Pith. sign in

REVIEW 2 major objections 4 minor 25 references

Causal inference for group-contaminated structured outcomes: observable quotients, lossless reduction and exact randomization inference

T0 review · 2 major / 4 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read Quotienting group-contaminated outcomes is statistically lossless exactly when the conditional law of the raw observation given the quotient is parameter-free.

desk verdict A sound theoretical core on observability vs. sufficiency for group-contaminated outcomes, with an empirical demonstration whose confirmatory p-value rests on an unverifiable allocation assumption and unadjusted pair selection. read the letter →

arxiv 2608.11954 v1 pith:ATAIDXRK submitted 2026-08-12 stat.ME cs.AI

Usef Faghihi , Amir Saki This is my paper · ORCID
classification stat.MEcs.AI MSC 62B1562D2062B05
keywords causalinferencegroupactionmaximalinvariantBlackwellsufficiencymicroscopyimagequotientspacerandomizationteststructuredoutcome
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

This paper studies experiments in which structured outcomes, such as microscopy images, are recorded after an unknown unit-specific transformation $\Gamma$ that may depend on treatment, covariates, and the outcome itself, so the observed object is $X = \Gamma \cdot Y(A)$ rather than the intrinsic potential outcome $Y(A)$. It separates three questions: a target is uniformly recoverable exactly when it is constant on group orbits; the interventional law of the quotient is identified by the ordinary covariate-adjustment formula; and quotient reduction is statistically lossless exactly when the conditional law of the raw observation given treatment, covariates, and the quotient has a parameter-free version, making the quotient experiment Blackwell-equivalent to the full experiment. This third boundary is the paper's central discovery because it distinguishes observability (invariance) from statistical sufficiency. The framework is carried through to an exact paired-swap randomization test for lattice images, which under the sharp null rejects in $0.052$ of simulation replicates at the nominal level, reaches power $0.992$ at unit effect strength, and reports $p = 0.0078$ for the primary RxRx1 HUVEC contrast.

What carries the argument

The machinery is the comparison of two statistical experiments, the full experiment $(C, A, X)$ and the quotient experiment $(C, A, M(X))$, in the Blackwell sense. The object carrying the argument is the quotient-faithful kernel: a Markov kernel $K$ from the quotient sample space to the full sample space that is supported on the fibre of $T$ over $s$ and reproduces every full-experiment law through $P^{\mathrm{quot}}_\theta K = P^{\mathrm{full}}_\theta$. Theorem 3.2 shows that such a parameter-free kernel exists precisely when the conditional law of the raw observation given the quotient is parameter-free, with the proof running through the disintegration identity for a common regular conditional distribution. Supporting constructions include the maximal invariant $M$, whose equality classes are exactly the group orbits, together with a Borel factorization theorem for invariant targets; the orbit kernel formed by integrating a Borel section against normalized Haar measure, which yields the conditional-Haar equivalence corollary; and, at the application level, the lattice canonicalizer that removes the translation normalizing the support bounding box and then selects the lexicographically smallest of the four quarter-turn variants, producing a Borel maximal invariant for $\mathbb{Z}^2 \rtimes C_4$ without placing a probability law on the noncompact translation group.

What would settle it

Re-run the paired-swap analysis on the eight RxRx1 confirmation blocks under a documented non-uniform within-plate assignment (for example, $P(Z_b = 1) = 0.7$ in some blocks) and check whether the sharp-null rejection rate exceeds the nominal level; if it does, the conditional-uniformity assumption behind Theorem 7.1 and the reported $p = 0.0078$ has been violated. Independently, the equivalence in Theorem 3.2 can be attacked directly: on a small finite group example, exhaustively search for a parameter-free quotient-faithful kernel $K$ with $P^{\mathrm{quot}}_\theta K = P^{\mathrm{full}}_\theta$ while $\mathrm{Law}_\theta(X \mid C, A, M(X))$ still depends on $\theta$; the theorem asserts that no such family exists.

