Pith. sign in

REVIEW 3 major objections 5 minor 60 references

For linear regression with a conditionally Poisson covariate, the SIMEX estimator is strongly consistent for the slope and intercept when the measurement-error variance is known or consistently estimated.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-05 05:56 UTC pith:IYH24LAB

load-bearing objection First SIMEX consistency claim for Poisson surrogates, but the main theorem leans on an unproved interchange of limits (R3) that is essentially the desired result. the 3 major comments →

arxiv 2509.04709 v1 pith:IYH24LAB submitted 2025-09-04 math.ST stat.TH

Strong Consistency of the SIMEX Estimator in Linear Regression with a Conditionally Poisson Covariate

Aijun Yang , Mary Lesperance , Farouk S. Nathoo This is my paper
classification math.ST stat.TH MSC 62F1262J05
keywords measurement errorPoisson surrogateSIMEXstrong consistencylinear regressionheteroscedastic errorsingle replicatevariance estimation
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper tries to prove that the simulation-extrapolation (SIMEX) correction for measurement error still works when the error-prone covariate is a conditionally Poisson count, a non-Gaussian, heteroscedastic, discrete surrogate. It introduces POI-SIMEX, which adds Gaussian pseudo-noise to the observed counts and extrapolates back to zero total error, and shows in simple linear regression that the limiting estimator recovers the true slope and intercept almost surely. The reason the trick works is a variance identity: under the Poisson model the observed covariate variance decomposes as Var(X)+E(X), and the added Gaussian noise toggles the E(X) term so that extrapolating λ to -1 cancels it. The paper also shows the per-subject error variance can be consistently estimated from a single replicate using the sample mean count. If the proof is right, POI-SIMEX offers bias correction for Poisson-surrogate data without internal validation data.

Core claim

The central result is Theorem 3.1: under regularity conditions R1-R4, lim_{N→∞} lim_{λ→-1} E[β̂_{l,U}(λ) | {(Y_i,W_i)}] = β_l almost surely for l = 0, 1. Thus the SIMEX estimator, defined as the conditional expectation of the least-squares estimator computed on data with added Gaussian pseudo-errors, is strongly consistent for both coefficients. The proof computes the almost-sure limit of the pseudo-data slope as β_1 Var(X) / (Var(X) + (1+λ)E(X)), then observes that λ = -1 removes the extra E(X) term from the denominator, leaving β_1. Proposition 2.1 gives a strongly consistent estimator of the heteroscedastic measurement-error variance from the observed counts alone, and Corollary 3.2 exten

What carries the argument

The POI-SIMEX estimator: start with W_{i,U}(λ) = W_i + λ^{1/2} σ U_i with U_i standard normal, compute the least-squares slope on the pseudo-data, take the conditional expectation given (Y, W), then extrapolate to λ = -1. The Poisson-specific identities Var(W) = Var(X) + E(X), Cov(X, W) = Var(X), and Var(W - X) = E(X) make the pseudo-slope's almost-sure limit β_1 Var(X) / (Var(X) + (1+λ)E(X)); the point λ = -1 cancels the E(X) term in the denominator, leaving β_1. Proposition 2.1's estimator σ̂²_i = ar V / A_i supplies a strongly consistent variance estimate from a single replicate, and Slutsky's theorem carries the consistency through to the estimated-variance version.

Load-bearing premise

The argument assumes rather than proves that the order of 'sample size goes to infinity' and 'average over added Gaussian noise' can be swapped (condition R3), and it does not show that the curve fitted to a finite grid of λ values in the practical algorithm converges to the exact λ = -1 limit; if either fails, the consistency guarantee in Theorem 3.1 does not apply.

