Pith. sign in

REVIEW 2 major objections 4 minor 180 references

A simulation-quantile threshold can control false alarms in online change point detection over any moving monitoring window.

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

2026-08-01 23:23 UTC pith:TDG3LTL7

load-bearing objection A genuinely useful new error-rate definition and a simple simulation-based calibration that works in the settings tested, but the paper never checks the central claim that control holds for every later monitoring window, not just the first. the 2 major comments →

arxiv 2607.15423 v1 pith:TDG3LTL7 submitted 2026-07-16 stat.ME

Sequential Control of False Positives in Online Change Point Detection

Melissa Lynne Martin , Theodore D. Satterthwaite , Ian J. Barnett This is my paper
classification stat.ME MSC 62L1062F0362P10
keywords online change point detectionsequential family-wise error ratemultiple testingmonitoring threshold calibrationnull simulationmoving windowmobile healthvariance component score test
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.

Online change point detection performs a new hypothesis test on each arriving observation, so false alarms accumulate in a way standard multiple-testing corrections do not handle. This paper introduces the sequential family-wise error rate (sFWER), the probability of at least one false alarm within any moving window of length Δ, and proposes to calibrate a single monitoring cutoff from simulated null data. The cutoff is the empirical α_Δ quantile of the distribution of the minimum of the monitoring statistics over the window. Simulation studies show the cutoff holds the sFWER near its target while common alternatives are either too conservative (fixed per-test corrections) or too liberal (unadjusted day-by-day thresholds). If correct, practitioners can choose a detection threshold directly from a clinically meaningful false-alarm window.

Core claim

The paper's central claim is that under the global null, the distribution of M = min_{t+1≤j≤t+Δ} S_j — the smallest daily monitoring statistic in a window — determines a correct monitoring cutoff. Generate B null datasets of length r+Δ, compute M for each, and set c to the empirical α_Δ quantile of these values. Then for every t ≥ r, the probability that the window {t+1,...,t+Δ} contains a false alarm is approximately α_Δ. The paper demonstrates the claim for the VC* statistic, a variance-component score test whose monitoring statistic is the minimum permutation p-value over recent candidate change-point days, and reports that the proposed threshold hits the target while corrections that ign

What carries the argument

The key object is the window-minimum statistic M, the smallest of the Δ daily monitoring statistics in a sliding window. Because the daily statistics use overlapping data, their joint distribution—and hence the distribution of M—has no closed form. The paper approximates this distribution by simulation under the null and defines the cutoff c as the empirical α_Δ quantile of M; a change is declared on day t whenever S_t ≤ c. In the demonstrations, S_t is the VC* monitoring statistic, defined as the minimum permutation p-value over candidate change-point days in the last d_b days.

Load-bearing premise

The distribution of the monitoring statistic under 'no change' is the same for every window after the run-in, so a threshold learned from the first window keeps its error rate later.

What would settle it

Simulate a no-change series with slow drift or autocorrelation, calibrate c under the paper's i.i.d. null model, then record the fraction of later Δ-windows (starting at various t beyond r) that contain at least one rejection; if that fraction clearly exceeds α_Δ, the sFWER guarantee fails.

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

If this is right

  • Practitioners can set a monitoring threshold directly from a desired false-alarm rate per Δ-day window, without knowing the number of future tests.
  • The same cutoff is claimed to control the sFWER for every window after the run-in, not just the calibration window, provided the null distribution is stationary.
  • The calibrated threshold preserves substantially more detection power than corrections that ignore the dependence among overlapping test statistics.
  • The procedure applies to multivariate data and can be adapted to any test statistic whose null window-minimum distribution can be simulated.
  • In daily mHealth monitoring the extra computation is small enough to be practical, since only null simulations are needed up front.

Where Pith is reading between the lines

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

  • The sFWER guarantee is only as good as the stationarity of the null: the cutoff is calibrated from the first window, and the paper's sensitivity check varies run-in length but never monitors later windows under a drifting null. Testing that directly would settle how robust the threshold is.
  • Because only the window-minimum distribution is used, the same calibration recipe should transfer to other change-point statistics (CUSUM, likelihood-ratio, etc.) with no theoretical change—only the simulation step differs.
  • For special null models (e.g., Gaussian AR processes), an analytic approximation of the window-minimum distribution might replace the simulation, removing Monte Carlo uncertainty and making the method lighter for high-frequency monitoring.
  • The real-data application resets monitoring after each detected change; the stated sFWER bound applies within one reset segment rather than across the whole follow-up, so cumulative false-alarm risk over an entire study may be higher than α_Δ.

