Pith. sign in

REVIEW 3 major objections 4 minor 98 references

Doubly Robust Inference on Causal Derivative Effects for Continuous Treatments

T0 review · 3 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read This paper constructs kernel-smoothed, doubly robust estimators for the derivative of the causal dose-response curve and proves they are asymptotically normal at the nonparametric rate $\sqrt{nh^3}$, with and without the positivity…

desk verdict Solid DR inference for the derivative of the dose-response curve under positivity; the no-positivity half is honest about its additive-model assumption in the body but the abstract overstates its causal scope. read the letter →

arxiv 2501.06969 v2 pith:IX45YXAA submitted 2025-01-12 stat.ME econ.EMmath.STstat.MLstat.TH

Yikun Zhang , Yen-Chi Chen This is my paper · ORCID
classification stat.MEecon.EMmath.STstat.MLstat.TH MSC 62G0562G0862G2062D20
keywords causalinferencedose-responsecurvederivativeestimationpositivityconditionkernelsmoothingdoublyrobustcontinuoustreatmentssupport
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

Most causal inference with continuous treatments estimates the dose-response curve $m(t) = \mathbb{E}[Y(t)]$, the average outcome under each treatment level. This paper argues that the derivative $\theta(t) = \frac{d}{dt}\mathbb{E}[Y(t)]$ is the quantity that actually signals treatment effect: two treatment levels can have the same expected outcome while the effect of one more unit of treatment differs between them, and the average derivative $\mathbb{E}[\theta(T)]$ can vanish even when $\theta(t)$ is nonzero. The paper constructs kernel-smoothed inverse-probability-weighted and doubly robust estimators of $\theta(t)$ and proves they are consistent and asymptotically normal at the nonparametric rate $\sqrt{nh^3}$, provided either the outcome model or the treatment-density model is correctly specified. It then treats the harder case where the positivity condition fails, showing that standard IPW and doubly robust estimators are inconsistent there and introducing bias-corrected versions that recover the same rate and asymptotic normality under an additive confounding structure. The result matters because it makes the derivative effect a testable, reportable object at each treatment level, including in observational settings where positivity fails and some covariate groups never receive certain doses.

What carries the argument

The workhorse is the doubly robust kernel estimator (8), which writes the derivative estimate as an average of a local-polynomial residual: each observation contributes $\frac{(T_i - t)/h \cdot K((T_i - t)/h)}{h \kappa_2 \widehat{p}_{T|S}(T_i|S_i)}\left(Y_i - \widehat{\mu}(t,S_i) - (T_i - t)\widehat{\beta}(t,S_i)\right)$ plus a regression adjustment $h\widehat{\beta}(t,S_i)$, with $\kappa_2 = \int u^2 K(u)\,du$. The local-polynomial correction pushes the inverse-probability residual to second order so that the Neyman orthogonality underlying double robustness holds as $h \to 0$. Without positivity, the central object is the modified IPW quantity (21), where the weight $1/p_{T|S}(T|S)$ is replaced by $p_\zeta(S|t)/p(T,S)$ using the $\zeta$-interior conditional density $p_\zeta$: the conditional density of $S$ given $T = t$ restricted to either the shrunken support $\mathcal{S}(t)\ominus\zeta$ or the level set $L_\zeta(t) = \{s : p_{S|T}(s|t) \geq \zeta\}$. Assumption A6 ensures the interior sits inside $\mathcal{S}(t+\delta)$ for small $\delta$, so the boundary discrepancy disappears; nuisance functions are estimated on independent folds via cross-fitting, and inference uses the sample variance of the influence function (9) or a multiplier bootstrap for uniform bands.

What would settle it

Generate data from design (28) with an added interaction, $Y = T^3 + T^2 + 10S + \gamma\,T\,S + \varepsilon$, so that model (13) fails while positivity still fails; if the bias-corrected DR estimator (26) develops a bias proportional to $\gamma$ that does not vanish as $n$ grows, the additive structural assumption is shown to be indispensable. A check of the positivity-half asymptotics: with the density model correct and the outcome model misspecified, the empirical bias of $\widehat{\theta}_{\mathrm{DR}}(t)$ should equal $h^2 B_\theta(t)$ with $B_\theta(t) = (\kappa_4/6\kappa_2)\cdot\mathbb{E}_S[\partial^3/\partial t^3\,\mu(t,S)]$ to leading order, so a detectable mismatch at $h \asymp n^{-1/5}$ would refute the claimed rate and bias expansion.

Watch

Extended reading notes

Core claim

On the paper's own terms the central discovery is an estimator together with a theorem about it. Under the identification conditions A1, smoothness conditions A3–A5, and positivity A2, the doubly robust estimator (8) satisfies, for fixed $t$ and bandwidth $h$ with $nh^7 \to c_3 \geq 0$, $\sqrt{nh^3}\big(\widehat{\theta}_{\mathrm{DR}}(t) - \theta(t) - h^2 B_\theta(t)\big) \to N(0, V_\theta(t))$, where $B_\theta(t)$ is an explicit bias term and $V_\theta(t)$ is the variance of the influence-function contribution $\phi_{h,t}$. The double robustness means the limit holds if either the outcome model $(\mu, \beta)$ is correct or the conditional density model $p_{T|S}$ is correct. When positivity fails, the paper proves that even oracle IPW estimators do not converge to $\theta(t)$: they converge instead to $\bar{m}'(t)\cdot \rho(t)$, where $\rho(t) = P(S \in \mathcal{S}(t))$ is the probability that the covariates fall in the conditional support of $S$ given $T = t$, because of the support discrepancy between $\mathcal{S}(t+uh)$ and $\mathcal{S}(t)$. The bias-corrected IPW and DR estimators (25)–(26) multiply by the density ratio $p_{S|T}(S|t)/p_S(S)$ and restrict to a $\zeta$-interior of the conditional support, which removes the boundary discrepancy and restores $\sqrt{nh^3}$-asymptotic normality with the same double robustness (Theorem 6), under the additive confounding model (13). The paper also shows the estimator attains the nonparametric efficiency bound of a smoothly approximated functional, a necessary detour because $\theta(t)$ itself lacks a pathwise derivative.

