Pith. sign in

REVIEW 2 major objections 6 minor 137 references

Optimal Designs with Robust Inference for Binary Treatment Effects

T0 review · 2 major / 6 minor · reviewed 2026-07-11 · grok-4.5

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

Pith's one-line read For binary outcomes under fixed covariates, balanced designs that cancel covariate imbalance make the difference-in-means estimator asymptotically variance-optimal, and a CMH variance estimator is conservative yet tight under local alternat

desk verdict Solid design-plus-inference package for binary outcomes under Neyman nonparametric Setting (b): exact variance, blocking optimality, and a usable CMH variance that is finite-sample conservative and locally tight. read the letter →

arxiv 2607.05768 v1 pith:HLEGA6XY submitted 2026-07-07 stat.ME

David Azriel , Abba M. Krieger , Adam Kapelner This is my paper
classification stat.ME MSC 62K0562K1062G05
keywords binarytreatmenteffectsNeymannonparametricmodeloptimalexperimentaldesignblockingdesignsCochran-Mantel-Haenszelvarianceconservativeinferencelocalalternativescovariatebalance
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 studies randomized two-arm experiments with binary responses when the covariates are already observed and fixed, but the potential outcomes remain random (Neyman's nonparametric model). It derives the exact variance of the usual difference-in-means estimator and shows that any balanced assignment whose covariance matrix drives the quadratic form in the average success probabilities to zero is asymptotically optimal. A broad class of blocking designs meets that condition once the number of blocks grows and the success probabilities are mildly smooth functions of the covariates. Because the exact variance depends on unknown probabilities, unbiased estimation is impossible; the authors therefore introduce a conservative Cochran–Mantel–Haenszel variance estimator that works for any balanced design and prove it becomes asymptotically exact under local alternatives when the design is sufficiently random. Simulations confirm that intermediate blocking (or other designs that combine balance with randomness) paired with this estimator yields higher power and shorter intervals than Wald or extended-Robins procedures while remaining properly sized.

What carries the argument

The exact variance formula Var[τ̂] = (1/n^{2})[(p_T + p_C)^T Σ_W (p_T + p_C) + 2(p_T^T(1-p_T) + p_C^T(1-p_C))] together with the CMH estimator V_CMH = (4/n^{2}) Y^T Σ_W Y, whose expectation exceeds the true variance by a non-negative quadratic form in the individual treatment effects.

What would settle it

Simulate binary outcomes under a local alternative in which half the units have treatment effect +c and half have -c (so average τ is small but (1/n)‖η‖^{2} stays order 1) using optimal blocking with growing B; if n(E[V_CMH] − Var[τ̂]) fails to vanish, the tightness claim is false.

Watch

Extended reading notes

Core claim

Any balanced design satisfying lim (1/n)(p_T + p_C)^T Σ_W (p_T + p_C) = 0 is asymptotically variance-optimal for the difference-in-means estimator; optimal uniform blocking with diverging block number satisfies the condition under Lipschitz response probabilities, and the CMH estimator V_CMH is finite-sample conservative with excess equal to (4/n^{2}) η^T (Σ_W ⊙ Σ_W) η and asymptotically tight under local alternatives whenever λ_max(Σ_W) stays bounded.

Load-bearing premise

Asymptotic tightness of the CMH variance estimator requires that the average squared individual treatment-effect differences also go to zero, not merely that their average shrinks like 1 over square-root n; if large positive and negative individual effects cancel, the estimator stays conservative even asymptotically.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 6 minor

Summary. The paper studies randomized experiments with binary outcomes under Neyman’s nonparametric model (fixed covariates, random independent Bernoulli potential outcomes). It derives the exact variance of the balanced difference-in-means estimator (Eq. 5), shows that any balanced design satisfying the covariate-balance condition (7) is asymptotically variance-optimal, and proves that optimal uniform blocking with B o∞ meets (7) under Lipschitz response probabilities (Theorem 1). Because unbiased variance estimation is impossible, the authors introduce a CMH-based conservative estimator V_CMH for general balanced designs (Theorem 2: E[V_CMH]−Var[τ̂]=(4/n^{2})ηᵀΣ_W^{⊙2}η) and an extension of Robins’ estimator for blocking designs. Under local alternatives with the second-moment condition (15) and bounded λ_max(Σ_W), V_CMH is asymptotically tight and consistent (Theorem 3), yielding asymptotically valid CIs. Simulations (N_sim=10 000) compare designs and inference procedures and support the claim that designs achieving both covariate balance and sufficient randomness perform well with CMH inference.