What would settle it

Simulate the exact model used in the paper (X ~ Gamma(1,10), β = (2,1,0.5), σ_ε = 5) with large N and many Monte Carlo draws of the Gaussian pseudo-errors, and compare the POI-SIMEX estimate computed at λ very close to -1 with the theoretical limit β_1 = 1. The theorem would be falsified if the estimate converges to a value other than 1 while R1-R4 hold; a more direct check of R3 is to compare the Monte Carlo average of the large-N pseudo-slope with the large-N limit of its conditional expectation for a fixed λ.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

Share X Bluesky LinkedIn Reddit HN

If this is right

  • For simple linear regression with a conditionally Poisson surrogate, POI-SIMEX is strongly consistent when the measurement-error variance is known or consistently estimated, with no validation subsample required.
  • The same Gaussian pseudo-error simulation used in standard SIMEX is valid for this non-Gaussian, heteroscedastic error structure, so existing SIMEX implementations can be adapted by substituting a per-subject variance estimate.
  • Per-subject measurement-error variance can be recovered from the average observed count, so the method applies when only one replicate of the surrogate is available.
  • The closed-form attenuation factor Var(X)/(Var(X) + (1+λ)E(X)) describes how Poisson noise shrinks the naive slope and exactly how extrapolating to λ = -1 corrects the shrinkage.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The theorem is proven only for simple linear regression with a single error-prone covariate; the numerical example includes an error-free covariate Z, but the proof does not, so extending strong consistency to multiple covariates is natural but unproven.
  • In practice POI-SIMEX fits an extrapolant (linear, quadratic, or nonlinear) to a finite grid of λ values, whereas the theorem treats the exact λ→-1 conditional expectation; finite-grid extrapolation error is an additional bias source the theorem does not quantify.
  • The variance identity relies on the conditional Poisson mean equaling the variance; if the surrogate is overdispersed, for instance zero-inflated or negative binomial, the cancellation at λ = -1 may be incomplete and the bias correction would need modification.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes POI-SIMEX, an adaptation of SIMEX for linear regression where the observed surrogate W is Poisson distributed conditional on the true covariate X. The main theoretical claim is strong consistency of the SIMEX estimator for the slope and intercept under simple linear regression, with the measurement-error variance either known or estimated. The proof derives the almost-sure limit of the naive slope estimator as β1 Var[X]/(Var[X]+(1+λ)E[X]) using the Poisson variance identity σ²=E[X], and then interchanges the N→∞ and λ→−1 limits to obtain β1. A numerical example with Gamma-distributed X illustrates finite-sample behavior.

Significance. If the double-limit result were established from primitive assumptions, the paper would fill a real gap: SIMEX theory has focused on homoscedastic additive Gaussian error, while conditional Poisson surrogates are heteroscedastic and discrete. The paper correctly identifies the variance identity σ²=E[X], and the algebra after the limit interchange is sound. It also gives a simple, strongly consistent estimator of the per-subject measurement-error variance from a single replicate. However, the central limit interchanges are assumed through regularity conditions R1–R3 rather than proved, and the theorem does not cover the fitted-extrapolation version of SIMEX that the algorithm section defines. In its current form, the advertised strong consistency is not demonstrated from model assumptions.