Load-bearing premise

For the no-positivity half, identification of the derivative effect rests on the additive confounding model $Y = \bar{m}(T) + \eta(S) + \varepsilon$ (equation 13), which forbids any interaction between treatment and covariates; if the true outcome contains a $T$-by-$S$ interaction, the bias-corrected IPW and DR estimators converge to a different quantity than the causal derivative $\theta(t)$.

Editorial extensions

If this is right

  • Choosing the bandwidth $h$ of order $n^{-1/5}$ makes the leading bias $h^2 B_\theta(t)$ asymptotically negligible, so the Wald intervals $\widehat{\theta}_{\mathrm{DR}}(t) \pm q_{1-\tau/2}\sqrt{\widehat{V}_\theta(t)/(nh^3)}$ have valid coverage at the standard nonparametric rate.
  • The rate conditions in Theorem 1 are satisfied by flexible machine-learned nuisance estimates, so the method can be used with neural-network fits of $\mu$, $\beta$, and $p_{T|S}$ under cross-fitting without changing the asymptotic theory.
  • Integrating the bias-corrected derivative estimators yields IPW and DR estimators of the dose-response curve $m(t)$ that remain consistent when positivity fails, extending the regression-adjustment construction to the full doubly robust family.
  • The multiplier bootstrap version gives uniform confidence bands over the whole treatment range, allowing simultaneous statements about where the derivative curve is significantly nonzero.

Reading between the lines

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

  • An automatic choice of the trimming level $\zeta$ is left open; the level-set threshold $0.5\cdot\max \widehat{p}$ used in the paper suggests a tunable bias-variance trade-off, and cross-validating $\zeta$ against the estimated variance rather than fixing the multiplier is a natural extension.
  • Proposition 3's bias formula implies a practical diagnostic: when positivity is doubtful, reporting $\rho(t) = P(S \in \mathcal{S}(t))$ alongside the estimate tells the user how far a naive IPW analysis is from the true derivative effect.
  • The boundary-correction idea transfers to other non-regular targets with partial overlap, such as the continuous-instrument DR estimator of Zeng et al. (2025), which the paper notes does not address positivity violations, or average derivative effects under trimmed support.
  • A misspecification test of the additive model (13), for instance checking whether the residual $\mathbb{E}[Y - \bar{m}(T) - \eta(S) \mid T, S]$ carries a $T\cdot S$ interaction, would delineate exactly where the no-positivity guarantee ends.
Share X Bluesky LinkedIn Reddit HN

Formalized claims in Lean

  1. Claim #1: The proposed DR estimator is asymptotically normal at the rate sqrt(n h^3) under positivity and either correct outcome or density model.

  2. Claim #2: Without positivity, conventional IPW estimators converge to the support-discrepancy limit instead of theta(t).

  3. Claim #3: Under additive confounding and without positivity, the bias-corrected DR estimator recovers sqrt(n h^3) asymptotic normality.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper develops kernel-smoothing estimators for the causal derivative effect θ(t)=d/dt E[Y(t)] with a continuous treatment. Under a positivity assumption, it proposes RA, IPW, and DR estimators of θ(t); Theorem 1 gives sqrt(nh^3)-asymptotic normality for the DR estimator, with double robustness in either the outcome model (μ and β) or the conditional density model. Without positivity, the paper shows that conventional IPW and DR estimators are inconsistent, imposes an additive confounding model of the form Y=mbar(T)+η(S)+ε, and proposes bias-corrected IPW and DR estimators based on an interior conditional density; Theorem 6 establishes analogous asymptotic normality under further rate conditions. Simulations and a Job Corps case study illustrate the methods.

Significance. Direct inference on the derivative of the dose-response curve is a useful and underexplored target, and the positivity-half of the paper is technically solid: the DR estimator is derived from the influence function of a smoothed functional, the proofs contain explicit remainder analyses, and the double-robustness claim is supported under the stated conditions. The no-positivity half is a valuable first step that connects causal inference to support and level-set estimation, but its causal scope is conditional on the additive confounding model (13), and the rate conditions for the interior-density estimator are not fully resolved for general covariate dimension. The paper also ships reproducible code and a detailed appendix, which strengthens its value.