Watch

Extended reading notes

Core claim

The paper's central claim is Theorem 3.2: under the unrestricted contamination model $X = \Gamma \cdot Y(A)$, the quotient experiment based on a maximal invariant $M$ is sufficient for, and Blackwell-equivalent to, the full experiment if and only if there exists a parameter-free quotient-faithful Markov kernel $K$ satisfying $P^{\mathrm{quot}}_\theta K = P^{\mathrm{full}}_\theta$ for every $\theta$, equivalently if and only if the conditional law $\mathrm{Law}_\theta(X \mid C, A, M(X))$ admits a version independent of $\theta$. Conditional Haar contamination on a compact group is derived as the special case that guarantees this condition, but it is not imposed in the main model. The paper also proves that invariance is exactly what makes a target observable (Theorem 2.1), that every measurable invariant target factors through a Borel maximal invariant (Theorem 2.4), that interventional quotient laws are identified by adjustment (Theorem 2.5), that componentwise canonicalization is maximal for independent sitewise actions but erases relative configuration under a shared diagonal action (Proposition 4.1), and that for finite-support multichannel lattice images a lexicographic canonicalizer is a Borel maximal invariant under integer translations and quarter turns (Theorem 6.1). Together with a characteristic Gaussian kernel on the quotient and exhaustive enumeration of the paired assignment distribution, these results ground an exact finite-sample randomization test (Theorem 7.1).

Load-bearing premise

The real-data conclusion rests on the assumption that the original within-plate allocation of the two siRNAs per block was conditionally uniform: the metadata audit can establish that the paired siRNAs belonged to the same randomization set, but, as the paper states in Section 9.2, it cannot itself prove that the allocation algorithm was uniform, and if it was not, the reported paired-swap $p$-values are not valid randomization $p$-values for the actual experiment.

Editorial extensions

If this is right

  • If the quotient experiment is Blackwell-equivalent to the full experiment, every bounded decision problem has the same attainable risk set under both, so analyses of invariant targets can be conducted on the quotient without sacrificing any experiment-relevant information.
  • The paired-swap test is exactly valid under the Fisher sharp null: rejection when the enumerated $p$-value is at most $\alpha$ has conditional probability at most $\alpha$, and in the simulations the test rejected in $0.052$ of replicates at the nominal $0.05$ level with power $0.992$ at unit effect strength.
  • Sitewise canonicalization is the correct quotient when sites undergo independent motions, but under one shared diagonal motion it erases relative configuration such as inter-site displacement; a diagonal quotient is required to preserve that information.
  • Under explicit Lipschitz regularity, approximate contamination propagates linearly: orbit-metric error $\varepsilon_a$ implies quotient-law Wasserstein error at most $L_M \varepsilon_a$ and population-MMD perturbation at most $L_k L_M (\varepsilon_1 + \varepsilon_0)$; the paper leaves the lexicographic canonicalizer's Lipschitz status open, so its perturbation table remains an empirical sensitivit
  • The primary RxRx1 HUVEC contrast is significant at $p = 0.0078$ with exact invariance holding across all imposed acquisition seeds, while no secondary contrast survives Holm adjustment; the eight-block design gives the enumeration a coarse resolution of $2^{-8}$.

Reading between the lines

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

  • The observability-sufficiency gap yields a diagnostic for any proposed invariant summary: test whether the conditional law of the raw outcome given the summary varies with the treatment parameter; if it does, the summary is not statistically lossless even when it is maximal.
  • The product-versus-diagonal split points to an intermediate model the paper does not pursue: a partially shared action, in which some sites share one motion while others do not, would require a quotient retaining relative pose only within the sharing structure, and the testable question is whether such relative configuration predicts the outcome.
  • The lattice canonicalizer embodies a transferable recipe, an explicit cross-section for the noncompact part of the group followed by finite lexicographic minimization over the compact residual, which should extend to other semidirect products with countable orbit structure, such as lattice actions with reflections or scaling.
  • Because the exact test enumerates only $256$ assignments, its $p$-values are coarse; combining the quotient kernel with more blocks or a continuous statistic would sharpen resolution while preserving the design-based validity guarantee.
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