major comments (3)
  1. [Theorem 3.1, R3 and Eq. (3)] The proof's key step is the equality lim_N E[β̂_l,U(λ)|data_N] = E[lim_N β̂_l,U(λ)|data_∞], stated as R3. This interchange is not a consequence of the Poisson model or of R4. The proof establishes a.s. convergence of β̂_{1,U}(λ) via SLLN and Slutsky, but a.s. convergence does not imply convergence of conditional expectations: the denominator Σ(W_i−W̄+λ^{1/2}σ(U_i−Ū))² can be arbitrarily close to zero for finite N, and R4 only bounds moments of W and X, not the reciprocal denominator. Thus R3 is essentially the desired conclusion, and the theorem as stated is conditional on an unproved analytic fact. Either R3 must be derived from more primitive moment and positivity conditions, or the theorem should be rephrased as a conditional limit result without the claim of strong consistency from model assumptions.
  2. [Theorem 3.1, R1–R2] The interchanges of lim_N and lim_λ also require R1 (uniform convergence in λ near −1) and R2 (existence of the λ→−1 limit for each N). These are not verified, and they are not immediate from R4 or from the Poisson assumption. In particular, uniform control of E[β̂_{1,U}(λ)|data_N] as λ approaches −1 requires control of the reciprocal denominator over the full data distribution. The manuscript should either prove R1–R2 under explicit conditions on the distribution of X and the moments of W, or state clearly that Theorem 3.1 is a theorem about the exact double limit under unverified analytic conditions. As written, the conclusion 'the SIMEX estimator is therefore strongly consistent' overstates what has been shown.
  3. [Section 2.2 vs. Theorem 3.1] The theorem concerns the exact λ→−1 limit of the conditional expectation E[β̂_l,U(λ)|data], with U integrated out. The POI-SIMEX algorithm in Section 2.2 uses a finite number B of simulated Gaussian pseudo-error vectors and then fits a linear, quadratic, or nonlinear extrapolant over a grid of λ values, extrapolating to λ=−1. The paper does not prove that the finite-B average converges to the conditional expectation as B→∞, nor that the fitted extrapolant converges to the exact limit. Consequently, Theorem 3.1 does not cover the estimator actually implemented in Section 4. This is a load-bearing gap for the paper's claim that POI-SIMEX is strongly consistent.
minor comments (5)
  1. [Theorem 3.1, R3 statement] The RHS of R3 conditions on {(Y_i,W_i)_{i=1}^N} on both sides, but the proof uses conditioning on the infinite sequence. The notation should be corrected to make clear the conditioning sigma-field.
  2. [Corollary 3.2] The corollary states strong consistency with σ̂²=W̄, but 'the proof proceeds in the same way' is too terse. Replacing σ² by a consistent estimator changes the denominator limit; a short derivation of the resulting limit would be useful.
  3. [Proposition 2.1] The condition E[(V_i−m)^4]≤a<∞ appears unnecessary for the SLLN used, which only requires finite first moment for an i.i.d. sequence. Also 'strongly consistent estimator of σ_i², i=1,...,N' is awkward phrasing; the estimator uses all N observations to estimate each subject-specific variance.
  4. [Section 4, Numerical Example] The numerical example reports only boxplots for 200 datasets. It does not provide Monte Carlo standard errors, bias tables, or comparison with existing corrected-score methodology, even though the introduction claims SIMEX generally outperforms corrected score. Such details would strengthen the empirical evidence.
  5. [Throughout] There are typographical and reference issues: 'Kiichenhoff' should be 'Kuchenhoff'; the reference list has inconsistent formatting for several entries (e.g., Li et al. 2004 and Carroll et al. 1995); and the notation β̂_l,U(λ) is used before l is defined for l=0,1.

Circularity Check

0 steps flagged

No significant circularity: the POI-SIMEX consistency proof derives the attenuation limit from the Poisson model rather than fitting the target, so the central claim is not forced by construction. The main risk is an unverified regularity condition (R3), which is a completeness gap, not circularity.

full rationale

The paper's central claim (Theorem 3.1) is not obtained by fitting a parameter to the data it then predicts, and no load-bearing self-citation is used. The proof derives the a.s. limit of the SIMEX slope from the conditional Poisson structure: β̂_{1,U}(λ) → β1 Var[X]/(Var[X]+(1+λ)E[X]) using SLLN/Slutsky and the identities Var[W]=Var[X]+E[X], Cov[X,W]=Var[X], σ²=Var[W−X]=E[X]; this is a genuine model-derived calculation, and the λ→−1 limit recovers β1. The intercept follows from E[W]=E[X]. The statement is conditional on regularity conditions R1–R3, and in particular R3 ('lim_N E[β̂|D_N] = E[lim_N β̂|D_N]') is assumed, not established from the Poisson model; a.s. convergence of β̂ does not imply convergence of conditional expectations without a uniform-integrability argument. This is an omitted proof / unverified premise that weakens the claimed 'strong consistency from primitive assumptions,' but it is not circular: R3 is not the theorem's conclusion, and the paper does not define the estimator in terms of the target value or fit the consistency factor to data. The variance estimator in Proposition 2.1 is likewise justified from the Poisson mean-variance identity, not by assuming the conclusion. The theorem also addresses the idealized limit lim_{λ→−1} E[β̂|D_N], not the fitted extrapolant used in the implemented POI-SIMEX algorithm, which is another gap rather than a circular reduction. No enumerated circularity pattern applies.