major comments (3)
  1. [Abstract; Section 4.1.1; Section 5.1] The abstract and the contribution list advertise bias-corrected IPW and DR estimators 'without positivity' without qualification. The identification in Section 4.1.1 requires the additive confounding model (13), and Proposition 5's limiting target is mbar'(t) under that model. For a non-additive outcome such as μ(t,s)=mbar(t)+η(s)+τ t s with a positivity violation, the modified IPW quantity (21) converges to ∫∂_t μ(t,s) pbarζ(s|t) ds, which is not the causal derivative θ(t). Please rephrase the abstract and the contribution list to state explicitly that the no-positivity results require the additive structural model.
  2. [Theorem 6; Remark 8] Theorem 6(a) requires sqrt(nh^3) ||bpζ(S|t)-pbarζ(S|t)||_{L2}=o(1). With the recommended bandwidth h ≍ n^{-1/5}, this is an L2 convergence rate of o(n^{-1/5}) for the interior conditional density estimator. The typical nonparametric rate O(n^{-2/(5+d)}) quoted in Remark 8 satisfies this requirement only when d<5. The remark's assertion that pbarζ can be viewed as a 'smooth surrogate' and that bpζ can be constructed with a smaller bandwidth is not supported by a concrete estimator or proof. As stated, Theorem 6's applicability to general fixed d is not established. Please provide an explicit construction of pbarζ and bpζ satisfying the rate, or restrict the theorem (and the advertised claims) to dimensions where such rates are available.
  3. [Section 6.1] The high-dimensional simulation (d=20) may not lie in the theoretical regime of Theorem 1. With h ≍ n^{-1/5}, condition (c) of Theorem 1 requires sqrt(nh) = n^{2/5} times the product of nuisance errors to be o(1). Using the neural-network rates O(n^{-2/(4+d)}) cited in Remark 8, the product does not vanish for d=20. The d=20 experiment can still be reported as an empirical illustration, but the paper should either verify the rate conditions for the implemented estimators or explicitly state that this simulation is outside the scope of the theorem.
minor comments (4)
  1. [Section 4.1.1] There is a duplicated word in the sentence defining β(t,s): 'estimators of of β(t, s)' should read 'estimators of β(t, s)'.
  2. [Theorem 1] Condition (b)(i) in the main-text statement says 'with only h ||βbar(t,S)-β(t,S)||_{L2} → 0'; since the same condition already assumes βbar=β, this is vacuous. The appendix version, which instead requires h times the estimation error of bβ to βbar to vanish, is the intended condition and should be stated consistently in the main text.
  3. [Figure 1] The vertical axis labels in the right-hand panel appear as '(t)' without the Greek symbol; please ensure the derivative-effect curve is labeled consistently with θ(t).
  4. [Section 6.2] The no-positivity simulation uses S ∈ R (d=1), which is consistent with the rate discussion in Remark 8. It would be helpful to note this explicitly when summarizing the simulation evidence for Theorem 6.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the DR estimators are derived from influence-function/local-polynomial expansions and verified against oracle IPW/RA benchmarks; the no-positivity results rest on an explicitly stated additive structural model whose identification is re-proved in Proposition C.1.

full rationale

The paper's derivation chain is self-contained rather than circular. Under positivity, the DR estimator (8) is constructed from a local-polynomial-approximated IPW term plus an RA term, and Theorem 1's asymptotic linear form, variance Vθ(t), and bias Bθ(t) are obtained by explicit Taylor expansions of the oracle quantities (Section F.3), not by assuming the conclusion. The double robustness property follows from a standard decomposition in which the first-order bias contains a product of the density error and the outcome-model error (Term X and Term XI of the proof), which vanishes under either correct specification; this is a derived property, not a relabeled fit. The efficiency claim (Theorem 2) is explicitly relative to the smoothed functional ϖ_{h,t}(P), and the paper states that the target θ(t) lacks a pathwise derivative, so the efficiency statement is honestly scoped rather than tautological. In the no-positivity half, the identification of θ(t)=m̄'(t) is imported from the authors' own prior work (Zhang et al., 2024) via the additive confounding model (13), but the paper states (13) as an explicit structural assumption, re-proves the identification theory in Proposition C.1, and grounds the assumption in external spatial-statistics literature (Paciorek, 2010; Schnell and Papadogeorgou, 2020). The bias-corrected estimators (25)-(26) are shown consistent for m̄'(t) by direct expectation calculations (Propositions 4 and 5) whose integrals use μ(t,s)=m̄(t)+η(s), so the conclusion tracks the stated assumption rather than being forced by construction. The liberty taken is in the abstract, which advertises bias-corrected IPW/DR estimators 'when the positivity condition is violated' without mentioning that their causal interpretation requires (13); the body (Sections 4.1.1 and 5) and Remark 8 qualify this clearly. That is a scope/overclaim concern ('correctness risk'), not circularity, because the paper never defines θ(t) in terms of the estimator and never relabels a fitted parameter as a prediction. The remaining self-citations (Zhang et al. 2024 for the additive identification; Fan et al. 2022 for uniform bootstrap validity) are provenance for externally checkable assumptions and re-derived results, so they are not load-bearing reductions. Score 1 reflects the self-citation provenance for the no-positivity identification while confirming the central DR derivations are independent.

