Pith. sign in

REVIEW 4 major objections 4 minor 55 references

Detecting Heterogeneous Treatment Effect with Instrumental Variables

T0 review · 4 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Tree-found subgroups can be tested on the same data in IV studies

desk verdict A useful synthesis of matching-IV and the absolute-difference honesty trick; structurally sound FWER control, but subset-level size is imported rather than proved. read the letter →

arxiv 1908.03652 v2 pith:D27W3ITC submitted 2019-08-09 stat.ME stat.APstat.ML

Michael Johnson , Jiongyi Cao , Hyunseung Kang This is my paper · ORCID
classification stat.MEstat.APstat.ML
keywords heterogeneouscomplieraveragecausaleffectinstrumentalvariablesmatchingclassificationandregressiontreesclosedtestingfamilywiseerrorrateOregonHealthInsuranceExperimentmodification
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper proposes a way to find and test subgroups with different causal effects in instrumental-variable studies, targeting what it calls the heterogeneous complier average causal effect (H-CACE), the usual complier effect restricted to a covariate-defined subgroup. The method matches units on pre-instrument covariates, uses a regression tree on the absolute value of the adjusted within-pair outcome differences to propose a grouping, and then runs closed testing on every union of the tree's leaves. Its central claim is that this pipeline can use the same data for discovery and inference: conditional on the data and the tree, the chance of at least one false rejection among the tested hypotheses is at most the nominal level. On the Oregon Health Insurance Experiment, the procedure finds that Medicaid's effect on days when health did not impede usual activities is concentrated in older, English-preferring, non-Asian men and younger, English-preferring people with at most a high school diploma or GED.

What carries the argument

The load-bearing device is the absolute-difference construction: the tree is trained on $|Y_i|$, the absolute value of the adjusted outcome difference within each matched pair, rather than on the signed difference, so the instrument assignment $Z$ is hidden from the tree and the randomized assignment distribution inside matched pairs remains intact. This is combined with the matched-pair IV test statistic $T_s(\lambda_0)/S_s(\lambda_0)$ from the matching-IV literature and closed testing over all unions of the tree's leaves; the absolute difference makes every test level $\alpha$ under the sharp null, and closed testing converts level-$\alpha$ tests into strong familywise error control. The entity being tested is the H-CACE $\lambda_s = \frac{\sum_{i\in s}\sum_j (r^{(1)}_{1sij}-r^{(0)}_{0sij})}{\sum_{i\in s}\sum_j (d_{1sij}-d_{0sij})}$, which extends the usual complier average causal effect to a covariate-defined subgroup.

What would settle it

Simulate data with no true effect but with a rare covariate combination that forces the tree to create a leaf with very few pairs, run Algorithm 1 repeatedly, and compare the empirical familywise error rate to $\alpha$. If the small-leaf statistic is not approximately standard normal, the rejection rate will exceed $\alpha$, which would falsify Proposition 1 as an unconditional guarantee.

Watch

Extended reading notes

Core claim

The paper's central result is Proposition 1: under the sharp null hypotheses used in Algorithm 1, the conditional probability, given the potential outcomes and covariates $F$, the instrument assignment $Z$, and the tree grouping $G$, that the algorithm makes at least one false rejection is at most $\alpha$. The mechanism is to feed the regression tree the absolute pairwise difference $|Y_i| = |(Z_{i1}-Z_{i2})(R_{i1}-\lambda_0 D_{i1} - (R_{i2}-\lambda_0 D_{i2}))|$ instead of the signed difference, so that the grouping depends only on $F$, not on $Z$; the instrument assignment therefore remains a fair coin within each matched pair, and the randomization-based null distribution used by the matching IV test statistics is preserved. The paper then runs closed testing over all unions of the tree's leaves, rejecting a subgroup hypothesis only when every superset hypothesis is rejected, which gives strong familywise error control. On the OHIE data the procedure reports significant H-CACEs in two subgroups, and simulations show high true discovery rates and low false-positive rates for effect modifiers, especially at low compliance rates.

Load-bearing premise

The guarantee rests on each subgroup test statistic $T_s(\lambda_0)/S_s(\lambda_0)$ being at least asymptotically level $\alpha$ under every sharp null visited by closed testing, even for the small, low-compliance leaves a regression tree can create, and on the tree being built from $X$ and $|Y|$ only so that the instrument assignment remains randomized within pairs.