Axiom & Free-Parameter Ledger

0 free parameters · 5 axioms · 0 invented entities

The derivation relies on the conditional Poisson error model, the linear regression structure, and unverified high-level regularity conditions. No ad hoc physical entities or fitted constants are introduced; the main unresolved burden is the limit-interchange assumptions R1-R3 and the unanalyzed extrapolation step.

axioms (5)
  • domain assumption W_i | X_i ~ Poisson(X_i A_i), with X_i i.i.d. and E[X_i]=m<∞
    The conditional Poisson error model is the entire setup, introduced in Section 2.1, Eq. (1)-(2).
  • domain assumption Y_i = β0 + β1 X_i + ε_i with ε_i i.i.d. N(0, τ²), and W_i ⊥ Y_i | X_i
    The linear regression and conditional independence structure, Section 2.1 Eq. (2).
  • ad hoc to paper Regularity conditions R1-R3 in Theorem 3.1 hold
    These are unverified limit-interchange assumptions that the proof needs; they are not derived from the Poisson model, and R3 effectively assumes the convergence used in the conclusion.
  • ad hoc to paper The fitted extrapolant (linear/quadratic/nonlinear) correctly represents E[[β̂_U(λ)|data] at λ=-1
    Section 2.2 describes the POI-SIMEX estimator via extrapolation, but Theorem 3.1 analyzes only the exact limit; no result shows the fitted curve's extrapolated value converges to the exact limit.
  • standard math SLLN, Slutsky's theorem, and Rudin's Theorem 7.11
    Invoked in the proof of Proposition 2.1 and Theorem 3.1.

pith-pipeline@v1.4.0-alltime-deepseek-medium · 7327 in / 15377 out tokens · 152306 ms · 2026-08-05T05:56:31.657194+00:00 · methodology

0 comments
Cite this review

Pith. "Pith review of Strong Consistency of the SIMEX Estimator in Linear Regression with a Conditionally Poisson Covariate." pith.science (2026). https://pith.science/paper/IYH24LAB

@misc{pith2026250904709,
  author       = {Pith},
  title        = {Pith review of: Strong Consistency of the SIMEX Estimator in Linear Regression with a Conditionally Poisson Covariate},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IYH24LAB}},
  note         = {Machine review of arXiv:2509.04709}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

This paper considers estimation for linear regression analysis with covariate measurement error arising from Poisson surrogates. We consider cases where covariates follow a conditional Poisson distribution, capturing non-Gaussian and heteroscedastic error structures. To address this, we extend the simulation extrapolation (SIMEX) algorithm to the conditional Poisson setting (POI-SIMEX), enabling robust adjustment in the absence of internal validation data. Theoretical analysis establishes strong consistency of the POI-SIMEX estimator under a linear regression framework.

Figures

Figures reproduced from arXiv: 2509.04709 by Aijun Yang, Farouk S. Nathoo, Mary Lesperance.

Figure 1
Figure 1. Figure 1: Boxplots of POI-SIMEX estimates for the regressio [PITH_FULL_IMAGE:figures/full_fig_p010_1.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

60 extracted references · 54 canonical work pages

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in ":" * " " * FUNCTION f...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in ":" * " " * FUNCTION f...

  3. [3]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in ":" * " " * FUNCTION f...