Assumptions & free parameters 2 free parameters · 8 assumptions · 0 invented entities

The core theoretical results rely on standard causal and smoothness assumptions plus the additive confounding model for the no-positivity setting. The bandwidth and zeta are tuning parameters. No new physical or conceptual entities are postulated.

free parameters (2)
  • Kernel bandwidth h = h = C_h sigma_T n^{-1/5} in simulations; theory allows h ~ n^{-1/5} for inference
    Bandwidth is a tuning parameter chosen by a scaling rule in simulations; the asymptotic theory requires an undersmoothing rate but does not pin down the constant.
  • Interior trimming parameter zeta = zeta = 0.5 max_i bp_{S|T}(S_i|t) in simulations
    Controls the interior conditional density for the no-positivity estimators. The paper recommends a heuristic threshold multiplier of 0.5 and notes users may adjust it.
assumptions (8)
  • domain assumption Assumption A1(a,b): consistency and unconfoundedness Y(t) independent of T given S
    Standard causal identification conditions for dose-response curves; invoked in Section 2.
  • domain assumption Assumption A1(c): positive conditional variance of T given S
    Ensures nontrivial support of (T,S) and is needed for identification, as noted in Section 2.
  • domain assumption Assumption A1(d): interchangeability of expectation and differentiation
    Required for the RA estimator and the identification of θ(t) as E[d/dt μ(t,S)]; a regularity condition on μ.
  • domain assumption Assumption A2: positivity of the conditional density p_{T|S}
    Central to the Section 3 results; the paper later relaxes this with an additive model.
  • standard math Assumptions A3-A5: smoothness of μ, densities, and regular kernel properties
    Common nonparametric smoothing conditions used throughout the proofs.
  • domain assumption Additive confounding model Y = mbar(T) + eta(S) + epsilon
    Used in Section 4.1.1 to maintain identification when positivity fails; a strong structural restriction that is not mentioned in the abstract.
  • domain assumption Assumption A6: Lipschitz-type support smoothness for S(t)
    Ensures the interior conditional density support remains inside S(t+delta), a key condition for the bias-corrected estimators.
  • domain assumption Nuisance rate conditions in Theorem 1(c) and Theorem 6(c)
    The asymptotic normality requires product-rate convergence of estimated nuisances; these conditions are stated but may be hard to verify in high-dimensional simulations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Doubly Robust Inference on Causal Derivative Effects for Continuous Treatments." pith.science (2026). https://pith.science/paper/IX45YXAA

@misc{pith2026250106969,
  author       = {Pith},
  title        = {Pith review of: Doubly Robust Inference on Causal Derivative Effects for Continuous Treatments},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IX45YXAA}},
  note         = {Machine review of arXiv:2501.06969}
}
read the original abstract

Statistical methods for causal inference with continuous treatments mainly focus on estimating the mean potential outcome function, commonly known as the dose-response curve. However, it is often not the dose-response curve but its derivative function that signals the treatment effect. In this paper, we investigate nonparametric inference on the derivative of the dose-response curve with and without the positivity condition. Under the positivity and other regularity conditions, we propose a doubly robust (DR) inference method for estimating the derivative of the dose-response curve using kernel smoothing. When the positivity condition is violated, we demonstrate the inconsistency of conventional inverse probability weighting (IPW) and DR estimators, and introduce novel bias-corrected IPW and DR estimators. In all settings, our DR estimator achieves asymptotic normality at the standard nonparametric rate of convergence with nonparametric efficiency guarantees. Additionally, our approach reveals an interesting connection to nonparametric support and level set estimation problems. Finally, we demonstrate the applicability of our proposed estimators through simulations and a case study of evaluating a job training program.

Figures

Figures reproduced from arXiv: 2501.06969 by the authors.

Figure 1
Figure 1. Illustrations of a dose-response curve m(t), its corresponding derivative effect curve θ(t), and average derivative effects E [θ(T)]. Here, m(t) is symmetric with respect to t = 3.5 so that m(t1) = m(t2) and E [θ(T)] = 0 but θ(t1) > 0 > θ(t2). Y (t) be the potential outcome (Rubin, 1974) that would have been observed under treatment level T = t. The focus of this work is the (causal) derivative effect curve t 7→ θ(t… view at source ↗
Figure 2
Figure 2. Graphical illustrations of the support discrepancy between [PITH_FULL_IMAGE:figures/full_fig_p018_2.png] view at source ↗
Figure 3
Figure 3. Comparisons between our proposed estimators and the finite-difference approaches by [PITH_FULL_IMAGE:figures/full_fig_p022_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: Comparisons between our bias-corrected estimators (NP) in [PITH_FULL_IMAGE:figures/full_fig_p023_4.png]
Figure 5
Figure 5. Figure 5: Estimated derivative effect curves with 95% confidence intervals using our proposed [PITH_FULL_IMAGE:figures/full_fig_p024_5.png]
Figure 6
Figure 6. Figure 6: Comparisons between our proposed estimators and the finite-difference approaches by [PITH_FULL_IMAGE:figures/full_fig_p039_6.png]
Figure 7
Figure 7. Figure 7: Comparisons between our proposed estimators and the finite-difference approaches by [PITH_FULL_IMAGE:figures/full_fig_p040_7.png]
Figure 8
Figure 8. Figure 8: Comparisons between our bias-corrected estimators (NP) in [PITH_FULL_IMAGE:figures/full_fig_p041_8.png]
Figure 9
Figure 9. Figure 9: Comparisons between our bias-corrected estimators (NP) in [PITH_FULL_IMAGE:figures/full_fig_p042_9.png]
Figure 10
Figure 10. Figure 10: Estimated derivative effect curves with 95% confidence intervals using our proposed [PITH_FULL_IMAGE:figures/full_fig_p043_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

98 extracted references · 60 canonical work pages

  1. [1]

    Bang and J

    H. Bang and J. M. Robins. Doubly robust estimation in missing data and causal inference models. Biometrics, 61 0 (4): 0 962--973, 2005