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

2 major / 4 minor

Summary. The paper introduces a sequential family-wise error rate (sFWER), defined as the probability of at least one false positive within a moving monitoring window of length Δ, and proposes a simulation-based calibration of a cutoff c for an online change point detection statistic. The cutoff is the empirical α_Δ quantile of the minimum of the monitoring statistic over the first monitoring window under a null model. The authors claim this controls the sFWER for any window after the run-in period, and present simulations comparing the proposed procedure with unadjusted, Bonferroni, and Šidák thresholds under the null and alternative hypotheses. They also illustrate the procedure on smartphone mobility data from an mHealth study. The central claim is that the calibrated cutoff balances error control and power better than standard corrections, which are either conservative or liberal.

Significance. If the central claim holds, the paper offers a practical, simulation-based alternative to Bonferroni/Šidák for online monitoring with strongly dependent test statistics. The sFWER definition is a useful formalization for mHealth and similar settings, and the simulation study is designed in a non-circular way: the cutoff is calibrated under an explicit null model and then evaluated on independent null datasets. The paper also transparently acknowledges the stationarity assumption and the illustrative nature of the real-data analysis. However, the simulations verify error control only for the first monitoring window, so the paper's principal guarantee—control for every moving window—is not empirically supported as stated.

major comments (2)
  1. [Section 2.2 and Eq. (1)] The cutoff c is calibrated as the empirical α_Δ quantile of M = min_{r+1≤j≤r+Δ} S_j from null datasets of length r+Δ. This directly controls the first monitoring window only. Eq. (1), however, requires P(∪_{j=t+1}^{t+Δ}{S_j≤c}) ≤ α_Δ for every t≥r. Section 3.1 varies the run-in length r, which shifts the pre-window data, but never evaluates windows beginning at later times. Section 3.2's null evaluation also uses datasets of length r+Δ and checks only the window (r+1,...,r+Δ). The claim that the procedure controls the sFWER for any moving window is therefore supported only for the first window. Please add simulations with long null series (T >> r+Δ) and report the empirical false-alarm rate for windows starting at t = r+1, r+2, ..., T-Δ, or at minimum at t = r+1 and t = r+Δ+1. Without such evidence, Eq. (1) remains an assumption rather than a demonstrated property.
  2. [Section 2.3 and Eq. (4)] Even under i.i.d. Gaussian null data, the distribution of S_t = min_{t-db≤k≤t-1} p_k may depend on t because each candidate p-value p_k is computed using data accumulated up to time t, and the amount of pre-candidate data changes as k increases. The Discussion acknowledges that the procedure assumes the data are stable under the null, but this stationarity of the raw data does not automatically imply stationarity of the joint distribution of (S_{t+1},...,S_{t+Δ}) for different t. The paper should either provide a formal argument or empirical evidence that the null distribution of S_t is invariant across monitoring days for the VC* statistic. In particular, Section 3.1's check across run-in lengths is not a check of later-window behavior. If the null distribution of S_t changes with t, the calibrated cutoff may fail to control the sFWER for later windows even under a perfectly stationary
minor comments (4)
  1. [Table 1] The unadjusted procedure's rejection rates are identical for α_Δ=0.1 and α_Δ=0.2 (e.g., 0.18 for p=1, Δ=7), which is expected because it uses a fixed 0.05 threshold. This duplication is not explained in the table or text; a note would avoid confusion.
  2. [Section 3.2] The Monte Carlo sample size is B=1000. For a target α_Δ=0.1, the standard error of the estimated type I error is approximately 0.0095; for α_Δ=0.2, approximately 0.013. The observed values (e.g., 0.23 for p=10, α_Δ=0.2, Δ=7) are within reasonable Monte Carlo variability, but adding standard errors or confidence intervals to Table 1 would help the reader assess deviations from the target.
  3. [Section 4.2] The real-data calibration uses a pooled correlation matrix estimated from all available data, including data after potential change points. Because the analysis is retrospective and explicitly illustrative, this is not fatal, but the text should state more clearly that the null model for cutoff calibration is not independent of the data being monitored. Supplement S2 is cited for insensitivity to correlation structure, but the main text would benefit from one sentence explaining this limitation.
  4. [Figure 1 caption] The caption says error bars show variability across bootstrap samples, but the number of bootstrap resamples and the resampling scheme are not specified in the main text. Please provide those details.