Significance. If the results hold, the paper supplies a coherent design-and-inference package for binary (incidence) outcomes under the practically relevant Setting (b): asymptotic variance optimality via a transparent balance condition, finite-sample conservative variance estimation that is not restricted to blocking, and local-alternative tightness that justifies CMH-based confidence intervals. The explicit variance formula, the Schur-product conservativeness argument, the suboptimal-cube bound for blocking, and the term-by-term covariance analysis for consistency are clean and reusable. The simulation design (fixed X, many designs including optimal and naïve blocking, rerandomization, greedy matching hybrids) is thorough and illustrates the balance–randomness trade-off. The work therefore advances both the theory of optimal randomization for binary responses and the practical toolkit for conservative yet asymptotically tight inference.

major comments (2)
  1. §2.4, Condition (15) and Theorem 3: asymptotic tightness of V_CMH (and therefore asymptotic validity of the CMH CIs) requires not only √n τ oτ_∞ but also (1/n)‖η‖^{2} o0. The manuscript notes that the condition can fail when individual effects remain large while averaging to a small τ, yet the main claims and the simulation design (logistic-linear model with fixed ‖β‖=3, β_T=0.5) stay inside the regime. A short discussion or numerical illustration of how large the difference term remains when (15) is violated would clarify the practical scope of the “asymptotically tight” claim.
  2. §2.3.1–2.3.2 and Proposition 1: the comparison of V_CMH and V_Robbins-ext is informative for BCRD, but the paper’s recommendation of CMH for general balanced designs rests on finite-sample conservativeness (Theorem 2) while Robbins-ext is only asymptotically conservative and restricted to blocking. The simulations show Robbins-ext badly missized for small blocks; a clearer statement of when (if ever) Robbins-ext is preferred, or an explicit recommendation to prefer CMH except in large-block settings, would strengthen the practical takeaway.
minor comments (6)
  1. Abstract and §1: “incidence outcomes” is used without definition; a brief parenthetical would help non-epidemiology readers.
  2. §2.2: the bound after Theorem 1 assumes Lipschitz constant 1 and unit-cube support; stating that the argument scales with the Lipschitz constant and diameter would make the result more transparent.
  3. §3.1: the Monte-Carlo approximation (18) for Σ_W of GreedyMD / BinaryMatchThenGreedyMD is correct but computationally heavy; a short remark on whether a closed-form or low-rank approximation is feasible would be useful.
  4. Figures 1–2 and Table 1: the U-shape over block size is clearest for p=1; a sentence noting that the pattern is attenuated (but still visible) for p=5,10 would help readers interpret the higher-dimensional panels.
  5. References: a few self-citations (Azriel et al. 2024, 2026; Kapelner et al. 2023, 2025) supply earlier variance expressions; ensuring that the present paper is self-contained for the key formulas (especially Eq. 5) is already largely achieved, but a one-sentence pointer would be helpful.
  6. Typographical: “Subjets” (§3.1, OptimalB), “Robbins” vs “Robins” inconsistency in a few places, and “hormetic U-shape” may be unfamiliar to some readers.

Circularity Check

1 steps flagged · score 1.0 of 10

Minor self-citations supply prior variance expressions that this paper extends; core optimality, conservativeness, and local-alternative tightness results are derived independently from the model and matrix algebra.

  1. self citation load bearing [Section 2.1 / Equation 5]
    "In Kapelner et al. (2023) we showed that if the design is balanced, then ˆτ is unbiased for τ and we proved that Var[ˆτ]=1/n^{2} ((p_T + p_C)⊤ Σ_W (p_T + p_C) + 2[p_T⊤(1-p_T)+p_C⊤(1-p_C)])."

    The exact variance expression that seeds the lower bound (Eq. 6), the asymptotic-optimality condition (Eq. 7), and the difference term in Theorem 2 is imported from the authors' own prior paper rather than re-derived from scratch in the main text. The supplement later uses this expression, so the citation is load-bearing for the chain, yet the subsequent optimality and tightness arguments remain independent algebraic consequences and do not loop back to redefine the inputs.

full rationale

The paper works entirely inside Neyman's nonparametric model (fixed covariates, independent Bernoulli potential outcomes, balanced designs with E[W]=0). Exact variance (Eq. 5), the covariate-balance optimality condition (Eq. 7), Theorem 1 (blocking under Lipschitz), Theorem 2 (E[V_CMH]-Var = (4/n^{2})ηᵀ Σ_W⊕2 η ≥ 0 by Schur), and Theorem 3 (tightness under local alternatives + bounded λ_max) are obtained by direct expansion, Jensen, Fan-Horn, Gershgorin, and a suboptimal cube-partition argument given in full in the supplement. V_CMH is not defined to equal the target variance; the difference term is non-negative by construction and vanishes asymptotically under the stated conditions. Self-citations (Kapelner et al. 2023 for the variance formula; Azriel et al. 2026 for the CMH identification and power; Azriel et al. 2024 for an auxiliary variance calculation) merely import intermediate expressions that are re-used and extended; they are not uniqueness theorems, ansatzes, or fitted parameters re-labeled as predictions. No step reduces a claimed first-principles result to its own inputs by definition. Score 1 reflects only the presence of non-load-bearing self-citation, consistent with ordinary sequential research.