Editorial extensions

If this is right

  • Data-driven subgroup discovery and inference can be done on the full sample, so the method avoids the power loss of sample splitting while still controlling the familywise error rate.
  • Any supervised grouping rule that uses only $X$ and $|Y|$ inherits the honesty property, so the same wrapper can be applied beyond the specific tree algorithm used here.
  • In the paper's simulations, the method achieves high true discovery rates and near-zero false positive rates for effect modifiers even when compliance is below 40 percent, a regime where IV inference is usually weak.
  • Subgroups with equal effect magnitudes but opposite signs are invisible to the method, because the absolute-value transformation collapses the direction of the effect.
  • The OHIE analysis locates Medicaid's effect in older, English-preferring, non-Asian men and younger, English-preferring adults with at most a high school diploma or GED.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial inference: replacing the single tree with a forest would require each leaf's test to remain level $\alpha$, so an ensemble version would need either leaf-level calibration or a different multiplicity correction before the Proposition 1 guarantee could be claimed.
  • Editorial inference: the closed-testing machinery is invertible in $\lambda_0$, so a natural extension would construct simultaneous confidence sets for all H-CACEs rather than testing a single fixed null value.
  • Editorial inference: the OHIE tree depth was chosen by pruning back for interpretability, a judgment call that is not demonstrated to be independent of outcome-dependent information; an automatic, outcome-free tuning rule would make the procedure fully reproducible.
  • Editorial inference: if the cited sharp-null-to-weak-null validity result holds generally, the procedure's error control would extend from sharp hypotheses to the weaker hypotheses $\lambda_s = \lambda_0$, which is what many users would care about in practice.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. This paper proposes a method for discovering and testing heterogeneity in the complier average causal effect (H-CACE) in instrumental variables studies with paired matching. The method first pair-matches units on covariates, then grows a CART tree on the absolute value of the adjusted pairwise outcome difference |Y_i|, which is a function of the potential outcomes only under the sharp null, and finally applies closed testing to all unions of tree leaves using the IV test statistic of Baiocchi et al. (2010) and Kang et al. (2016). Proposition 1 claims strong familywise error control at level α conditional on (F, Z, G). The authors present simulations comparing the method with BCF-IV and apply it to the Oregon Health Insurance Experiment, reporting Medicaid effects among older English-preferring non-Asian men and younger English-preferring individuals with at most a high school diploma or GED.

Significance. The proposal addresses a real gap: existing tree-based heterogeneity methods for IV settings either rely on sample splitting or do not provide strong familywise error control, whereas the use of |Y| to decouple discovery from the instrument assignment is elegant and, if valid, permits honest inference without sample splitting. The H-CACE estimand is defined before estimation, so the analysis is not circular. The closed-testing structure is sound conditional on having level-α tests for every intersection hypothesis, and the OHIE application is a useful demonstration. However, the strength of the contribution depends on closing the gap between the imported asymptotic test and the sparse, data-dependent subsets used by CART, and on clarifying the tuning and matching assumptions.