  4. [4]

    & Crainiceanu, C

    Carroll, R., Ruppert, D., Stefanski, L. & Crainiceanu, C. (1995). Measurement Error in Nonlinear Models. Chapman and Hall, London

    work page 1995

  5. [5]

    & Stefanski, L

    Carroll, R., Kiichenhoff, H., Lombard, F. & Stefanski, L. (1996). Asymptotics for the SIMEX estimator in nonlinear measurement error models. American Statistical Association. 91 pp. 242-250

    work page 1996

  6. [6]

    & Stefanski, L

    Carroll, R. & Stefanski, L. (1997).Asymptotic Theory for the Simex Estimator in Measurement Error Models. Statistics For Industry And Technology. https://doi.org/10.1007/978-1-4612-2308-5-10

    work page doi:10.1007/978-1-4612-2308-5-10 1997

  7. [7]

    & Hall, P

    Carroll, R., Delaigle, A. & Hall, P. (1999). Nonparametric Regression in the Presence of Measurement Error. Biometrika. 86, http://www.jstor.org/stable/2673653

    work page arXiv 1999

  8. [8]

    & Crainiceanu, C

    Carroll, R., Ruppert, D., Stefanski, L. & Crainiceanu, C. (2006) Measurement Error in Nonlinear Models: A Modern Perspective. Chapman and Hall, London

    work page 2006

  9. [9]

    & Stefanski, L

    Cook, J. & Stefanski, L. (1994) Simulation-Extrapolation Estimation in Parametric Measurement Error Models. J Am Stat Assoc. 89, https://doi.org/10.1080/01621459.1994.10476871

    arXiv 1994

  10. [10]

    & Stefanski, L

    Devanarayan, V. & Stefanski, L. (2002). Empirical simulation extrapolation for measurement error models with replicate measurements. Statistics & Probability Letters. 59, 219-225

    work page 2002

  11. [11]

    & Hall, P

    Carroll, R., Delaigle , A. & Hall, P. (2009). Nonparametric Prediction in Measurement Error Models. J Am Stat Assoc. 104, 10.1198/jasa.2009.tm07543

    work page doi:10.1198/jasa.2009.tm07543 2009

  12. [12]

    Ferguson, T. (1996). A Course in Large Sample Theory (1st ed.). Chapman, ISBN 9780412043710

    work page 1996

  13. [13]

    Fuller, W. (1986). Measurement error models. John Wiley & Sons, Inc., USA

    work page 1986

  14. [14]

    & Lesaffre, E

    Kuchenhoff, H., Mwalili, S. & Lesaffre, E. (2006). A General Method for Dealing with Misclassification in Regression: The Misclassification Simex. Biometrics. 62

    work page 2006

  15. [15]

    A measurement error model with a Poisson distributed surrogate

    Li, L, Palta, M, Shao, J (2004). A measurement error model with a Poisson distributed surrogate. Stat Med. 23(16), 2527-2536, https://doi: 10.1002/sim.1838

    work page doi:10.1002/sim.1838 2004

  16. [16]

    Nakamura, T. (1990). Corrected score functions for errors-in-variables models: methodology and application to generalized linear models. Biometrika.77

    work page 1990

  17. [17]

    Nakamura, T. (1992). Proportional hazards model with covariates subject to measurement error. Biometrika . 79, 387-395

    work page 1992

  18. [18]

    & Rosner, B

    Spiegelman, D., McDermott, A. & Rosner, B. (1997). Regression calibration method for correcting measurement-error bias in nutritional epidemiology. Am J Clin Nutr. 65, https://doi:10.1093/ajcn/65.4.1179S

    work page doi:10.1093/ajcn/65.4.1179s 1997

  19. [19]

    & Brenner, B

    Parveen, N., Moodie, E. & Brenner, B. (2021). The non-zero mean SIMEX: Improving estimation in the face of measurement error. Observational Studies. 1 pp. 90-123, https://api.semanticscholar.org/CorpusID:63160969

    work page 2021

  20. [20]

    & Gilks , W

    Richardson, S. & Gilks , W. (1993). A Bayesian approach to measurement error problems in epidemiology using conditional independence models.Am J Epidemiol. 138, https://doi:10.1093/oxfordjournals.aje.a116875

    work page doi:10.1093/oxfordjournals.aje.a116875 1993

  21. [21]

    Sarkar, A. (2014). Bayesian Semiparametric Density Deconvolution and Regression in the Presence of Measurement Errors. Doctoral Dissertation, Texas A & M University, https : / /hdl .handle .net /1969 .1 /153327

    work page 2014

  22. [22]