2 major / 4 minor

Summary. The paper develops a theory of causal inference for structured outcomes observed as X = Γ·Y(A), where Γ is an unobserved group element whose distribution may depend arbitrarily on treatment, covariates and intrinsic outcomes. It proves an observability theorem (measurable targets are uniformly recoverable iff they are G-invariant; a maximal invariant factorizes all invariant targets), a losslessness theorem (quotient reduction is Blackwell-equivalent to the full experiment exactly when a parameter-free, quotient-faithful reconstruction kernel exists, with conditional-Haar contamination as a special case), a product-versus-diagonal action dichotomy, and an approximate-contamination Wasserstein/MMD stability bound. The paper then constructs a maximal invariant for multichannel lattice images under integer translations and quarter turns, defines a characteristic Gaussian kernel on the quotient, and combines these with a complete paired-swap randomization test under the sharp null. The methods are applied to simulations and to an RxRx1 HUVEC confirmation experiment, with a reported primary paired-swap p-value of 0.0078. The text is unusually explicit about its assumptions and limitations.

Significance. If the main theorems hold (and they appear correct), the paper makes a valuable conceptual contribution by separating observability from statistical sufficiency for group-contaminated outcomes: Theorem 3.2 gives a precise condition under which quotienting is decision-theoretically lossless, and the lattice maximal invariant of Theorem 6.1 provides an explicit noncompact construction with a characteristic kernel. The paired-swap randomization test in Theorem 7.1 is clean, and the simulation evidence (0.052 rejection at the null, 0.992 power at effect strength 1.0) is supportive. The authors are also to be credited for flagging the unverifiable assumptions behind the RxRx1 p-value and for refusing to present Theorem 5.1 as certifying the approximate-action stress test. The main uncertainty is about the empirical demonstration, where the headline p-value depends on assumptions that metadata cannot verify.

major comments (2)
  1. [Section 7 and Abstract] Theorem 7.1's exactness requires that, conditional on the unordered eligible wells, their complete potential outcomes and pretreatment design information, Z is uniform on {0,1}^B. The manuscript states (Section 7, repeated in Section 10) that the metadata audit can establish that paired siRNAs occur within the same plate randomization set but cannot prove the original allocation algorithm was uniform. Since the abstract's headline p-value of 0.0078 is a randomization p-value for the real RxRx1 experiment only under this assumption, the confirmatory language in the abstract and Section 9.2 overstates the evidence. Please either carry this caveat into the abstract and the results section, explicitly labeling the RxRx1 finding as conditional on an unverifiable uniform-within-plate allocation, or add a sensitivity analysis over plausible non-uniform allocation mechanisms (for example, plate-level or batch-level randomization) to quantify how the p-value would change.
  2. [Section 9.1] The selection of the ten confirmation pairs uses public 128-dimensional embeddings from discovery rows only, but the paper concedes that 'the row split does not prove that the public embedding generator was trained independently of every confirmation image.' If the embedding model was trained on confirmation images, the selected pairs are not outcome-independent, and the 'prespecified' characterization of the primary and secondary contrasts is weakened. Please provide evidence of the embedding generator's training provenance, or run a sensitivity analysis with alternative selection rules (for example, random eligible pairs or embeddings from a model trained exclusively on discovery images), and otherwise describe the RxRx1 contrasts as exploratory rather than confirmatory.
minor comments (4)
  1. [Section 7, Eq. (7.2)] Q_1(z) and Q_0(z) are written with set braces, but the MMD statistic uses them as indexed lists; use multisets or tuples to avoid ambiguity when quotient outcomes coincide.
  2. [Section 6.1] The key vector in Theorem 6.1 is described as containing bytes in 'site–channel–row–column order'; for a single image there is no site index, so the description should be 'channel–row–column order'.
  3. [Table 5] The header arrangement makes it unclear which rows correspond to which method; add an explicit Method column and state that |Δp| is the change in the paired-swap p-value, not a p-value itself.
  4. [Abstract] The phrase 'retain the original simulations and RxRx1 HUVEC study' is vague; specify that these analyses are preserved from the prior workflow and that the RxRx1 results are conditional on the assumptions listed in Section 10.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the theoretical results are proven from definitions and standard references, and the one fragile empirical assumption is explicitly flagged as unverifiable rather than smuggled in.