major comments (4)
  1. [Web Appendix B; Eqs. (2)-(3); Algorithm 1] The proof of Proposition 1 in Web Appendix B reduces familywise error control to the claim that each intersection hypothesis H0L is tested at level α by |T_s(λ0)/S_s(λ0)|, and it states the equality P(|T/S| ≥ z_{1−α/2} | F,Z,G) = α/2. This equality is not exact in finite samples: the normal approximation is imported from Baiocchi et al. (2010) and Kang et al. (2016), where it is derived for a fixed matched sample with I growing, not for the variable, data-dependent unions of CART leaves that Algorithm 1 visits. In the OHIE analysis the nodes in Figure 4 have I_s in the hundreds and compliance near 0.29, so the subset statistics can have a much smaller effective sample of compliers; no uniform asymptotic or bounded-variance condition is provided for S_s(λ0). Please either prove the size property under explicit conditions on the class of possible groupings, or replace the normal approximation with an exact randomization test conditional on (F,Z,G), so that Proposition 1 holds exactly as stated.
  2. [Section 4.3; Algorithm 1, Step 3] Proposition 1 relies on G being a function of X and |Y| only. Section 4.3 reports that the OHIE tree depth was chosen by forming trees of larger depth and pruning back "until a more interpretable tree was obtained." If this search uses estimated subgroup effects, confidence intervals, or p-values, then G depends on Z and the conditional randomization argument fails. Please specify the exact tuning procedure for the complexity parameter, maximum depth, and pruning, and verify that it is a deterministic function of (X, |Y|) alone.
  3. [Section 2.3, Eqs. (4)-(5); Section 4.1] The definition of H-CACE at covariate value x is rewritten as a pair-subset estimand λ_s under the assumption that units within a matched pair have identical covariate values. The OHIE matching in Section 4.1 uses an optimal caliper and a robust rank-based Mahalanobis distance on age and several categorical variables, which does not generally yield exact covariate agreement within pairs. The paper does not define the pair-level covariate Xi used by CART or explain how approximate matching changes the interpretation of λ_s as a covariate-defined H-CACE. Please state the exact-matching assumption explicitly or redefine the estimand and the discovered subgroups at the level of matched pairs.
  4. [Web Appendix A, Table 1] The simulation evidence for finite-sample FWER is limited to one configuration with a strong signal (λ00=2, others 0), and the text itself notes that the trees from Y and |Y| are nearly identical at the upper levels in this configuration, so the simulation is not informative about the level of the subset-specific tests in the problematic small or low-compliance intersections. Please add simulations that report the size of |T_s(λ0)/S_s(λ0)| for sparse and low-compliance intersection hypotheses, or otherwise justify the asymptotic approximation in those settings.
minor comments (4)
  1. [Figures 1-2, 7-10] The simulation plots show averages over 1000 replications without Monte Carlo error bars or confidence bands; adding them would help interpret claims such as a false positive rate of zero.
  2. [Figure 9] The legend labels the comparison method as "BCV-IV"; this should read "BCF-IV."
  3. [Section 2.4] The statement that the honesty property is preserved for "any supervised machine learning algorithm" and "subsequent visual heuristics" is broader than what is proved; visual heuristics are not formal deterministic algorithms, and their independence from Z should be justified or the sentence qualified.
  4. [Web Appendix B] The proof writes "H0K : λs = λ0" for the subset null, but the algorithm's nulls are sharp unit-level nulls; aligning the notation would make the closed-testing argument easier to follow.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the FWER guarantee is a standard closed-testing argument, and the level-α test statistic is imported from independent prior work rather than fitted to this paper's data.

full rationale

The paper's central claim, Proposition 1, is proved in Web Appendix B by a conventional closed-testing reduction: under the sharp nulls, any false rejection of H0K requires rejection of the intersection H0T of all true hypotheses, and that intersection test is level α because |T(λ0)/S(λ0)| ≥ z_{1−α/2} has probability α/2 conditional on (F,Z,G). The paper does not fit any parameter and then rename it as a prediction; the H-CACE estimand is defined before estimation in Section 2.3, and the discovered subgroups are produced by CART on |Y|, which is a function of F only, preserving the randomization distribution of Z. The level-α property of the test statistic is cited from Baiocchi et al. (2010) and Kang et al. (2016), both published and externally established results that do not include the paper's target claim as an input; this is self-citation in the sense of shared authorship, but it is real, independent support rather than a circular reduction. The OHIE findings are outputs of the tree plus closed testing, not quantities forced by construction from the inputs. Concerns about the asymptotic approximation for sparse or low-compliance CART subsets are correctness or robustness issues, not circularity, and no specific equation in the paper reduces to an earlier fitted value or to the paper's own conclusion.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The theoretical method has no fitted constants in its central claim. The listed free parameters are hand-chosen CART tuning values that affect the simulations and the empirical tree, but they are not part of the FWER proof. The axioms are the IV identifying assumptions, the sharp-null form needed for the guarantee, the imported asymptotic normal approximation, and the conditional independence of the tree grouping from the random assignment. No new physical or conceptual entities are postulated; H-CACE is a defined estimand rather than an entity with independent evidence.