    Ross, S. (2006). Introduction to Probability Models (9th ed.). Academic Press

    work page 2006

  23. [23]

    Rudin, W. (1976). Principles of Mathematical Analysis (3rd ed.). McGraw-Hill

    work page 1976

  24. [24]

    Sevilimedu, V. & Yu, L. (2022). Simulation extrapolation method for measurement error: A review. Stat Med Med Res. 31, 1617-1636, https://doi:10.1177/09622802221102619

    work page doi:10.1177/09622802221102619 2022

  25. [25]

    & Wang, S

    Sinha, S. & Wang, S. Semiparametric Bayesian analysis of censored linear regression with errors-in-covariates. Statistical Methods In Medical Research. 26, 1389-1415, doi:10.1177/0962280215580668

    work page doi:10.1177/0962280215580668

  26. [26]

    & Gustafson, P

    Yi, G., Delaigle, A. & Gustafson, P. (2021) Handbook of Measurement Error Models (1st ed.). Chapman and Hall/CRC, New York, https://doi.org/10.1201/9781315101279

    doi:10.1201/9781315101279 2021

  27. [27]

    Kong, F., Huang, W. & Li, X. (1998). Estimating survival curves under proportional hazards model with covariate measurement errors. Scand. J. Statist. 25 pp. 573-587

    work page 1998

  28. [28]

    Kong, F. & Gu, M. (1999) Consistent estimation in Cox proportional hazards model with covariate measurement errors. Statistica Sinica. 9 pp. 953-969

    work page 1999

  29. [29]

    & Davidian, M

    Tsiatis, A. & Davidian, M. (2001) A semiparametric estimator for the proportional hazards model with longitudinal covariates measured with error. Biometrika. 88, 447-458

    work page 2001

  30. [30]

    Augustin, T. (2004). An Exact Corrected Log-Likelihood Function for Cox’s Proportional Hazards Model under Measurement Error and Some Extensions. Scandinavian Journal Of Statistics. 31, 43-50, https://www.jstor.org/stable/4616810

    work page arXiv 2004

  31. [31]

    & Lawless, J.(2007) A corrected likelihood method for the proportional hazards model with covariates subject to measurement error

    Yi, G. & Lawless, J.(2007) A corrected likelihood method for the proportional hazards model with covariates subject to measurement error. J Stat Plan Infer. 137 pp. 1816-1828

    work page 2007

  32. [32]

    & Yi, G.(2015)

    Yan, Y. & Yi, G.(2015). A corrected profile likelihood method for survival data with covariate measurement error under the Cox model. Can J Stat. 43, 454-480

    work page 2015

  33. [33]

    & Wong, M

    Cao, Z. & Wong, M. (2022). Approximate profile likelihood estimation for Cox regression with covariate measurement error. Stat Med.. 41, 910-931, https://doi: 10.1002/sim.9324

    work page doi:10.1002/sim.9324 2022

  34. [34]

    Prentice, R. (1982). Covariate measurement errors and parameter estimation in a failure time regression model. Biometrika. 69, 331-342

    work page 1982

  35. [35]

    & Prentice, R

    Wang, C., Hsu, L., Feng, Z. & Prentice, R. (1997)

    work page 1997

  36. [36]

    & Wang, X

    Wang, L. & Wang, X. (2008). A new method for survival analysis with covariates subject to measurement error. Joural Of The Royal Statistics Society:Series B. 70, 519-530

    work page 2008

  37. [37]

    & Xiong, J

    He, W., Yi, G. & Xiong, J. Accelerated failure time models with covariates subject to measurement error. Stat Med. 26, 4817-4832

    work page

  38. [38]

    Zhang, J., He, W. & H, L. (2014). A semiparametric approach for accelerated failure time models with covariates subject to measurement error. Commun Stat Simul Comput. 43, 329-341

    work page 2014

  39. [39]

    & Cai, J

    Greene, W. & Cai, J. Measurement error in covariates in the marginal hazards model for multivariate failure time data. Biometrics. 60, 987-996

    work page

  40. [40]