Assumptions & free parameters 3 free parameters · 6 assumptions · 2 invented entities

The central claims rest on the Neyman nonparametric binary model (fixed covariates, independent Bernoulli potential outcomes), balanced equally-likely designs, standard matrix analysis, and mild regularity (Lipschitz response surfaces; local-alternative second-moment vanishing; bounded design eigenvalues). No new physical entities are postulated. Simulation free parameters affect only empirical illustrations, not the theorems.

free parameters (3)
  • Simulation covariate effect norm ‖β‖ = 3
    Fixed at 3 in §3.1 to magnify design differences; not used in theorems but drives reported power gaps.
  • Simulation treatment coefficient β_T = 0.5 (power); 0 (size)
    Set to 0.5 for power/coverage and 0 for size; chooses effect size for Monte Carlo comparison only.
  • Rerandomization acceptance threshold = 1%
    Keep lowest 1% Mahalanobis-distance BCRD allocations; design hyperparameter for simulations.
assumptions (6)
  • domain assumption Potential outcomes Y_{T,i}~Bernoulli(p_{T,i}), Y_{C,i}~Bernoulli(p_{C,i}) independent given fixed covariates (Neyman nonparametric Setting (b)).
    Stated in §2.1; all variance and estimator results condition on this model.
  • domain assumption Designs are balanced: ∑ W_i = 0 a.s., E[W]=0, n even.
    §1 and §2.1; unbiasedness of τ̂ and the variance formula (5) use this.
  • domain assumption p_{T,i}=h_T(x_i), p_{C,i}=h_C(x_i) with h_T,h_C Lipschitz and ‖x_i‖ bounded (for blocking optimality).
    Theorem 1 hypothesis; without smoothness, within-block variance of v_i need not vanish as B→∞.
  • domain assumption Local alternatives: √n τ → τ_∞ and (1/n)‖η‖²→0; λ_max(Σ_W) bounded (and max row ℓ1 of Σ_W bounded for consistency).
    Conditions (14)–(15) and Theorem 3; required for asymptotic tightness of V_CMH and valid CIs.
  • standard math Schur product theorem: Σ_W ⊙ Σ_W is PSD when Σ_W is PSD.
    Used in Theorem 2 to prove E[V_CMH]≥Var[τ̂].
  • standard math Fan–Horn inequality and Gershgorin circle theorem for eigenvalue/row-sum bounds.
    Used in §2.3.1 and proof of Theorem 3 to control η^T Σ_W^{⊙2} η and Var[n V_CMH].
invented entities (2)
  • V_CMH = (4/n²) Y^T Σ_W Y (general balanced designs)
    purpose: Conservative variance estimator for τ̂ that does not require estimating p_T,p_C and reduces to a known CMH form under blocking.
    Defined in §2.3.1 Eq. (8); conservativeness proved in Theorem 2. Independent evidence is mathematical (expectation identity), not external empirical discovery.
  • V_Robbins-ext (blockwise Robins plus Wald-type Bernoulli variance term)
    purpose: Extend Robins (1988) conservative variance from fixed potential outcomes under BCRD to blocking under Setting (b).
    §2.3.2 Eqs. (12)–(13); asymptotic conservativeness relies on large n_B for consistent block means.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Optimal Designs with Robust Inference for Binary Treatment Effects." pith.science (2026). https://pith.science/paper/HLEGA6XY

@misc{pith2026260705768,
  author       = {Pith},
  title        = {Pith review of: Optimal Designs with Robust Inference for Binary Treatment Effects},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HLEGA6XY}},
  note         = {Machine review of arXiv:2607.05768}
}
read the original abstract

We study randomized experiments with binary outcomes under Neyman's nonparametric model, where covariate measurements are fixed but potential outcomes are random. In this setting we derive the exact variance of the difference-in-means estimator and characterize designs that minimize it. We show that any balanced design satisfying a covariate-balance condition is asymptotically optimal, and we prove that a broad class of blocking designs satisfies this condition under mild smoothness assumptions. Because the variance depends on unknown success probabilities, unbiased variance estimation is impossible. We therefore develop two conservative estimators: a generalization of the Cochran-Mantel-Haenszel (CMH) statistic applicable to any balanced design, and an extension of Robins' variance estimator for blocking designs. We establish conditions under which the CMH-based estimator is asymptotically tight under local alternatives, thereby yielding asymptotically valid confidence intervals. Our theoretical and simulations results show that blocking and other designs that achieve covariate balance and sufficient degree of randomness, perform well when they are equipped with the CMH-based inference. Thus, we provide an experimental design and inference framework for incidence outcomes that is simultaneously variance-optimal, conservative in finite samples and asymptotically tight.