    2005

  2. [2]

    A. G. Baydin, B. A. Pearlmutter, A. A. Radul, and J. M. Siskind. Automatic differentiation in machine learning: a survey. Journal of Machine Learning Research, 18 0 (153): 0 1--43, 2018

    2018

  3. [3]

    Bickel, C

    P. Bickel, C. Klaassen, Y. Ritov, and J. Wellner. Efficient and Adaptive Estimation for Semiparametric Models. Springer New York, 1998

    1998

  4. [4]

    Blondel and V

    M. Blondel and V. Roulet. The elements of differentiable programming. arXiv preprint arXiv:2403.14606, 2024

    arXiv 2024

  5. [5]

    Local causal effects with continuous exposures: A matching estimator for the average causal derivative effect

    S. Bong and K. Lee. Local causal effects with continuous exposures: A matching estimator for the average causal derivative effect. arXiv preprint arXiv:2311.18532, 2023

    work page Pith review arXiv 2023

  6. [6]

    Bonvini and E

    M. Bonvini and E. H. Kennedy. Fast convergence rates for dose-response estimation. arXiv preprint arXiv:2207.11825, 2022

    arXiv 2022

  7. [7]

    Bonvini, A

    M. Bonvini, A. McClean, Z. Branson, and E. H. Kennedy. Incremental causal effects: an introduction and review. In Handbook of matching and weighting adjustments for causal inference, pages 349--372. Chapman and Hall/CRC, 2023

    2023

  8. [8]

    Branson, E

    Z. Branson, E. H. Kennedy, S. Balakrishnan, and L. Wasserman. Causal effect estimation after propensity score trimming with continuous treatments. arXiv preprint arXiv:2309.00706, 2023

    arXiv 2023

Show all 98 references
  1. [9]

    B. Cadre. Kernel estimation of density level sets. Journal of Multivariate Analysis, 97 0 (4): 0 999--1023, 2006

    2006

  2. [10]

    Calonico, M

    S. Calonico, M. D. Cattaneo, and M. H. Farrell. On the effect of bias estimation on coverage accuracy in nonparametric inference. Journal of the American Statistical Association, 113 0 (522): 0 767--779, 2018

    2018

  3. [11]

    Carone, A

    M. Carone, A. R. Luedtke, and M. J. van der Laan. Toward computerized efficient estimation in infinite-dimensional models. Journal of the American Statistical Association, 114 0 (527): 0 1174--1190, 2019

    2019

  4. [12]

    M. D. Cattaneo, R. K. Crump, and M. Jansson. Robust data-driven inference for density-weighted average derivatives. Journal of the American Statistical Association, 105 0 (491): 0 1070--1083, 2010

    2010

  5. [13]

    Chen and Z

    X. Chen and Z. Liao. Sieve m inference on irregular parameters. Journal of Econometrics, 182 0 (1): 0 70--86, 2014

    2014

  6. [14]

    X. Chen, Z. Liao, and Y. Sun. Sieve inference on possibly misspecified semi-nonparametric time series models. Journal of Econometrics, 178: 0 639--658, 2014

    2014

  7. [15]

    Cheng and Y.-C

    G. Cheng and Y.-C. Chen. Nonparametric inference via bootstrapping the debiased estimator . Electronic Journal of Statistics, 13 0 (1): 0 2194 -- 2256, 2019

    2019

  8. [16]

    Chernozhukov, D

    V. Chernozhukov, D. Chetverikov, M. Demirer, E. Duflo, C. Hansen, W. Newey, and J. Robins. Double/debiased machine learning for treatment and structural parameters . The Econometrics Journal, 21 0 (1): 0 C1--C68, 01 2018

    2018

  9. [17]

    Chernozhukov, W

    V. Chernozhukov, W. K. Newey, and R. Singh. Automatic debiased machine learning of causal and structural effects. Econometrica, 90 0 (3): 0 967--1027, 2022

    2022

  10. [18]

    Colangelo and Y.-Y

    K. Colangelo and Y.-Y. Lee. Double debiased machine learning nonparametric inference with continuous treatments. arXiv preprint arXiv:2004.03036, 2020

    2004 arXiv

  11. [19]

    S. R. Cole and M. A. Hern \'a n. Constructing inverse probability weights for marginal structural models. American Journal of Epidemiology, 168 0 (6): 0 656--664, 2008

    2008

  12. [20]

    A. Cuevas. Set estimation: Another bridge between statistics and geometry. Bolet \' n de Estad \' stica e Investigaci \'o n Operativa , 25 0 (2): 0 71--85, 2009

    2009

  13. [21]

    Cuevas and R

    A. Cuevas and R. Fraiman. A plug-in approach to support estimation . The Annals of Statistics, 25 0 (6): 0 2300 -- 2312, 1997

    1997

  14. [22]

    Devroye and G

    L. Devroye and G. L. Wise. Detection of abnormal behavior via nonparametric estimation of the support. SIAM Journal on Applied Mathematics, 38 0 (3): 0 480--488, 1980

    1980

  15. [23]

    D \' az and N

    I. D \' az and N. S. Hejazi. Causal mediation analysis for stochastic interventions. Journal of the Royal Statistical Society Series B: Statistical Methodology, 82 0 (3): 0 661--683, 2020

    2020

  16. [24]

