al. (2008), Goel and Salganik (2010), Goodreau et al. (2009), Goodreau (2007), Handcock et al. (2007), and Hunter et al. (2008).”
Identification of treatment response with social interactions
Manski, C. F. (2013) · Econometrics Journal · doi:10.1111/j.1368-423x.2012.00368.x · Cited by 283
Fuller, W. A. (2009) · Wiley eBooks · doi:10.1002/9780470523551 · Cited by 237
Hartley, H. O., & Ross, A. (1954) · Nature · doi:10.1038/174270a0 · Cited by 231
Normal approximation under local dependence
Chen, L. H. Y., & Shao, Q.-M. (2004) · The Annals of Probability · doi:10.1214/009117904000000450 · Cited by 217
Peer pressure against prejudice: A high school field experiment examining social network change
Paluck, E. L. (2010) · Journal of Experimental Social Psychology · doi:10.1016/j.jesp.2010.11.017 · Cited by 208 · cited 1×
“This simulated experiment resembles various studies of network persuasion campaigns (Chen et al., 2010; Aral and Walker, 2011; Paluck, 2011).”
MethodGraph 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
Causal Diagrams for Interference
Ogburn, E. L., & VanderWeele, T. J. (2014) · Statistical Science · doi:10.1214/14-sts501 · Cited by 149
Large Sample Randomization Inference of Causal Effects in the Presence of Interference
Liu, L., & Hudgens, M. G. (2013) · Journal of the American Statistical Association · doi:10.1080/01621459.2013.844698 · Cited by 133
Cluster–Robust Variance Estimation for Dyadic Data
Aronow, P. M., Samii, C., & Assenova, V. (2015) · Political Analysis · doi:10.1093/pan/mpv018 · Cited by 128
Self-citationReasoning about Interference Between Units: A General Framework
Bowers, J., Fredrickson, M., & Panagopoulos, C. (2012) · Political Analysis · doi:10.1093/pan/mps038 · Cited by 125
A Class of Robust Sampling Designs for Large-Scale Surveys
Brewer, K. R. W. (1979) · Journal of the American Statistical Association · doi:10.1080/01621459.1979.10481053 · Cited by 122 · cited 2×
“Combining expressions we obtain a conservative variance estimator for Var 6 Asymptotics and intervals Consider a sequence of subpopulations, U ( b ) , with b = 1 ,..., B (Brewer, 1979; Isaki and Fuller, 1982), where each subpopulation U ( b ) consists of 1 ≤ n b < ∞ units.”
BackgroundMethodAn Essay on the Logical Foundations of Survey Sampling, Part One*
Basu, D. (2010) · Springer eBooks · doi:10.1007/978-1-4419-5825-9_24 · Cited by 95 · cited 6×
“9 The richness of these data makes Add Health a canonical dataset for methodological research related to social networks, as with Bramoulle et al. (2009), Chung et al. (2008), Goel and Salganik (2010), Goodreau et al. (2009), Goodreau (2007), Handcock et al. (2007), and Hunter et al. (2008). We simulate experiments in…”
BackgroundMethodInfluentialDesigning experiments to measure spillover effects
Baird, S., Bohren, J., Mcintosh, C., & Ozler, B. (2014) · World Bank, Washington, DC eBooks · doi:10.1596/1813-9450-6824 · Cited by 95 · cited 1×
“The estimands and methods presented here may be useful in evaluating alternative experimental designs for estimating causal effects in the presence of interference (Airoldi, 2016; Baird et al., 2016; Eckles, Karrer and Ugander, 2014; Toulis and Kao, 2013; Ugander et al., 2013).”
MethodFattorini, L. (2006) · Biometrika · doi:10.1093/biomet/93.2.269 · Cited by 91 · cited 1×
“One may nonetheless approximate the π i s and joint probabilities with arbitrary precision through replication (Fattorini, 2006).”
BackgroundEstimation of Causal Peer Influence Effects
Toulis, P., & Kao, E. K. (2013) · Cited by 87
On a Method of Using Multi-Auxiliary Information in Sample Surveys
Raj, D. (1965) · Journal of the American Statistical Association · doi:10.1080/01621459.1965.10480789 · Cited by 84 · cited 1×
“A first method of covariance adjustment is based on the so-called “difference estimator” (Des Raj, 1965; S ¨ arndal et al., 1992, Ch. 6).”
MethodUnbiased Estimation of the Average Treatment Effect in Cluster-Randomized Experiments
Middleton, J. A., & Aronow, P. M. (2015) · Statistics Politics and Policy · doi:10.1515/spp-2013-0002 · Cited by 74
Self-citationA Class of Unbiased Estimators of the Average Treatment Effect in Randomized Experiments
Aronow, P. M., & Middleton, J. A. (2013) · Journal of Causal Inference · doi:10.1515/jci-2012-0009 · Cited by 72 · cited 1×
“Middleton and Aronow (2011) and Aronow and Middleton (2013) demonstrate that ŷT G(dk) is location invariant.”