Figures

Figures reproduced from arXiv: 2607.05768 by the authors.

Figure 1
Figure 1. Power and size results for legal inference when [PITH_FULL_IMAGE:figures/full_fig_p015_1.png] view at source ↗
Figure 2
Figure 2. Coverage and confidence interval length results analogous to Figure 1. [PITH_FULL_IMAGE:figures/full_fig_p016_2.png] view at source ↗
Figure 3
Figure 3. Power performance of the CMH estimator over different block sizes [PITH_FULL_IMAGE:figures/full_fig_p017_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Power and size results for legal inference when [PITH_FULL_IMAGE:figures/full_fig_p029_4.png]
Figure 5
Figure 5. Figure 5: Coverage and confidence interval length results for legal inference when [PITH_FULL_IMAGE:figures/full_fig_p030_5.png]
Figure 6
Figure 6. Figure 6: Power and size results for legal inference when [PITH_FULL_IMAGE:figures/full_fig_p031_6.png]
Figure 7
Figure 7. Figure 7: Coverage and confidence interval length results for legal inference when [PITH_FULL_IMAGE:figures/full_fig_p032_7.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

137 extracted references · 137 canonical work pages

  1. [1]

    Statistics in Medicine , volume=

    Implementing unequal randomization in clinical trials with heterogeneous treatment costs , author=. Statistics in Medicine , volume=

    work page

  2. [2]

    Statistics and Its Interface , volume=

    Optimal responses-adaptive designs based on efficiency, ethic, and cost , author=. Statistics and Its Interface , volume=

    work page

  3. [3]

    Statistics in Medicine , volume=

    A comparative study of restricted randomization procedures for multiarm trials with equal or unequal treatment allocation ratios , author=. Statistics in Medicine , volume=

    work page

  4. [4]

    Biometrics , pages=

    Optimum sample sizes for the comparison of the control and treatment , author=. Biometrics , pages=

    work page

  5. [5]

    Journal of the American Statistical Association , volume=

    Implementing optimal allocation in sequential binary response experiments , author=. Journal of the American Statistical Association , volume=

    work page

  6. [6]

    Statistics in Medicine , volume=

    Optimum treatment allocation rules under a variance heterogeneity model , author=. Statistics in Medicine , volume=

    work page

  7. [7]

    Contemporary clinical trials , volume=

    The use of unequal randomisation ratios in clinical trials: a review , author=. Contemporary clinical trials , volume=

    work page

  8. [8]

    Journal of Health Services Research & Policy , volume=

    Unequal randomisation can improve the economic efficiency of clinical trials , author=. Journal of Health Services Research & Policy , volume=. 1997 , publisher=

    work page 1997

Show all 137 references
  1. [9]

    Journal of clinical epidemiology , volume=

    Rationale for unequal randomization in clinical trials is rarely reported: a systematic review , author=. Journal of clinical epidemiology , volume=. 2014 , publisher=

    2014

  2. [10]

    Journal of the American Statistical Association , volume=

    Causal inference using potential outcomes: Design, modeling, decisions , author=. Journal of the American Statistical Association , volume=

  3. [11]

    arXiv preprint arXiv:2103.13051 , year=

    Pair-switching rerandomization , author=. arXiv preprint arXiv:2103.13051 , year=

    arXiv

  4. [12]

    arXiv preprint arXiv:2205.09736 , year=

    Balanced and robust randomized treatment assignments: the Finite Selection Model , author=. arXiv preprint arXiv:2205.09736 , year=

    arXiv

  5. [13]

    arXiv preprint arXiv:1911.03071 , year=

    Balancing covariates in randomized experiments with the Gram--Schmidt Walk design , author=. arXiv preprint arXiv:1911.03071 , year=

    1911 arXiv

  6. [14]

    2013 , edition=

    Applied logistic regression , author=. 2013 , edition=

    2013

  7. [15]

    2002 , edition=

    Categorical data analysis , author=. 2002 , edition=

    2002

  8. [16]

    Canadian Journal of Statistics , year=

    Better experimental design by hybridizing binary matching with imbalance optimization , author=. Canadian Journal of Statistics , year=

  9. [17]

    Journal of Statistical Planning and Inference , volume=

    Optimal rerandomization designs via a criterion that provides insurance against failed experiments , author=. Journal of Statistical Planning and Inference , volume=