Circularity Check

2 steps flagged

The null sFWER 'demonstration' re-evaluates the fitted quantile, and the any-window guarantee is asserted from a first-window calibration.

specific steps
  1. fitted input called prediction [Section 3.2 (sFWER evaluation), Table 1; Section 2.2 calibration]
    "To evaluate sFWER, we generate an additional B=1000 null datasets of length r+Δ and record whether at least one false alarm occurred during the monitoring window, i.e., whether M≤c."

    The cutoff c was defined in Section 2.2 as the empirical α_Δ quantile of M = min_{r+1≤j≤r+Δ} S_j under null simulations of length r+Δ. The null sFWER evaluation counts the event M≤c on new datasets of the same length and same null model. This is the empirical CDF at the fitted quantile, so the rejection rate is forced to approximate α_Δ up to Monte Carlo error. The claimed 'demonstration' of error control is therefore a restatement of the calibration step, not an independent prediction.

  2. self definitional [Section 2.2, Eq. (1)-(3)]
    "By construction, this procedure controls the probability of at least one false alarm within any monitoring window of length Δ at approximately α_Δ."

    The construction in Section 2.2 generates only the first monitoring window (r+1,...,r+Δ) and sets c to its empirical quantile. Eq. (1) requires control for every t≥r. For t>r+1, control does not follow from the construction; it requires the additional, untested assumption that the joint null distribution of (S_{t+1},...,S_{t+Δ}) equals that of (S_{r+1},...,S_{r+Δ}). The phrase 'any monitoring window' converts the first-window quantile control into a universal guarantee by definition, though the later-window property is never simulated or verified.

full rationale

The core derivation has a genuine circular component: the simulation-based cutoff is the empirical quantile of M, and the null sFWER evaluation uses the same M over the same window length, so the empirical rejection rate is the fitted quantile's CDF evaluated at itself. This validation cannot fail to approximate the target. Separately, the central inequality Eq. (1) promises control for all t≥r, but the calibration only constructs the first window; the paper's 'By construction ... any monitoring window' is an overclaim unless the null distribution is stationary across later windows, a condition the simulations never exercise (they stop at r+Δ). The Discussion admits the stationarity assumption ('the proposed calibration procedure assumes that the data are stable under the null hypothesis'), which confirms the gap but does not make it circular. The self-citations to Martin et al. (2026) for the VC* statistic and preprocessing are not load-bearing for the calibration logic; the real-data application's use of full data to estimate the pooled correlation matrix is a retrospective data-dependence acknowledged as illustrative. Overall, parts of the central claim reduce to construction, but the method retains independent content in power comparisons and practical implementation, yielding partial circularity rather than complete equivalence.

Axiom & Free-Parameter Ledger

5 free parameters · 4 axioms · 0 invented entities

The central claim rests on the null simulation model, the first-window-to-all-windows stationarity assumption, and the validity of the VC* permutation p-values. The calibrated cutoff c is not a hidden fitted parameter in the simulations (it is the object being estimated), but in the real-data application, correlation matrices are computed from all available data, which weakens the prospective interpretation. No invented entities are introduced.