free parameters (4)
  • CART complexity parameter in OHIE analysis = 0
    Section 4.3 selects cp=0. This is a data-dependent choice and is outside the formal FWER proof.
  • CART maximum depth in OHIE analysis = 4
    Section 4.3 forms deeper trees and prunes back until an interpretable tree is obtained. The text does not show this choice is based only on X and |Y|.
  • CART complexity parameter in main simulations = 0.005
    Section 3 uses cp=0.005, half of the rpart default. This arbitrary tuning affects the reported power and false positive rates.
  • OHIE-like simulation CART settings = cp=0.001, max depth=7, min split=90
    Web Appendix F selects these values to allow more splits. They affect the simulation results but are not part of the theoretical claim.
assumptions (4)
  • domain assumption Instrumental variables assumptions (A1)-(A4): relevance, exclusion restriction, ignorability or exchangeability, and monotonicity.
    Assumed in Section 2.2 to identify CACE and H-CACE. The exclusion restriction is the fragile one and is discussed as potentially violated in Section 4.2.
  • domain assumption Sharp null hypotheses H0L: r(d1)-r(d0)=lambda0(d1-d0) for all units in L.
    Proposition 1 guarantees FWER control only for sharp nulls. Weak-null validity is borrowed from Fogarty et al. (2020) rather than proved here.
  • domain assumption Asymptotic standard normality of T_s(lambda0)/S_s(lambda0) under the sharp null.
    The closed-testing threshold z_{1-alpha/2} in Algorithm 1 requires this distribution. The paper cites Baiocchi et al. (2010) and Kang et al. (2016) instead of deriving it for the subgroup setting.
  • domain assumption The CART grouping G is a function of F only, not of the instrument assignment Z.
    The absolute-value transform makes |Y_i| independent of Z under the sharp null, but the proof also requires every hyperparameter and tree-building choice to respect this. The OHIE depth selection is not explicitly verified.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Detecting Heterogeneous Treatment Effect with Instrumental Variables." pith.science (2026). https://pith.science/paper/D27W3ITC

@misc{pith2026190803652,
  author       = {Pith},
  title        = {Pith review of: Detecting Heterogeneous Treatment Effect with Instrumental Variables},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/D27W3ITC}},
  note         = {Machine review of arXiv:1908.03652}
}
read the original abstract

There is an increasing interest in estimating heterogeneity in causal effects in randomized and observational studies. However, little research has been conducted to understand heterogeneity in an instrumental variables study. In this work, we present a method to estimate heterogeneous causal effects using an instrumental variable approach. The method has two parts. The first part uses subject-matter knowledge and interpretable machine learning techniques, such as classification and regression trees, to discover potential effect modifiers. The second part uses closed testing to test for the statistical significance of the effect modifiers while strongly controlling familywise error rate. We conducted this method on the Oregon Health Insurance Experiment, estimating the effect of Medicaid on the number of days an individual's health does not impede their usual activities, and found evidence of heterogeneity in older men who prefer English and don't self-identify as Asian and younger individuals who have at most a high school diploma or GED and prefer English.

Figures

Figures reproduced from arXiv: 1908.03652 by the authors.

Figure 1
Figure 1. True discovery rate as a function of the compliance rate and heterogeneity settings. The dashed and [PITH_FULL_IMAGE:figures/full_fig_p011_1.png] view at source ↗
Figure 2
Figure 2. F-score and false positive rate as a function of compliance rate and heterogeneity settings. The shapes [PITH_FULL_IMAGE:figures/full_fig_p012_2.png] view at source ↗
Figure 3
Figure 3. Covariate balance as measured by difference in means of the covariates between the treated and [PITH_FULL_IMAGE:figures/full_fig_p014_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Results of our proposed method on the effect of enrolling in Medicaid on the number of days physical [PITH_FULL_IMAGE:figures/full_fig_p016_4.png]
Figure 5
Figure 5. Figure 5: Illustration of closed testing to test the null hypothesis [PITH_FULL_IMAGE:figures/full_fig_p017_5.png]
Figure 6
Figure 6. Figure 6: Histogram of p-values obtained from using |Y | and Y as the outcomes in CART in discovery of potential effect modifiers. The black dashed line denotes the alpha level of 0.05 of the hypothesis tests. 19 [PITH_FULL_IMAGE:figures/full_fig_p019_6.png]
Figure 7
Figure 7. Figure 7: True discovery rate as a function of overall compliance rate for the four treatment and four compliance [PITH_FULL_IMAGE:figures/full_fig_p023_7.png]
Figure 8
Figure 8. Figure 8: F-score and false positive rates (FPR) as a function of overall compliance rate for the four treatment [PITH_FULL_IMAGE:figures/full_fig_p024_8.png]
Figure 9
Figure 9. Figure 9: True discovery rate as a function of overall compliance rate for four compliance heterogeneity settings. [PITH_FULL_IMAGE:figures/full_fig_p025_9.png]
Figure 10
Figure 10. Figure 10: False positive rate (FPR) and F-score as a function of overall compliance rate for the four compliance [PITH_FULL_IMAGE:figures/full_fig_p026_10.png]
Figure 11
Figure 11. Figure 11: True discovery rate as a function of overall compliance rate for the strong heterogeneity and complex [PITH_FULL_IMAGE:figures/full_fig_p027_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 39 canonical work pages