  10. [18]

    arXiv preprint arXiv:2209.00490 , year=

    The Role of Pairwise Matching in Experimental Design for an Incidence Outcome , author=. arXiv preprint arXiv:2209.00490 , year=

    arXiv

  11. [19]

    Australian & New Zealand Journal of Statistics , volume=

    The role of pairwise matching in experimental design for an incidence outcome , author=. Australian & New Zealand Journal of Statistics , volume=

  12. [20]

    Foster, HE and Hanno, PM and Nickel, JC and Payne, CK and Mayer, RD and Burks, DA and Yang, CC and Chai, TC and Kreder, KJ and Peters, KM and Lukacz, ES and FitzGerald, MP and Cen, L and Landis, JR and Propert, KJ and Yang, W and Kusek, JW and Nyberg, LM , journal =

  13. [21]

    Machine Learning , volume=

    Active learning for logistic regression: an evaluation , author=. Machine Learning , volume=

  14. [22]

    Journal of the American Statistical Association , volume=

    Experimental design for binary data , author=. Journal of the American Statistical Association , volume=

  15. [23]

    Journal of the Royal Statistical Society, Series B , volume=

    On the experimental attainment of optimum conditions , author=. Journal of the Royal Statistical Society, Series B , volume=

  16. [24]

    Chemometrics and Intelligent Laboratory Systems , volume=

    Comparing D-optimal designs with common mixture experimental designs for logistic regression , author=. Chemometrics and Intelligent Laboratory Systems , volume=

  17. [25]

    International Journal of Experimental Design and Process Optimisation , volume=

    Choice of second-order response surface designs for logistic and Poisson regression models , author=. International Journal of Experimental Design and Process Optimisation , volume=

  18. [26]

    Quality and Reliability Engineering International , volume=

    Separation in D-optimal experimental designs for the logistic regression model , author=. Quality and Reliability Engineering International , volume=

  19. [27]

    Journal of Statistical Planning and Inference , volume=

    Some alphabetic optimal designs for the logistic regression model , author=. Journal of Statistical Planning and Inference , volume=

  20. [28]

    Journal of the American Statistical Association , volume=

    Design of experiment for bioassay , author=. Journal of the American Statistical Association , volume=

  21. [29]

    Bmj , volume=

    Montelukast and fluticasone compared with salmeterol and fluticasone in protecting against asthma exacerbation in adults: one year, double blind, randomised, comparative trial , author=. Bmj , volume=. 2003 , publisher=

    2003

  22. [30]

    The Lancet , volume=

    Outcomes in hypertensive patients at high cardiovascular risk treated with regimens based on valsartan or amlodipine: the VALUE randomised trial , author=. The Lancet , volume=. 2004 , publisher=

    2004

  23. [31]

    Journal of clinical oncology , volume=

    Randomized trial comparing conventional-dose with high-dose conformal radiation therapy in early-stage adenocarcinoma of the prostate: long-term results from proton radiation oncology group/american college of radiology 95-09 , author=. Journal of clinical oncology , volume=. ...

    2010

  24. [32]

    IEEE Transactions on information theory , volume=

    Optimal binary sequences for spread spectrum multiplexing (Corresp) , author=. IEEE Transactions on information theory , volume=

  25. [33]

    Linear algebra and its applications , volume=

    Geometry of the Welch bounds , author=. Linear algebra and its applications , volume=

  26. [34]

    1966 , institution=

    Weight distribution formula for some class of cyclic codes , author=. 1966 , institution=

    1966

  27. [35]

    2006 , edition=

    Testing statistical hypotheses , author=. 2006 , edition=

    2006

  28. [36]

    arXiv preprint arXiv:1912.02633 , year=

    Another look at the Lady Tasting Tea and permutation-based randomization tests , author=. arXiv preprint arXiv:1912.02633 , year=

    1912 arXiv

  29. [37]

    Statistical applications in genetics and molecular biology , volume=

    Permutation P-values should never be zero: calculating exact P-values when permutations are randomly drawn , author=. Statistical applications in genetics and molecular biology , volume=

  30. [38]

    Test , volume=

    Exact testing with random permutations , author=. Test , volume=

  31. [39]

    The Annals of Mathematical Statistics , pages=

    Modified randomization tests for nonparametric hypotheses , author=. The Annals of Mathematical Statistics , pages=

  32. [40]

    Journal of the American Statistical Association , volume=

    General forms of finite population central limit theorems with applications to causal inference , author=. Journal of the American Statistical Association , volume=. 2017 , publisher=

    2017

  33. [41]

    Some extensions of the

    H. Some extensions of the. The Annals of Mathematical Statistics , pages=

  34. [42]

    Advances in Neural Information Processing Systems , pages=

    Removing Hidden Confounding by Experimental Grounding , author=. Advances in Neural Information Processing Systems , pages=

  35. [43]