    D \' az and M

    I. D \' az and M. J. van der Laan. Targeted data adaptive estimation of the causal dose--response curve. Journal of Causal Inference, 1 0 (2): 0 171--192, 2013

    2013

  17. [25]

    Einmahl and D

    U. Einmahl and D. M. Mason. Uniform in bandwidth consistency of kernel-type function estimators . The Annals of Statistics, 33 0 (3): 0 1380 -- 1403, 2005

    2005

  18. [26]

    Fan and I

    J. Fan and I. Gijbels. Local polynomial modelling and its applications, volume 66. Chapman & Hall/CRC, 1996

    1996

  19. [27]

    Fan, Y.-C

    Q. Fan, Y.-C. Hsu, R. P. Lieli, and Y. Zhang. Estimation of conditional average treatment effects with high-dimensional data. Journal of Business & Economic Statistics, 40 0 (1): 0 313--327, 2022

    2022

  20. [28]

    M. H. Farrell, T. Liang, and S. Misra. Deep neural networks for estimation and inference. Econometrica, 89 0 (1): 0 181--213, 2021

    2021

  21. [29]

    C. Flores. Estimation of dose-response functions and optimal doses with a continuous treatment. Technical report, Department of Economics, University of Miami, 2007. URL https://core.ac.uk/download/pdf/7169663.pdf

    2007

  22. [30]

    C. A. Flores and A. Flores-Lagunes. Identification and estimation of causal mechanisms and net effects of a treatment under unconfoundedness. IZA Discussion Papers 4237, Institute of Labor Economics (IZA), 2009

    2009

  23. [31]

    C. A. Flores, A. Flores-Lagunes, A. Gonzalez, and T. C. Neumann. Estimating the effects of length of exposure to instruction in a training program: The case of job corps. Review of Economics and Statistics, 94 0 (1): 0 153--171, 2012

    2012

  24. [32]

    A. F. Galvao and L. Wang. Uniformly semiparametric efficient estimation of treatment effects with a continuous treatment. Journal of the American Statistical Association, 110 0 (512): 0 1528--1542, 2015

    2015

  25. [33]

    Gasser and H.-G

    T. Gasser and H.-G. M \"u ller. Estimating regression functions and their derivatives by the kernel method. Scandinavian Journal of Statistics, pages 171--185, 1984

    1984

  26. [34]

    R. D. Gill and J. M. Robins. Causal inference for complex longitudinal data: the continuous case. Annals of Statistics, 29 0 (6): 0 1785--1811, 2001

    2001

  27. [35]

    Godambe and V

    V. Godambe and V. Joshi. Admissibility and bayes estimation in sampling finite populations. i. The Annals of Mathematical Statistics, 36 0 (6): 0 1707--1722, 1965

    1965

  28. [36]

    Z. Guo, W. Yuan, and C.-H. Zhang. Decorrelated local linear estimator: Inference for non-linear effects in high-dimensional additive models. arXiv preprint arXiv:1907.12732, 2019

    1907 arXiv

  29. [37]

    H \"a rdle and T

    W. H \"a rdle and T. M. Stoker. Investigating smooth multiple regression by the method of average derivatives. Journal of the American Statistical Association, 84 0 (408): 0 986--995, 1989

    1989

  30. [38]

    J. D. Hart and P. Vieu. Data-driven bandwidth choice for density estimation based on dependent data. The Annals of Statistics, pages 873--890, 1990

    1990

  31. [39]

    Hines, K

    O. Hines, K. Diaz-Ordaz, and S. Vansteelandt. Optimally weighted average derivative effects. arXiv preprint arXiv:2308.05456, 2023

    2023 arXiv

  32. [40]

    Hirano and G

    K. Hirano and G. W. Imbens. The Propensity Score with Continuous Treatments, chapter 7, pages 73--84. John Wiley & Sons, Ltd, 2004

    2004

  33. [41]

    D. A. Hirshberg and S. Wager. Debiased inference of average partial effects in single-index models: Comment on wooldridge and zhu. Journal of Business & Economic Statistics, 38 0 (1): 0 19--24, 2020

    2020

  34. [42]

    M. Huber. Identifying causal mechanisms (primarily) based on inverse probability weighting. Journal of Applied Econometrics, 29 0 (6): 0 920--943, 2014

    2014

  35. [43]

    Huber, Y.-C

    M. Huber, Y.-C. Hsu, Y.-Y. Lee, and L. Lettry. Direct and indirect effects of continuous treatments based on generalized propensity score weighting. Journal of Applied Econometrics, 35 0 (7): 0 814--840, 2020

    2020

  36. [44]

    Ichimura and W

    H. Ichimura and W. K. Newey. The influence function of semiparametric estimators. Quantitative Economics, 13 0 (1): 0 29--61, 2022

    2022

  37. [45]

    Imai and D

    K. Imai and D. A. van Dyk. Causal inference with general treatment regimes: Generalizing the propensity score. Journal of the American Statistical Association, 99 0 (467): 0 854--866, 2004

    2004

  38. [46]

    Kallus and A

    N. Kallus and A. Zhou. Policy evaluation and optimization with continuous treatments. In International Conference on Artificial Intelligence and Statistics, pages 1243--1251. PMLR, 2018

    2018

  39. [47]

    E. H. Kennedy. Semiparametric doubly robust targeted double machine learning: a review. Handbook of Statistical Methods for Precision Medicine, pages 207--236, 2024

    2024

  40. [48]