    Tang, Y., Song, X. & Yi, G. (2022) Bayesian analysis under accelerated failure time models with error-prone time-to-event outcomes. Lifetime Data Anal. 28 pp. 139-168, https://doi.org/10.1007/s10985-021-09543-3

    work page doi:10.1007/s10985-021-09543-3 2022

  41. [41]

    Havasi, A., Cainap, S. (2023). Ovarian Cancer—Insights into Platinum Resistance and Overcoming It. Medicina . 59, https://doi.org/10.3390/medicina59030544

    work page doi:10.3390/medicina59030544 2023

  42. [42]

    Kononen, J., Bubendorf, L., Kallionimeni, A., Bärlund, M., Schraml, P., Leighton, S.B., Torhorst, J., Mihatsch, M.J., Sauter, G., & Kallionimeni, O. (1998). Tissue microarrays for high-throughput molecular profiling of tumor specimens. Nature Medicine, 4, 844-847

    work page 1998

  43. [43]

    BioChain Institute Inc. (2024). The Tissue Microarray: Tissue Analysis from Research to Treatment - Blog-. , Accessed at 2024-2-26. https://www.biochain.com/blog/the-tissue-microarray-tissue-analysis-from-research-to-treatment/

    work page 2024

  44. [44]

    & Boyd, K

    Eskaros, A., Egloff, S. & Boyd, K. (2017). Larger core size has superior technical and analytical accuracy in bladder tissue microarray. Lab Invest. 97 pp. 335-342, https://doi.org/10.1038/labinvest.2016.151

    work page doi:10.1038/labinvest.2016.151 2017

  45. [45]

    & Guido, S

    Ronald, S. & Guido, S. (2002). Tissue microarrays for miniaturized high-throughput molecular profiling of tumors.Experimental Hematology. 30, 1365-1372, https://doi.org/10.1016/S0301-472X(02)00965-7

    work page doi:10.1016/s0301-472x(02)00965-7 2002

  46. [46]

    & Carroll, R.(2000)

    Lin, X. & Carroll, R.(2000). Nonparametric function estimation for clustered data when the predictor is measured without/with error. Am Stat Assoc. 95 pp. 520-534

    work page 2000

  47. [47]

    Efron, B. (1979). Bootstrap Methods: Another Look at the Jackknife. The Annals Of Statistics. 7, 1-26, http://links.jstor.org/sici?sici=0090-5364

    work page 1979

  48. [48]

    & Hanley, G

    Arora, N., Talhouk, A., McAlpine, J., Law, M. & Hanley, G. (2018). Long-term mortality among women with epithelial ovarian cancer: a population-based study in British Columbia, Canada. BMC Cancer. 18 pp. 1039, https://doi.org/10.1186/s12885-018-4970-9

    work page doi:10.1186/s12885-018-4970-9 2018

  49. [49]

    & Drapkin, V

    Labidi-Galy, S., Papp, E., Hallberg, D., Niknafs, N., Adleff, V., Noe, M., Bhattacharya, R., Novak, M., Jones, S., Phallen , J., Hruban, C., Hirsch, M., Lin, D., Schwartz, L., Maire, C., Tille, J., Bowden, M., Ayhan, Wood, L., Scharpf, R., Kurman, T., Shih, I., Karchi, R. & Drapkin, V. (2017). High grade serous ovarian carcinomas originate in the fallopia...

    work page doi:10.1038/s41467-017-00962-1 2017

  50. [50]

    & Zou, W

    Curiel, T., Coukos, G., Zou, L., Alvarez, X., Cheng , P., Mottram, P., Evdemon-Hogan, M., Conejo-Garcia, J., Zhang, L., Burow, Zhu, Y., Wei, S., Kryczek, I., Daniel, B., Gordon, A., Myers, L., Lackner, A., Disis, M., Knutson, K., Chen, L. & Zou, W. Specific recruitment of regulatory T cells in ovarian carcinoma fosters immune privilege and predicts reduce...

    doi:10.1038/nm1093 2004

  51. [51]