    Machine learning , volume=

    Random forests , author=. Machine learning , volume=

  36. [44]

    , author=

    The State of Tennessee's Student/Teacher Achievement Ratio (STAR) Project: Technical Report (1985-1990). , author=. 1990 , institution=

    1985

  37. [45]

    Discrete mathematics , volume=

    A lower bound on the probability of a union , author=. Discrete mathematics , volume=

  38. [46]

    The Annals of Statistics , pages=

    Rerandomization to improve covariate balance in experiments , author=. The Annals of Statistics , pages=

  39. [47]

    1982 , note =

    The Differencing Method of Set Partitioning , author=. 1982 , note =

    1982

  40. [48]

    The differencing algorithm

    Yakir, B , journal=. The differencing algorithm. 1996 , publisher=

    1996

  41. [49]

    Operations Research , volume=

    The power of optimization over randomization in designing experiments involving small samples , author=. Operations Research , volume=

  42. [50]

    International Statistical Review , volume=

    Econometric causality , author=. International Statistical Review , volume=

  43. [51]

    Fisher, R A , title =

  44. [52]

    Biostatistics , volume=

    Optimal multivariate matching before randomization , author=. Biostatistics , volume=

  45. [53]

    Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume=

    Optimal a priori balance in the design of controlled experiments , author=. Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume=. 2018 , publisher=

    2018

  46. [54]

    Biometrika , pages=

    Comparison between balanced and random arrangements of field plots , author=. Biometrika , pages=. 1938 , publisher=

    1938

  47. [55]

    Pocock, S J and Simon, R , journal =

  48. [56]

    Biometrics , volume=

    Matching on-the-fly: Sequential allocation with higher power and efficiency , author=. Biometrics , volume=

  49. [57]

    Biometrics , pages=

    A matching procedure for sequential experiments that iteratively learns which covariates improve power , author=. Biometrics , pages=

  50. [58]

    Freedman, D A , journal =

  51. [59]

    Journal of Computational and Graphical Statistics , volume =

    Hansen, B B and Klopfer, S O , title =. Journal of Computational and Graphical Statistics , volume =

  52. [60]

    Statistics in medicine , volume=

    Metrics for covariate balance in cohort studies of causal effects , author=. Statistics in medicine , volume=

  53. [61]

    arXiv preprint arXiv:1604.00698 , year=

    Asymptotic Theory of Rerandomization in Treatment-Control Experiments , author=. arXiv preprint arXiv:1604.00698 , year=

    arXiv

  54. [62]

    Hansen, B E , year=2018, title =

    2018

  55. [63]

    Biometrika , volume=

    The covariate-adaptive biased coin design for balancing clinical trials in the presence of prognostic factors , author=. Biometrika , volume=

  56. [64]

    2016 , edition=

    Randomization in clinical trials: theory and practice , author=. 2016 , edition=

    2016

  57. [65]

    2006 , publisher=

    Optimal design of experiments , author=. 2006 , publisher=

    2006

  58. [66]

    2009 , publisher=

    Introduction to algorithms , author=. 2009 , publisher=

    2009

  59. [67]

    Management Science , year=

    Near-Optimal AB Testing , author=. Management Science , year=

  60. [68]

    Controlled clinical trials , volume=

    Properties of simple randomization in clinical trials , author=. Controlled clinical trials , volume=. 1988 , publisher=

    1988

  61. [69]

    1973 , publisher=

    Linear statistical inference and its applications , author=. 1973 , publisher=

    1973

  62. [70]

    Fisher, R A , publisher =

  63. [71]

    American journal of mental deficiency , volume=

    Principles of research , author=. American journal of mental deficiency , volume=

  64. [72]

    Biometrika , volume=

    Nearly random designs with greatly improved balance , author=. Biometrika , volume=

  65. [73]

    Econometrica: Journal of the Econometric Society , pages=

    An automatic method of solving discrete programming problems , author=. Econometrica: Journal of the Econometric Society , pages=. 1960 , publisher=

    1960

  66. [74]

    , author=

    Empirical model-building and response surfaces. , author=. 1987 , publisher=

    1987

  67. [75]

    1998 , publisher=

    Theory of point estimation , author=. 1998 , publisher=

    1998

  68. [76]

    2012 , publisher=

    Philosophical theories of probability , author=. 2012 , publisher=

    2012

  69. [77]

    The Annals of Mathematical Statistics , pages=

    Design for the control of selection bias , author=. The Annals of Mathematical Statistics , pages=

  70. [78]

    Journal of Nonparametric Statistics , volume=

    Bias properties and nonparametric inference for truncated binomial randomization , author=. Journal of Nonparametric Statistics , volume=

  71. [79]

    Technical University of Denmark , volume=

    The matrix cookbook , author=. Technical University of Denmark , volume=

  72. [80]