    E. H. Kennedy, Z. Ma, M. D. McHugh, and D. S. Small. Nonparametric methods for doubly robust estimation of continuous treatment effects. Journal of the Royal Statistical Society Series B: Statistical Methodology, 79 0 (4): 0 1229--1245, 2017

    2017

  41. [49]

    S. Klosin. Automatic double machine learning for continuous treatment effects. arXiv preprint arXiv:2104.10334, 2021

    2021 arXiv

  42. [50]

    D. S. Lee. Training, wages, and sample selection: Estimating sharp bounds on treatment effects. The Review of Economic Studies, 76 0 (3): 0 1071--1102, 2009

    2009

  43. [51]

    Y.-Y. Lee. Partial mean processes with generated regressors: Continuous treatment effects and nonseparable models. arXiv preprint arXiv:1811.00157, 2018

    2018 arXiv

  44. [52]

    Lee and C.-A

    Y.-Y. Lee and C.-A. Liu. Lee bounds with a continuous treatment in sample selection. arXiv preprint arXiv:2411.04312, 2024

    2024

  45. [53]

    E. L. Lehmann. Elements of large-sample theory. Springer, 1999

    1999

  46. [54]

    Li and J

    Q. Li and J. Racine. Cross-validated local linear nonparametric regression. Statistica Sinica, 14: 0 485--512, 2004

    2004

  47. [55]

    A. Luedtke. Simplifying debiased inference via automatic differentiation and probabilistic programming. arXiv preprint arXiv:2405.08675, 2024

    2024 arXiv

  48. [56]

    Luedtke and I

    A. Luedtke and I. Chung. One-step estimation of differentiable hilbert-valued parameters. The Annals of Statistics, 52 0 (4): 0 1534--1563, 2024

    2024

  49. [57]

    Mack and H.-G

    Y. Mack and H.-G. M \"u ller. Derivative estimation in nonparametric regression with random predictor variable. Sankhy \=a : The Indian Journal of Statistics, Series A , pages 59--72, 1989

    1989

  50. [58]

    McClean, Y

    A. McClean, Y. Li, S. Bae, M. A. McAdams-DeMarco, I. D \' az, and W. Wu. Fair comparisons of causal parameters with many treatments and positivity violations. arXiv preprint arXiv:2410.13522, 2024

    2024

  51. [59]

    Meier, S

    L. Meier, S. van de Geer, and P. B \"u hlmann. High-dimensional additive modeling . The Annals of Statistics, 37 0 (6B): 0 3779 -- 3821, 2009

    2009

  52. [60]

    J. Meloche. Asymptotic behaviour of the mean integrated squared error of kernel density estimators for dependent observations. The Canadian Journal of Statistics/La Revue Canadienne de Statistique, pages 205--211, 1990

    1990

  53. [61]

    Neugebauer and M

    R. Neugebauer and M. van der Laan. Nonparametric causal effects based on marginal structural models. Journal of Statistical Planning and Inference, 137 0 (2): 0 419--434, 2007

    2007

  54. [62]

    W. K. Newey and J. R. Robins. Cross-fitting and fast remainder rates for semiparametric estimation. arXiv preprint arXiv:1801.09138, 2018

    2018 arXiv

  55. [63]

    W. K. Newey and T. M. Stoker. Efficiency of weighted average derivative estimators and index models. Econometrica, 61 0 (5): 0 1199--1223, 1993

    1993

  56. [64]

    J. Neyman. Optimal asymptotic tests of composite hypotheses. Probability and Statsitics, pages 213--234, 1959

    1959

  57. [65]

    J. Neyman. C( ) tests and their use. Sankhy \=a : The Indian Journal of Statistics, Series A , 41 0 (1/2): 0 1--21, 1979

    1979

  58. [66]

    C. J. Paciorek. The importance of scale for spatial-confounding bias and precision of spatial regression estimators. Statistical Science, 25 0 (1): 0 107--125, 2010

    2010

  59. [67]

    Paszke, S

    A. Paszke, S. Gross, S. Chintala, G. Chanan, E. Yang, Z. DeVito, Z. Lin, A. Desmaison, L. Antiga, and A. Lerer. Automatic differentiation in pytorch. In NIPS 2017 Workshop on Autodiff, 2017

    2017

  60. [68]

    Paszke, S

    A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga, A. Desmaison, A. Kopf, E. Yang, Z. DeVito, M. Raison, A. Tejani, S. Chilamkurthy, B. Steiner, L. Fang, J. Bai, and S. Chintala. Pytorch: An imperative style, high-per...

    2019

  61. [69]

    J. L. Powell, J. H. Stock, and T. M. Stoker. Semiparametric estimation of index coefficients. Econometrica, 57 0 (6): 0 1403--1430, 1989

    1989

  62. [70]

    Ratkovic and D

    M. Ratkovic and D. Tingley. Estimation and inference on nonlinear and heterogeneous effects. The Journal of Politics, 85 0 (2): 0 421--435, 2023

    2023

  63. [71]

    J. Robins. A new approach to causal inference in mortality studies with a sustained exposure period—application to control of the healthy worker survivor effect. Mathematical Modelling, 7 0 (9-12): 0 1393--1512, 1986

    1986

  64. [72]

    J. M. Robins, M. A. Hernan, and B. Brumback. Marginal structural models and causal inference in epidemiology. Epidemiology, 11 0 (5): 0 550--560, 2000