    Curiel, T. (2008). Regulatory T cells and treatment of cancer. Current Opinion In Immunology. 20, 241-246, https://www.sciencedirect.com/science/article/pii/S0952791508000502

    work page 2008

  52. [52]

    Elkoshi, Z. (2022). On the Prognostic Power of Tumor Infiltrating Lymphocytes – A Critical Commentary. Front. Immunol. 13, https:doi: 10.3389/fimmu.2022.892543

    work page arXiv 2022

  53. [53]

    & Knutson, K

    Preston, C., Maurer, M., Oberg, A., Visscher, D., Kalli, K., Hartmann, L., Goode, E. & Knutson, K. (2013). The ratios of CD8+ T cells to CD4+CD25+ FOXP3+ and FOXP3- T cells correlate with poor clinical outcome in human serous ovarian cancer. PLoS One. 8, https://10.1371/journal.pone.0080063

    work page doi:10.1371/journal.pone.0080063 2013

  54. [54]

    & Aloulou, M.(2022)

    Liston, A. & Aloulou, M.(2022). A fresh look at a neglected regulatory lineage: CD8+Foxp3+ Regulatory T cells. Immunology Letters. 247, https://doi.org/10.1016/j.imlet.2022.05.004

    work page doi:10.1016/j.imlet.2022.05.004 2022

  55. [55]

    & Jirstrom, K

    Warfvinge, C., Lundgren, S., Elebro, J., Margareta Heby, M., Krzyzanowska, A., Bjartell, A., Nodin, B., Jakob Eberhard, J., Leandersson, K. & Jirstrom, K. (2017). The prognostic impact of CD3, CD8, FoxP3, and IL17 tumor-infiltrating immune cells in periampullary cancer differs by morphological type and adjuvant chemotherapy. Clinical Oncology. 35, https:/...

    work page doi:10.1200/jco.2017.35.7-suppl.53 2017

  56. [56]

    Leffers, N., Gooden, M., Jong, R., Hoogeboom , B., Hoor, K., Hollema, H. & et al. (2009). Prognostic Significance of Tumor-Infiltrating T-Lymphocytes in Primary and Metastatic Lesions of Advanced Stage Ovarian Cancer. Cancer Immunol Immun. 58, https://doi: 10.1007/s00262-008-0583-5

    work page doi:10.1007/s00262-008-0583-5 2009

  57. [57]

    & Nelson, B.(2009)

    Milne, K., Köbel, M., Kalloger, S., Barnes , R., Gao, D., Gilks, C., Watson, P. & Nelson, B.(2009). Systematic Analysis of Immune Infiltrates in High-Grade Serous Ovarian Cancer Reveals CD20, FoxP3 and TIA-1 as Positive Prognostic Factors. PLoS ONE. 4, https://doi.org/10.1371/journal.pone.0006412

    work page doi:10.1371/journal.pone.0006412 2009

  58. [58]

    Myeloid-derived suppressor cells: more mechanisms for inhibiting antitumor immunity

    Ostrand-Rosenberg, S.(2010). Myeloid-derived suppressor cells: more mechanisms for inhibiting antitumor immunity. Cancer Immunology, Immunotherapy. 59, 1593-1600, https://doi.org/10.1007/s00262-010-0855-8

    work page doi:10.1007/s00262-010-0855-8 2010

  59. [59]

    & Nagaraj, S

    Gabrilovich, D. & Nagaraj, S. 2009. Myeloid-derived suppressor cells as regulators of the immune system. Nature Reviews Immunology. 9, 162-174, https://doi.org/10.1038/nri2506

    work page doi:10.1038/nri2506 2009

  60. [60]

    & Zhang, H

    Liu, C., Yu, S., Kappes, J., Wang, J., Grizzle, W., Zinn, K. & Zhang, H. 2015. Expansion of Myeloid-Derived Suppressor Cells (MDSCs) in Cancer Is Driven by MTDH-Dependent IL-6 Secretion. Cellular And Molecular Immunology . 12, 609-622, https://doi.org/10.1038/cmi.2014.84

    work page doi:10.1038/cmi.2014.84 2015