  1. [1]

    Abadie, A. (2003). Semiparametric instrumental variable estimation of treatment response models. Journal of Econometrics , 113(2):231--263

    2003

  2. [2]

    and Chen, X

    Ai, C. and Chen, X. (2003). Efficient estimation of models with conditional moment restrictions containing unknown functions. Econometrica , 71(6):1795--1843

    2003

  3. [3]

    D., Imbens, G

    Angrist, J. D., Imbens, G. W., and Rubin, D. B. (1996). Identification of causal effects using instrumental variables. Journal of the American Statistical Association , 91(434):444--455

    1996

  4. [4]

    and Imbens, G

    Athey, S. and Imbens, G. (2016). Recursive partitioning for heterogeneous causal effects. Proceedings of the National Academy of Sciences , 113(27):7353--7360

    2016

  5. [5]

    Recursive Partitioning for Heterogeneous Causal Effects

    Athey, S. and Imbens, G. W. (2015). Machine learning methods for estimating heterogeneous causal effects. arXiv:1504.01132v1 [stat.ML]

    work page Pith review arXiv 2015

  6. [6]

    Athey, S., Tibshirani, J., Wager, S., et al. (2019). Generalized random forests. The Annals of Statistics , 47(2):1148--1178

    work page 2019

  7. [7]

    Baiocchi, M., Cheng, J., and Small, D. S. (2014). Instrumental variable methods for causal inference. Statistics in Medicine , 33(13):2297--2340

    work page 2014

  8. [8]

    S., Lorch, S., and Rosenbaum, P

    Baiocchi, M., Small, D. S., Lorch, S., and Rosenbaum, P. R. (2010). Building a stronger instrument in an observational study of perinatal care for premature infants. Journal of the American Statistical Association , 105(492):1285--1296

    2010

Show all 55 references
  1. [9]

    and Pearl, J

    Balke, A. and Pearl, J. (1997). Bounds on treatment effects from studies with imperfect compliance. Journal of the American Statistical Association , 92(439):1171--1176

    1997

  2. [10]

    J., De-Witte, K., and Gnecco, G

    Bargagli-Stoffi, F. J., De-Witte, K., and Gnecco, G. (2019). Heterogeneous causal effects with imperfect compliance: a novel bayesian machine learning approach. arXiv:1905.12707 [stat.ME]

    2019 arXiv

  3. [11]

    Bargagli-Stoffi, F. J. and Gnecco, G. (2018). Estimating heterogeneous causal effects in the presence of irregular assignment mechanisms. In 2018 IEEE 5th International Conference on Data Science and Advanced Analytics (DSAA) , pages 1--10. IEEE

    2018

  4. [12]

    Blundell, R., Chen, X., and Kristensen, D. (2007). Semi-nonparametric iv estimation of shape-invariant engel curves. Econometrica , 75(6):1613--1669

    2007

  5. [13]

    and Powell, J

    Blundell, R. and Powell, J. L. (2003). Endogeneity in nonparametric and semiparametric regression models. Econometric Society Monographs , 36:312--357

    2003

  6. [14]

    Breiman, L., Friedman, J., Olshen, R., and Stone, C. (1984). Classification and Regression Trees. New York: Chapman and Hall/CRC

    1984

  7. [15]

    and Pouzo, D

    Chen, X. and Pouzo, D. (2012). Estimation of nonparametric conditional moment models with possibly nonsmooth generalized residuals. Econometrica , 80(1):277--321

    2012

  8. [16]