free parameters (5)
  • Calibrated detection threshold c
    Empirical α_Δ quantile of simulated null window minimum; the object the procedure estimates, not a hidden tuning parameter.
  • Candidate lookback db = 7
    Fixed to 7 in all simulations and application; affects dependence of S_t.
  • Run-in length r = 30 (simulations), 14 (application)
    Chosen by hand; underlies the stationarity assumption.
  • Simulation size B / permutation resamples = B=1000; 5000 permutations
    Computational choices that set Monte Carlo error of the cutoff and p-values.
  • Covariance regularization parameter = 0.1
    Used only in the real-data application for covariance estimation.
axioms (4)
  • domain assumption The monitoring statistic sequence S_t is stationary under the null after the run-in period, so the first-window null distribution of M equals the distribution for every later window.
    Section 2.2 calibrates using only days r+1..r+Δ (Eq. 1 requires control for every t≥r); Section 3.1 tests sensitivity to r but does not verify windows at arbitrary t.
  • domain assumption The null data generation model (i.i.d. N_p(0,I_p) in simulations; pooled correlation matrix in application) correctly represents the real data-generating process under no change point.
    Calibration is only as valid as the simulation model; acknowledged in Discussion ('assumes that the data are stable under the null').
  • domain assumption Permutation-based p-values from VC* are valid under the null (uniform or conservative) so that S_t=min p_k is a meaningful statistic.
    Section 2.3 relies on permutation p-values from Martin et al. (2026); no proof of validity is given in this paper.
  • standard math Standard probability/quantile estimation (Monte Carlo) is valid.
    Empirical quantile converges to the true quantile as B grows; used throughout Section 2.2.

pith-pipeline@v1.3.0-alltime-deepseek · 8646 in / 12645 out tokens · 113307 ms · 2026-08-01T23:23:37.079711+00:00 · methodology

0 comments
read the original abstract

Online change point detection is the process of identifying distributional changes in time-ordered data in real time. In applications such as mobile health (mHealth), repeated testing is often performed as new data arrive, creating a multiple testing problem. Traditional approaches for controlling the family-wise error rate (FWER) are not well suited to this setting because the tests are highly dependent and the number of tests is not fixed in advance. In this work, we introduce a sequential family-wise error rate (sFWER), defined as the probability of at least one false positive within a moving monitoring window. We propose a simulation-based calibration procedure to estimate monitoring thresholds that control the sFWER at a desired level. Through simulation studies, we demonstrate that the proposed procedure achieves the desired error control, while commonly used alternatives are either overly conservative or fail to adequately control false alarms. Finally, we illustrate the proposed approach using passively collected smartphone data from a cohort of adolescents and young adults with affective instability.

Figures

Figures reproduced from arXiv: 2607.15423 by Ian J. Barnett, Melissa Lynne Martin, Theodore D. Satterthwaite.

Figure 1
Figure 1. Figure 1: Estimated calibrated cutoffs across run-in lengths. Points show the estimated cutoff, and [PITH_FULL_IMAGE:figures/full_fig_p012_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Detected change points for each participant using the proposed monitoring procedure. [PITH_FULL_IMAGE:figures/full_fig_p013_2.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

180 extracted references · 104 canonical work pages · 3 internal anchors

  1. [1]

    Biometrika , author =

    Group sequential methods in the design and analysis of clinical trials , volume =. Biometrika , author =. 1977 , pages =. doi:10.1093/biomet/64.2.191 , abstract =

    doi:10.1093/biomet/64.2.191 1977

  2. [2]

    Journal of the Royal Statistical Society Series B: Statistical Methodology , author =

    Optimal. Journal of the Royal Statistical Society Series B: Statistical Methodology , author =. 2003 , pages =. doi:10.1111/1467-9868.00389 , abstract =

    arXiv 2003

  3. [3]

    Hui, S. L. and Walter, S. D. , year =. Estimating the. Biometrics , publisher =. doi:10.2307/2530508 , abstract =

    work page doi:10.2307/2530508

  4. [5]

    doi:10.1002/sim.9097 , abstract =

    Estimation of diagnostic test accuracy:. doi:10.1002/sim.9097 , abstract =

    work page doi:10.1002/sim.9097

  5. [6]

    Statistics in Biopharmaceutical Research , author =

    Evaluating. Statistics in Biopharmaceutical Research , author =

  6. [7]

    Estimating

    Bradbrook, Keighly , month = jan, year =. Estimating. doi:https://doi.org/10.25772/9RB0-1G73 , journal =

    work page doi:10.25772/9rb0-1g73

  7. [8]