    2000

  65. [73]

    D. B. Rubin. Estimating causal effects of treatments in randomized and nonrandomized studies. Journal of Educational Psychology, 66 0 (5): 0 688--701, 1974

    1974

  66. [74]

    A. Schick. On Asymptotically Efficient Estimation in Semiparametric Models . The Annals of Statistics, 14 0 (3): 0 1139 -- 1151, 1986

    1986

  67. [75]

    Schindl, S

    K. Schindl, S. Shen, and E. H. Kennedy. Incremental effects for continuous exposures. arXiv preprint arXiv:2409.11967, 2024

    2024

  68. [76]

    Schnell and G

    P. Schnell and G. Papadogeorgou. Mitigating unobserved spatial confounding when estimating the effect of supermarket access on cardiovascular disease deaths. Annals of Applied Statistics, 14: 0 2069--2095, 12 2020

    2020

  69. [77]

    P. Z. Schochet, J. Burghardt, and S. Glazerman. National job corps study: The impacts of job corps on participants' employment and related outcomes. Mathematica policy research reports, Mathematica Policy Research, 2001

    2001

  70. [78]

    P. Z. Schochet, J. Burghardt, and S. McConnell. Does job corps work? impact findings from the national job corps study. American Economic Review, 98 0 (5): 0 1864--1886, 2008

    2008

  71. [79]

    J. Shao. Mathematical Statistics. Springer Science & Business Media, 2003

    2003

  72. [80]

    R. M. Stolzenberg. The measurement and decomposition of causal effects in nonlinear and nonadditive models. Sociological Methodology, 11: 0 459--488, 1980

    1980

  73. [81]

    C. J. Stone. Additive regression and other nonparametric models. The Annals of Statistics, 13 0 (2): 0 689--705, 1985

    1985

  74. [82]

    L. Su, T. Ura, and Y. Zhang. Non-separable models with high-dimensional data. Journal of Econometrics, 212 0 (2): 0 646--677, 2019

    2019

  75. [83]

    Swaminathan and T

    A. Swaminathan and T. Joachims. The self-normalized estimator for counterfactual learning. In C. Cortes, N. Lawrence, D. Lee, M. Sugiyama, and R. Garnett, editors, Advances in Neural Information Processing Systems, volume 28, 2015

    2015

  76. [84]

    Takatsu and T

    K. Takatsu and T. Westling. Debiased inference for a covariate-adjusted regression function. Journal of the Royal Statistical Society Series B: Statistical Methodology, 87 0 (1): 0 33--55, 2025

    2025

  77. [85]

    H. F. Trotter and J. W. Tukey. Conditional monte carlo for normal samples. In Symposium on Monte Carlo Methods, pages 64--79. John Wiley and Sons, 1956

    1956

  78. [86]

    A. B. Tsybakov. On nonparametric estimation of density level sets. The Annals of Statistics, 25 0 (3): 0 948--969, 1997

    1997

  79. [87]

    M. J. van der Laan and J. M. Robins. Unified methods for censored longitudinal data and causality. Springer, 2003

    2003

  80. [88]

    M. J. van der Laan, A. Bibaut, and A. R. Luedtke. Cv-tmle for nonpathwise differentiable target parameters. In M. J. van der Laan and S. Rose, editors, Targeted Learning in Data Science: Causal Inference for Complex Longitudinal Studies, pages 455--481. Springer, 2018

    2018

  81. [89]

    van der Vaart

    A. van der Vaart. On differentiable functionals. The Annals of Statistics, 19 0 (1): 0 178--204, 1991

    1991

  82. [90]

    A. W. van der Vaart. Asymptotic Statistics. Cambridge Series in Statistical and Probabilistic Mathematics. Cambridge University Press, 1998

    1998

  83. [91]

    M. P. Wand and M. C. Jones. Kernel Smoothing. CRC press, 1994

    1994

  84. [92]

    Wasserman

    L. Wasserman. All of nonparametric statistics. Springer Science & Business Media, 2006

    2006

  85. [93]

    Westreich and S

    D. Westreich and S. R. Cole. Invited commentary: positivity in practice. American Journal of Epidemiology, 171 0 (6): 0 674--677, 2010

    2010

  86. [94]

    X. Wu, F. Mealli, M.-A. Kioumourtzoglou, F. Dominici, and D. Braun. Matching on generalized propensity scores with continuous exposures. Journal of the American Statistical Association, 119 0 (545): 0 757--772, 2024

    2024

  87. [95]

    Y. Xu, N. Sani, A. Ghassami, and I. Shpitser. Multiply robust causal mediation analysis with continuous treatments. arXiv preprint arXiv:2105.09254, 2021

    2021 arXiv

  88. [96]

    Z. Zeng, A. W. Levis, J. Lee, E. H. Kennedy, and L. Keele. Nonparametric estimation of local treatment effects with continuous instruments. arXiv preprint arXiv:2504.03063, 2025

    2025 arXiv

  89. [97]

    Zhang, Y.-C

    Y. Zhang, Y.-C. Chen, and A. Giessing. Nonparametric inference on dose-response curves without the positivity condition. arXiv preprint arXiv:2405.09003, 2024

    2024 arXiv

  90. [98]

    Zhou and D

    S. Zhou and D. A. Wolfe. On derivative estimation in spline regression. Statistica Sinica, 10 0 (1): 0 93--108, 2000

    2000

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.