    Chernozhukov, V., Demirer, M., Duflo, E., and Fernandez-Val, I. (2018). Generic machine learning inference on heterogenous treatment effects in randomized experiments. National Bureau of Economic Research

    2018

  9. [17]

    Darolles, S., Fan, Y., Florens, J.-P., and Renault, E. (2011). Nonparametric instrumental regression. Econometrica , 79(5):1541--1565

    2011

  10. [18]

    Ding, P. (2017). A paradox from randomization-based causal inference. Statistical Science , 32(3):331--345

    2017

  11. [19]

    P., Allen, H., Baicker, K., and Group, O

    Finkelstein, A., Taubman, S., Wright, B., Bernstein, M., Gruber, J., Newhouse, J. P., Allen, H., Baicker, K., and Group, O. H. S. (2012). The oregon health insurance experiment: evidence from the first year. The Quarterly Journal of Economics , 127(3):1057--1106

    2012

  12. [20]

    Fogarty, C. B. (2018). Regression-assisted inference for the average treatment effect in paired experiments. Biometrika , 105(4):994--1000

    2018

  13. [21]

    Fogarty, C. B. (2020). Studentized sensitivity analysis for the sample average treatment effect in paired observational studies. Journal of the American Statistical Association , 115(531):1518--1530

    2020

  14. [22]

    B., Lee, K., Kelz, R

    Fogarty, C. B., Lee, K., Kelz, R. R., and Keele, L. J. (2020). Biased encouragements and heterogeneous effects in an instrumental variable study of emergency general surgical outcomes. Journal of the American Statistical Association , pages 1--32

    2020

  15. [23]

    R., Murray, J

    Hahn, P. R., Murray, J. S., and Carvalho, C. (2017). Bayesian regression tree models for causal inference: regularization, confounding, and heterogeneous effects. arXiv:1706.09523 [stat,ME]

    2017 arXiv

  16. [24]

    and Horowitz, J

    Hall, P. and Horowitz, J. L. (2005). Nonparametric methods for inference in the presence of instrumental variables. The Annals of Statistics , 33(6):2904--2929

    2005

  17. [25]

    Hern \'a n, M. A. and Robins, J. M. (2006). Instruments for causal inference: an epidemiologist's dream? Epidemiology , 17(4):360--372

    2006

  18. [26]

    Hill, J. L. (2011). Bayesian nonparametric modeling for causal inference. Journal of Computational and Graphical Statistics , 20(1):217--240

    2011

  19. [27]

    Hodges, J. L. and Lehmann, E. L. (1963). Estimates of location based on rank tests. The Annals of Mathematical Statistics , 34(2):598--611

    1963

  20. [28]

    Y., Small, D

    Hsu, J. Y., Small, D. S., and Rosenbaum, P. R. (2013). Effect modification and design sensitivity in observational studies. Journal of the American Statistical Association , 108(501):135--148

    2013

  21. [29]

    Y., Zubizarreta, J

    Hsu, J. Y., Zubizarreta, J. R., Small, D. S., and Rosenbaum, P. R. (2015). Strong control of the familywise error rate in observational studies that discover effect modification by exploratory methods. Biometrika , 102(4):767--782

    2015

  22. [30]

    Imbens, G. W. (2010). Better late than nothing: Some comments on deaton (2009) and heckman and urzua (2009). Journal of Economic Literature , 48(2):399--423

    2010

  23. [31]

    Kang, H., Kreuels, B., Adjei, O., Krumkamp, R., May, J., and Small, D. S. (2013). The causal effect of malaria on stunting: a mendelian randomization and matching approach. International Journal of Epidemiology , 42(5):1390--1398

    2013

  24. [32]

    S., et al

    Kang, H., Kreuels, B., May, J., Small, D. S., et al. (2016). Full matching approach to instrumental variables estimation with application to the effect of malaria on stunting. The Annals of Applied Statistics , 10(1):335--364

    2016

  25. [33]

    Kang, H., Peck, L., and Keele, L. (2018). Inference for instrumental variables: a randomization inference approach. Journal of the Royal Statistical Society: Series A (Statistics in Society) , 181(4):1231--1254

    2018

  26. [34]

    S., and Dominici, F

    Lee, K., Small, D. S., and Dominici, F. (2018a). Discovering effect modification and randomization inference in air pollution studies. arXiv:1802.06710 [stat.AP]

    2018 arXiv

  27. [35]

    S., Hsu, J

    Lee, K., Small, D. S., Hsu, J. Y., Silber, J. H., and Rosenbaum, P. R. (2018b). Discovering effect modification in an observational study of surgical mortality at hospitals with superior nursing. Journal of the Royal Statistical Society: Series A (Statistics in Society) , 181(...