full rationale

The paper's theoretical core is self-contained and does not reduce to its own inputs. Theorem 2.1 is a direct characterization of orbit-invariant recoverability: the proof exhibits the decoder for invariant targets and uses the identity group element to force invariance, so it is a definitional equivalence rather than a circular derivation. Theorem 2.4 invokes standard descriptive set theory and the measurable factorization lemma; Theorem 3.2 proves the equivalence between quotient sufficiency and existence of a parameter-free conditional law by disintegration and Blackwell-dominance arguments, and its two conditions are shown equivalent by proof, not assumed identical. Corollary 3.4 uses the standard right-invariance of normalized Haar measure. The lattice maximal invariant in Theorem 6.1 is constructed explicitly and verified against the semidirect-product action, with no appeal to the conclusion. The randomization test in Theorem 7.1 is the usual finite-sample permutation argument, and the Gaussian bandwidth is fixed on Q_pool, which is invariant under paired label swaps, so it does not enter the randomization distribution and is not a fitted parameter masquerading as a prediction. The main load-bearing concern in the empirical demonstration is the assumption of conditional uniform within-plate allocation, but the paper explicitly states that the metadata audit 'cannot itself prove that the original allocation algorithm was uniform' (Section 7, restated in Section 10), so this is an acknowledged correctness risk, not a hidden circular step. There are no load-bearing self-citations: the reference list consists of external classical and methodological sources, and the phrase 'original outcome-level framework' is mentioned without being cited as evidence for any result. The approximate-contamination theorem is also honestly qualified: the paper states that Theorem 5.1 applies only when its Lipschitz assumptions are verified, and the stress test is presented as empirical sensitivity analysis rather than as certified by the theorem. Overall, the derivation chain is independent of its conclusions, and no circularity is present.

Assumptions & free parameters 4 free parameters · 7 assumptions · 0 invented entities

The paper introduces no new physical or mathematical entities. The free parameters are tuning and selection choices: the kernel bandwidth, the crop size, the secondary Euler threshold count, and the unspecified split-half scoring. The axioms are mostly standard descriptive set theory plus domain assumptions about the observation model, the causal identification conditions, and the two empirical assumptions about the RxRx1 assignment mechanism and embedding independence that the paper itself flags as unverifiable.

free parameters (4)
  • Gaussian kernel bandwidth sigma^2 = median of pooled pairwise squared distances divided by 2 (Section 7)
    Set from the data before examining condition labels. It is label-invariant, so exactness of the randomization test is preserved, but the statistic and power depend on this choice.
  • Crop size L = 28 in simulation, 64 in RxRx1
    Hand-chosen image lattice bound that defines the space X_q,L and the canonicalizer. Not fitted to target, but a modeling choice.
  • Euler signature threshold count = 25 shape-adaptive thresholds in 8 directions
    Hand-chosen for the secondary endpoint; not used in the primary quotient-pixel test.
  • Split-half score weights = unspecified combination of within-half signal-to-noise and agreement of contrast direction
    Used to select the primary RyRx1 pair. The exact formula is not given, and the selection is not adjusted for in the reported p-value.
assumptions (7)
  • domain assumption Observation model X = Γ·Y(A) with Γ unrestricted and jointly Borel group action
    Central model (Section 2.1). The entire analysis assumes the only corruption is a single group element acting on the intrinsic outcome.
  • standard math Y, C, S, G are standard Borel spaces
    Needed for regular conditional probabilities, measurable factorization, and Lusin separation arguments (Sections 2, 3, Appendix A).
  • domain assumption Existence of a Borel maximal invariant and, for Corollary 3.4, a Borel section s: S0 → Y
    Required for the Haar Blackwell-equivalence corollary. The paper assumes S0 = M(Y) is Borel and a Borel section exists, which is not automatic for general non-compact group actions.
  • domain assumption Conditional exchangeability Q(a) ⊥⊥ A | C, positivity, and quotient consistency (2.1)
    Required for identification of the interventional quotient law (Theorem 2.5). Standard causal assumptions applied to quotient outcomes.
  • domain assumption For the exact test: Z is uniform on {0,1}^B conditional on the design, no cross-well interference, sharp null
    Theorem 7.1. The randomization validity is conditional on these design assumptions, which the metadata audit can only partially verify.
  • domain assumption The original RxRx1 within-plate allocation is uniform
    Section 9.2 explicitly states the metadata audit cannot prove the original allocation was uniform. The empirical p-values depend on this.
  • domain assumption The public RxRx1 embedding generator was trained independently of every confirmation image
    Section 9.1: the split-half selection uses discovery rows only, but complete algorithmic independence is flagged as an assumption, not proven.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Causal inference for group-contaminated structured outcomes: observable quotients, lossless reduction and exact randomization inference." pith.science (2026). https://pith.science/paper/ATAIDXRK