BackgroundSelf-citationDesign and Analysis of Experiments in Networks: Reducing Bias from Interference
Eckles, D., Karrer, B., & Ugander, J. (2017) · Journal of Causal Inference · doi:10.1515/jci-2015-0021 · Cited by 49
Graph cluster randomization: network exposure to multiple universesPreprint
Ugander, J. H., Karrer, B., Bäckström, L., & Kleinberg, J. M. (2013) · arXiv (Cornell University) · doi:10.48550/arxiv.1305.6979 · Cited by 41
Latent transition analysis: Inference and estimation
Chung, H., Lanza, S. T., & Loken, E. (2007) · Statistics in Medicine · doi:10.1002/sim.3130 · Cited by 38 · cited 1×
“9 The richness of these data makes Add Health a canonical dataset for methodological research related to social networks, as with Bramoulle et al. (2009), Chung et al. (2008), Goel and Salganik (2010), Goodreau et al. (2009), Goodreau (2007), Handcock et al. (2007), and Hunter et al. (2008).”
Rosenbaum, P. R. (1999) · Biometrics · doi:10.1111/j.0006-341x.1999.00560.x · Cited by 35 · cited 1×
“We consider a simple “dilated effects” scenario (Rosenbaum, 1999) where potential outcomes are such that y i ( d 11 ) = 2 × y i ( d 00 ) , y i ( d 10 ) = 1 .”
BackgroundOn the Convergence of the Horvitz‐Thompson Estimator
Robinson, P. M. (1982) · Australian Journal of Statistics · doi:10.1111/j.1467-842x.1982.tb00829.x · Cited by 26 · cited 2×
“Proof of Proposition 6.2. We follow a proof technique similar to that of Aronow, Samii and Assenova (2015) to establish convergence of NV̂ar [τ̂HT (dk, dl)], though for a considerably more general setting.”
BackgroundMethodInfluentialOn the covariance between related Horvitz-Thompson estimators
Wood, J. B. (2008) · Journal of Official Statistics · Cited by 23
Unbiased Estimation of the Average Treatment Effect in Cluster-Randomized ExperimentsPreprint
Middleton, J. A., & Aronow, P. M. (2011) · SSRN Electronic Journal · doi:10.2139/ssrn.1803849 · Cited by 23
Self-citationConservative variance estimation for sampling designs with zero pairwise inclusion probabilities
Aronow, P. M., & Samii, C. (2012) · Survey methodology · Cited by 10
Self-citation
Research works printed but not linked: 4
- Aronow, P. M. (2013). Model Assisted Causal Inference. Ph.D. thesis, Dept. Political Science, Yale Univ., New Haven, CT.
- Basu, D. (1971). An essay on the logical foundations of survey sampling. I. In Foundations of Statistical Inference (V. Godambe and D. A. Sprott, eds.) 203–242. Holt, Rinehart and Winston, Toronto.
- Cox, D. R. (1958). Planning of Experiments. Wiley, New York.
- Manski, C. F. (1995). Identification Problems in the Social Sciences. Harvard Univ. Press, Cambridge, MA.
Cited by
Sorted by citations
The Econometrics of Randomized Experiments
Athey, S. C., & Imbens, G. W. (2017) · Handbook of economic field experiments · doi:10.1016/bs.hefe.2016.10.003 · Cited by 561
“Aronow and Samii (2013) study estimation in this general network setting.”
BackgroundMatching Methods for Causal Inference with Time‐Series Cross‐Sectional Data
Imai, K., Kim, I. S., & Wang, E. H. (2023) · American Journal of Political Science · doi:10.1111/ajps.12685 · Cited by 349
“Although the methodological literature has begun to relax the assumption of no spillover effect in experimental settings (e.g. Aronow and Samii 2017; Hudgens and Halloran 2008; Imai et al. 2021; Tchetgen Tchetgen and VanderWeele 2010).”
BackgroundA causal test of the strength of weak ties
Rajkumar, K., Saint-Jacques, G., Bojinov, I., Brynjolfsson, E., & Aral, S. (2022) · Science · doi:10.1126/science.abl4476 · Cited by 216
Exactp-Values for Network Interference
Athey, S. C., Eckles, D., & Imbens, G. W. (2016) · Journal of the American Statistical Association · doi:10.1080/01621459.2016.1241178 · Cited by 189
Peer Effects in Networks: A Survey
Bramoullé, Y., Djebbari, H., & Fortin, B. (2020) · Annual Review of Economics · doi:10.1146/annurev-economics-020320-033926 · Cited by 181
Metadata
- Authors
- Aronow, P. M., & Samii, C.
- Affiliations
- Yale University (US)
- Venue
- The Annals of Applied Statistics
- Publisher
- Institute of Mathematical Statistics
- Published
- 2017-12-01
- Volume
- 11
- Issue
- 4
- Pages
- 1912-1947
- ISSN
- 1932-6157 (ISSN-L)
- 1941-7330
- OpenAlex
- W2735728227
- Cited by
- 347
- References
- 51
- Citations by year
- 2026: 47 · 2025: 47 · 2024: 35 · 2023: 47 · 2022: 30 · 2021: 46 · 2020: 39 · 2019: 21 · 2018: 20 · 2017: 10
- FWCI
- 17.12
- Type
- journal article
- Language
- en
- Open access
- bronze
- doi.org
- Locations
- The Annals of Applied Statistics (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- Project Euclid (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Advanced Causal Inference Techniques
- Social Sciences › Decision Sciences › Management Science and Operations Research › Optimal Experimental Design Methods
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Clinical Trials
- Keywords
- randomization inference, inverse probability weighting, variance estimation, local dependence, covariate adjustment, ratio estimation, school students
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