    2018

  28. [36]

    S., and Rosenbaum, P

    Lee, K., Small, D. S., and Rosenbaum, P. R. (2018c). A powerful approach to the study of moderate effect modification in observational studies. Biometrics , 74(4):1161--1170

    2018

  29. [37]

    Marcus, R., Eric, P., and Gabriel, K. R. (1976). On closed testing procedures with special reference to ordered analysis of variance. Biometrika , 63(3):655--660

    1976

  30. [38]

    Newey, W. K. and Powell, J. L. (2003). Instrumental variable estimation of nonparametric models. Econometrica , 71(5):1565--1578

    2003

  31. [39]

    and Kang, H

    Park, C. and Kang, H. (2020). A groupwise approach for inferring heterogeneous treatment effects in causal inference. arXiv preprint arXiv:1908.04427v2

    2020 arXiv

  32. [40]

    Rosenbaum, P. R. (2002a). Covariance adjustment in randomized experiments and observational studies. Statistical Science , 17(3):286--327

    2002

  33. [41]

    Rosenbaum, P. R. (2002b). [covariance adjustment in randomized experiments and observational studies]: Rejoinder. Statistical Science , 17(3):321--327

    2002

  34. [42]

    Rosenbaum, P. R. (2010). Design of Observational Studies . New York: Springer

    2010

  35. [43]

    Rosenbaum, P. R. (2020). Modern algorithms for matching in observational studies. Annual Review of Statistics and Its Application , 7:143--176

    2020

  36. [44]

    Rubin, D. B. (1980). Randomization analysis of experimental data: The fisher randomization test comment. Journal of the American Statistical Association , 75(371):591--593

    1980

  37. [45]

    Rubin, D. B. (2001). Using propensity scores to help design observational studies: application to the tobacco litigation. Health Services and Outcomes Research Methodology , 2(3-4):169--188

    2001

  38. [46]

    Staiger, D. O. and Stock, J. H. (1997). Instrumental variables regression with weak instruments. Econometrica , 65(3):557--586

    1997

  39. [47]

    H., Wright, J

    Stock, J. H., Wright, J. H., and Yogo, M. (2002). A survey of weak instruments and weak identification in generalized method of moments. Journal of Business & Economic Statistics , 20(4):518--529

    2002

  40. [48]

    Stuart, E. A. (2010). Matching methods for causal inference: A review and a look forward. Statistical Science , 25(1):1--21

    2010

  41. [49]

    Su, L., Murtazashvili, I., and Ullah, A. (2013). Local linear gmm estimation of functional coefficient iv models with an application to estimating the rate of return to schooling. Journal of Business & Economic Statistics , 31(2):184--207

    2013

  42. [50]

    M., and Li, B

    Su, X., Tsai, C.-L., Wang, H., Nickerson, D. M., and Li, B. (2009). Subgroup analysis via recursive partitioning. Journal of Machine Learning Research , 10:141--158

    2009

  43. [51]

    Swanson, S. A. and Hern \'a n, M. A. (2013). Commentary: how to report instrumental variable analyses (suggestions welcome). Epidemiology , 24(3):370--374

    2013

  44. [52]

    Swanson, S. A. and Hern \'a n, M. A. (2014). Think globally, act globally: an epidemiologist’s perspective on instrumental variable estimation. Statistical Science , 29(3):371--374

    2014

  45. [53]

    Therneau, T., Atkinson, B., and Ripley, B. (2015). Package ‘rpart’ . R package version 4.1-15. https://cran.r-project.org/package=rpart

    2015

  46. [54]

    and Athey, S

    Wager, S. and Athey, S. (2018). Estimation and inference of heterogeneous treatment effects using random forests. Journal of the American Statistical Association , 113(523):1228--1242

    2018

  47. [55]

    Yu, R. (2019). bigmatch: Making Optimal Matching Size-Scalable Using Optimal Calipers . R package version 0.6.1. https://CRAN.R-project.org/package=bigmatch

    2019

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.