    Journal of the American Statistical Association , volume=

    Using multivariate matched sampling and regression adjustment to control bias in observational studies , author=. Journal of the American Statistical Association , volume=

  73. [81]

    The American Statistician , volume=

    Optimal nonbipartite matching and its statistical applications , author=. The American Statistician , volume=

  74. [82]

    Statistical science , volume=

    Matching methods for causal inference: A review and a look forward , author=. Statistical science , volume=

  75. [83]

    Journal of Theoretical Probability , volume=

    Asymptotic distribution of quadratic forms and applications , author=. Journal of Theoretical Probability , volume=

  76. [84]

    Biometrics , pages=

    Confidence intervals from randomization tests , author=. Biometrics , pages=

  77. [85]

    The Annals of Mathematical Statistics , volume=

    The distribution of a quadratic form of normal random variables , author=. The Annals of Mathematical Statistics , volume=

  78. [86]

    Journal of the American Statistical Association , volume=

    Interference between units in randomized experiments , author=. Journal of the American Statistical Association , volume=

  79. [87]

    Journal of multivariate analysis , volume=

    Some representations of the multivariate Bernoulli and binomial distributions , author=. Journal of multivariate analysis , volume=

  80. [88]

    Biometrika , volume=

    Robustness of normal tolerance intervals , author=. Biometrika , volume=

  81. [89]

    2015 , publisher=

    Causal inference in statistics, social, and biomedical sciences , author=. 2015 , publisher=

    2015

  82. [90]

    The Annals of Statistics , pages=

    On the robustness and efficiency of some randomized designs , author=. The Annals of Statistics , pages=

  83. [91]

    Statistics in Medicine , volume=

    Seven myths of randomisation in clinical trials , author=. Statistics in Medicine , volume=

  84. [92]

    Statistical Science , volume=

    Covariance adjustment in randomized experiments and observational studies , author=. Statistical Science , volume=

  85. [93]

    Proceedings of the National Academy of Sciences , volume=

    Asymptotic theory of rerandomization in treatment--control experiments , author=. Proceedings of the National Academy of Sciences , volume=

  86. [94]

    Biometrika , volume=

    On the standard deviations of adjusted and interpolated values of an observed polynomial function and its constants and the guidance they give towards a proper choice of the distribution of observations , author=. Biometrika , volume=

  87. [95]

    The Annals of Mathematical Statistics , pages=

    Optimum designs in regression problems , author=. The Annals of Mathematical Statistics , pages=

  88. [96]

    Technometrics , volume=

    Experimental design: review and comment , author=. Technometrics , volume=

  89. [97]

    The Annals of Applied Statistics , volume=

    Agnostic notes on regression adjustments to experimental data: Reexamining Freedman's critique , author=. The Annals of Applied Statistics , volume=

  90. [98]

    When second best is good enough:

    Berk, Richard and Barnes, Geoffrey and Ahlman, Lindsay and Kurtz, Ellen , journal=. When second best is good enough:

  91. [99]

    Experimental randomization:

    Harville, David A , journal=. Experimental randomization:

  92. [100]

    The American Statistician , volume=

    Harmonizing Optimized Designs with Classic Randomization in Experiments , author=. The American Statistician , volume=. 2021 , publisher=

    2021

  93. [101]

    R package version 1.5.1

    nbpMatching: Functions for Optimal Non-Bipartite Matching. R package version 1.5.1. , author=. 2016 , publisher=

    2016

  94. [102]

    Statistics and Computing , volume=

    A comparison of efficient approximations for a weighted sum of chi-squared random variables , author=. Statistics and Computing , volume=

  95. [103]

    Australian Journal of Statistics , volume=

    An approximation to the distribution of quadratic forms in normal random variables , author=. Australian Journal of Statistics , volume=

  96. [104]

    1990 , publisher=

    Generalized Additive Models , author=. 1990 , publisher=

    1990

  97. [105]

    , author=

    Justifying the use and increasing the power of at test for a randomized experiment with a convenience sample. , author=. Psychological methods , volume=

  98. [106]

    arXiv preprint arXiv:2008.05980 , year=

    Improving the Power of the Randomization Test , author=. arXiv preprint arXiv:2008.05980 , year=

    2008 arXiv

  99. [107]

    , author=

    Estimating causal effects of treatments in randomized and nonrandomized studies. , author=. J of Educ Psychol , volume=

  100. [108]

    Journal of Causal Inference , doi =

    Properties of restricted randomization with implications for experimental design , author =. Journal of Causal Inference , doi =. 2022 , lastchecked =

    2022

  101. [109]

    Efron, B , journal =

  102. [110]