@misc{pith2026260811954,
  author       = {Pith},
  title        = {Pith review of: Causal inference for group-contaminated structured outcomes: observable quotients, lossless reduction and exact randomization inference},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ATAIDXRK}},
  note         = {Machine review of arXiv:2608.11954}
}
read the original abstract

Structured potential outcomes such as microscopy images may be recorded after an unknown, unit-specific transformation. If that transformation can depend on treatment, covariates or the intrinsic outcome, raw-coordinate analyses may mix biological effects with acquisition geometry. We study the unrestricted observation model X = {\Gamma} . Y(A) and characterize its observable information: a target is uniformly recoverable exactly when it is constant on group orbits, while a Borel maximal invariant retains every measurable invariant target. We then distinguish observability from statistical losslessness. A quotient-faithful reconstruction theorem shows that quotient reduction is sufficient for the full transformed experiment exactly when the conditional law of the raw observation given treatment, covariates and the quotient has a parameter-free version. Conditional Haar contamination on a compact group yields Blackwell equivalence as a special case; it is not imposed in the main model. We also separate independent site-specific product actions from shared diagonal actions and show why componentwise canonicalization can discard relative cross-site information. Under explicit metric and kernel regularity, an approximate-contamination theorem bounds quotient-law Wasserstein error and the induced perturbation of population maximum mean discrepancy. For finite-support multichannel lattice images, we construct a maximal invariant under integer translations and quarter turns, combine its characteristic Gaussian kernel with a complete paired-swap test, and retain the original simulations and RxRx1 HUVEC study. Under the sharp null, the quotient test rejected in 0.052 of simulation replicates; at unit effect strength its power was 0.992. The primary RxRx1 contrast had an enumerated paired-swap p-value of 0.0078.

Figures

Figures reproduced from arXiv: 2608.11954 by the authors.

Figure 1
Figure 1. Simulation rejection fractions. Finite-sample causal validity under informative acquisition per￾tains to the quotient and post-canonical Euler endpoints; the other methods are diagnostics. The exact quotient begins near the nominal 0.05 level and rises to 0.992, raw pixels remain at 1.0, and moment regis￾tration and the Euler endpoint begin near 0.05 and approach 1.0 as effect strength increases. 9 RxRx1 confirmatio… view at source ↗
Figure 1
Figure 1. Simulation rejection fractions across intrinsic effect strengths; causal validity under [PITH_FULL_IMAGE:figures/full_fig_p021_1.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 20 canonical work pages

  1. [1]

    and Xue, L

    Bhattacharjee, S., Li, B., Wu, X. and Xue, L. (2025) Doubly robust estimation of causal effects for random object outcomes with continuous treatments.arXiv:2506.22754

    arXiv 2025

  2. [2]

    (1953) Equivalent comparisons of experiments.Ann

    Blackwell, D. (1953) Equivalent comparisons of experiments.Ann. Math. Statist., 24, 265–272

    work page 1953

  3. [3]

    Clopper, C. J. and Pearson, E. S. (1934) The use of confidence or fiducial limits illustrated in the case of the binomial.Biometrika, 26, 404–413

    work page 1934

  4. [4]

    and Turner, K

    Curry, J., Mukherjee, S. and Turner, K. (2022) How many directions determine a shape and other sufficiency results for two topological transforms.Trans. Amer. Math. Soc. Ser. B, 9, 1006–1043. 19

    work page 2022

  5. [5]

    Eaton, M. L. (1989)Group Invariance Applications in Statistics. Institute of Mathematical Statistics, Hayward, CA

    work page 1989

  6. [6]

    Fisher, R. A. (1935)The Design of Experiments. Oliver and Boyd, Edinburgh

    work page 1935

  7. [7]

    M., Rasch, M

    Gretton, A., Borgwardt, K. M., Rasch, M. J., Sch¨ olkopf, B. and Smola, A. (2012) A kernel two-sample test.J. Mach. Learn. Res., 13, 723–773

    work page 2012

  8. [8]

    (1979) A simple sequentially rejective multiple test procedure.Scand

    Holm, S. (1979) A simple sequentially rejective multiple test procedure.Scand. J. Stat., 6, 65–70

    work page 1979

Show all 25 references
  1. [9]

    (2021)Foundations of Modern Probability, 3rd edn

    Kallenberg, O. (2021)Foundations of Modern Probability, 3rd edn. Springer, Cham

    2021

  2. [10]

    Kechris, A. S. (1995)Classical Descriptive Set Theory. Springer, New York

    1995

  3. [11]

    and M¨ uller, H.-G

    Kurisu, D., Zhou, Y., Otsu, T. and M¨ uller, H.-G. (2024) Geodesic causal inference. arXiv:2406.19604

    2024 arXiv

  4. [12]

    (1986)Asymptotic Methods in Statistical Decision Theory

    Le Cam, L. (1986)Asymptotic Methods in Statistical Decision Theory. Springer, New York

    1986

  5. [13]

    Lehmann, E. L. and Romano, J. P. (2005)Testing Statistical Hypotheses, 3rd edn. Springer, New York

    2005

  6. [14]

    (1979) A threshold selection method from gray-level histograms.IEEE Trans

    Otsu, N. (1979) A threshold selection method from gray-level histograms.IEEE Trans. Syst. Man Cybern., 9, 62–66

    1979

  7. [15]

    P., Luo, H., Strait, J

    Raykov, Y. P., Luo, H., Strait, J. D. and KhudaBukhsh, W. R. (2025) Kernel-based estimators for functional causal effects.arXiv:2503.05024

    2025 arXiv

  8. [16]

    [dataset] Recursion (2023) RxRx1: an image set for cellular morphological variation across many biological perturbations.https://www.rxrx.ai/rxrx1

    2023

  9. [17]

    Robins, J. M. (1986) A new approach to causal inference in mortality studies with a sustained exposure period.Math. Modelling, 7, 1393–1512

    1986

  10. [18]

    Rosenbaum, P. R. and Rubin, D. B. (1983) The central role of the propensity score in obser- vational studies for causal effects.Biometrika, 70, 41–55

    1983

  11. [19]

    Rubin, D. B. (1974) Estimating causal effects of treatments in randomized and nonrandomized studies.J. Educ. Psychol., 66, 688–701

    1974

  12. [20]

    and Oh, H.-S

    Shin, H.-Y., Kim, K., Lee, K. and Oh, H.-S. (2024) Absolute average and median treatment effects as causal estimands on metric spaces.arXiv:2407.03726

    2024 arXiv

  13. [21]

    K., Gretton, A., Fukumizu, K., Sch¨ olkopf, B

    Sriperumbudur, B. K., Gretton, A., Fukumizu, K., Sch¨ olkopf, B. and Lanckriet, G. R. G. (2010) Hilbert space embeddings and metrics on probability measures.J. Mach. Learn. Res., 11, 1517–1561

    2010

  14. [22]

    Sypetkowski, M. et al. (2023) RxRx1: a dataset for evaluating experimental batch correction methods.arXiv:2301.05768

    2023 arXiv

  15. [23]

    (1991)Comparison of Statistical Experiments

    Torgersen, E. (1991)Comparison of Statistical Experiments. Cambridge University Press, Cam- bridge. 20

    1991

  16. [24]

    and Boyer, D

    Turner, K., Mukherjee, S. and Boyer, D. M. (2014) Persistent homology transform for modeling shapes and surfaces.Information and Inference, 3, 310–344

    2014

  17. [25]

    Wijsman, R. A. (1990)Invariant Measures on Groups and Their Use in Statistics. Institute of Mathematical Statistics, Hayward, CA. Figure caption list Figure 1. Simulation rejection fractions across intrinsic effect strengths; causal validity under informative acquisition perta...

    1990

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.