    Biometrics , volume=

    Some methods for strengthening the common ^2 tests , author=. Biometrics , volume=. 1954 , publisher=

    1954

  103. [111]

    Journal of the national cancer institute , volume=

    Statistical aspects of the analysis of data from retrospective studies of disease , author=. Journal of the national cancer institute , volume=. 1959 , publisher=

    1959

  104. [112]

    2024 , journal=

    The Optimality of Blocking Designs in Equally and Unequally Allocated Randomized Experiments with General Response , author=. 2024 , journal=

    2024

  105. [113]

    Stat , volume=

    The Pairwise Matching Design Is Optimal Under Extreme Noise and Extreme Assignments , author=. Stat , volume=. 2025 , publisher=

    2025

  106. [114]

    2010 , publisher=

    Normal approximation by Stein’s method , author=. 2010 , publisher=

    2010

  107. [115]

    Biometrika , volume=

    Odds ratio estimators when the data are sparse , author=. Biometrika , volume=. 1981 , publisher=

    1981

  108. [116]

    , journal =

    Gart, John J. , journal =. On the Combination of Relative Risks , urldate =

  109. [117]

    The Analysis of Several 2 × 2 Contingency Tables , urldate =

    Zelen, Marvin , journal =. The Analysis of Several 2 × 2 Contingency Tables , urldate =

  110. [118]

    Statistics in medicine , volume=

    Confidence intervals for causal parameters , author=. Statistics in medicine , volume=. 1988 , publisher=

    1988

  111. [119]

    Statistical Science , pages=

    Randomization does not justify logistic regression , author=. Statistical Science , pages=. 2008 , publisher=

    2008

  112. [120]

    , title =

    Sekhon, Jasjeet S. , title =. The Oxford Handbook of Political Methodology , editor =. 2008 , pages =. doi:10.1093/oxfordhb/9780199286546.003.0011 , isbn =

    2008 doi

  113. [121]

    Statistics in medicine , volume=

    Randomization inference for treatment effects on a binary outcome , author=. Statistics in medicine , volume=. 2015 , publisher=

    2015

  114. [122]

    Proceedings of the 24th ACM Conference on Economics and Computation , pages=

    Fast computation of exact confidence intervals for randomized experiments with binary outcomes , author=. Proceedings of the 24th ACM Conference on Economics and Computation , pages=

  115. [123]

    arXiv preprint arXiv:2508.03834 , year=

    Exact and Conservative Inference for the Average Treatment Effect in Stratified Experiments with Binary Outcomes , author=. arXiv preprint arXiv:2508.03834 , year=

    arXiv

  116. [124]

    arXiv preprint arXiv:2405.03910 , year=

    A primer on the analysis of randomized experiments and a survey of some recent advances , author=. arXiv preprint arXiv:2405.03910 , year=

    arXiv

  117. [125]

    and Fr \"a nti, Pasi

    Malinen, Mikko I. and Fr \"a nti, Pasi. Balanced K-Means for Clustering. Structural, Syntactic, and Statistical Pattern Recognition. 2014

    2014

  118. [126]

    Canadian Journal of Mathematics , year =

    Jack Edmonds , title =. Canadian Journal of Mathematics , year =

  119. [127]

    Essay on principles

    On the application of probability theory to agricultural experiments. Essay on principles. Section 9 , author=. Statistical Science , pages=. 1990 , publisher=

    1990

  120. [128]

    Azriel et al

    Block designs that provide optimal power in the Cochran--Mantel--Haenszel test: D. Azriel et al. , author=. Statistical Papers , volume=. 2026 , publisher=

    2026

  121. [129]

    and Johnson, Charles R

    Horn, Roger A. and Johnson, Charles R. , year=. Topics in Matrix Analysis , publisher=

  122. [130]

    Spielman and Peng Zhang , title =

    Christopher Harshaw and Fredrik Sävje and Daniel A. Spielman and Peng Zhang , title =. Journal of the American Statistical Association , volume =. 2024 , publisher =

    2024

  123. [131]

    2000 , publisher=

    Asymptotic statistics , author=. 2000 , publisher=

    2000

  124. [132]

    2004 , publisher=

    Gersgorin and His Circles , author=. 2004 , publisher=

    2004

  125. [133]

    2013 , publisher =

    Matrix Analysis , author =. 2013 , publisher =

    2013

  126. [134]

    Annals of Operations Research , year =

    Ulrich Derigs , title =. Annals of Operations Research , year =

  127. [135]

    Cole Beck and Bo Lu and Robert Greevy , year =

  128. [136]

    Klau , title =

    Martin Papenberg and Gunnar W. Klau , title =. Psychological Methods , year =

  129. [137]

    Martin Papenberg , year =

Pith tools

Reviewed July 11, 2026 · model on record in the stance chip above.