    Evaluating and

    He, Jun , month = jan, year =. Evaluating and. doi:https://doi.org/10.25772/W7XY-9056 , journal =

    work page doi:10.25772/w7xy-9056

  8. [9]

    and Hackworth, Olivia and Almirall, Daniel , year =

    Seewald, Nicholas J. and Hackworth, Olivia and Almirall, Daniel , year =. Sequential,. Principles and. doi:10.1007/978-3-319-52677-5_280-1 , pages =

    work page doi:10.1007/978-3-319-52677-5_280-1

  9. [10]

    Isabel and Ortega, María Victoria and Garrido-Aranda, Alicia and Lavado-Valenzuela, Rocío and Álvarez, Martina and Alba, Emilio and Comino-Méndez, Iñaki , month = may, year =

    Velasco-Suelto, Jesús and Quirós-Ortega, María Elena and Pascual, Javier and Godoy-Ortiz, Ana and Alba-Bernal, Alfonso and López-López, Esperanza and Domínguez-Recio, María Emilia and Jiménez-Rodríguez, Begoña and Peralta-Linero, Jesús and Bellagarza-García, Estefanía and Troyano-Ramos, Laura and Garrido-Ruiz, Guadalupe and Hierro-Martín, M. Isabel and Or...

  10. [11]

    and To, Briana and Mestres-Villanueva, Mariella A

    Prasath, Vishnu and Jhawar, Sachin R. and To, Briana and Mestres-Villanueva, Mariella A. and Schneider, Bryan P. and Reátegui, Eduardo and Stover, Daniel G. , month = oct, year =. Detection and. JCO Precision Oncology , publisher =. doi:10.1200/PO-25-00170 , number =

    work page doi:10.1200/po-25-00170

  11. [12]

    and Pan, Tien-chi and Pant, Dhruv and Chen, Yan and Woodfield, George and Graves, Jewell and Lawrence-Paul, Matthew R

    Chislock, Elizabeth M. and Pan, Tien-chi and Pant, Dhruv and Chen, Yan and Woodfield, George and Graves, Jewell and Lawrence-Paul, Matthew R. and Belka, George K. and Wang, Jianping and Bayne, Lauren and Blanchard, Tatiana and Smith, Meaghan and Ji, Xiaodan and Shih, Natalie N. C. and Soucier-Ernst, Danielle and Goodman, Noah and Nivar, Isoris and Goodspe...

  12. [13]

    and Manon, Nayelis A

    Magbanua, Mark Jesus M. and Manon, Nayelis A. and Wolf, Denise M. and Rivero-Hinojosa, Samuel and Ahmed, Ziad and Sayaman, Rosalyn W. and Tin, Antony and Renner, Derrick and Kalashnikova, Ekaterina and Brown-Swigart, Lamorna and Hirst, Gillian L. and Yau, Christina and Li, Wen and Isaacs, Claudine and Shatsky, Rebecca A. and Clark, Amy S. and Zimmer, Alex...

  13. [14]

    Nature Medicine , author =

    Targeting dormant tumor cells to prevent recurrent breast cancer: a randomized phase 2 trial , volume =. Nature Medicine , author =. 2025 , keywords =. doi:10.1038/s41591-025-03877-3 , abstract =

    work page doi:10.1038/s41591-025-03877-3 2025

  14. [15]

    2022 , pages =

    Annals of Behavioral Medicine , author =. 2022 , pages =. doi:10.1093/abm/kaab061 , abstract =

    work page doi:10.1093/abm/kaab061 2022

  15. [16]

    Journal of Clinical Epidemiology , author =

    Sequential multiple assignment randomized trial studies should report all key components: a systematic review , volume =. Journal of Clinical Epidemiology , author =. 2022 , keywords =. doi:10.1016/j.jclinepi.2021.11.007 , abstract =

    work page doi:10.1016/j.jclinepi.2021.11.007 2022

  16. [17]

    Patient Education and Counseling , author =

    Adaptive web-based stress management programs among adults with a cardiovascular disease:. Patient Education and Counseling , author =. 2022 , keywords =. doi:10.1016/j.pec.2021.10.020 , abstract =

    work page doi:10.1016/j.pec.2021.10.020 2022

  17. [18]

    Pain Medicine , author =

    The design and rationale of the. Pain Medicine , author =. 2025 , pages =. doi:10.1093/pm/pnaf032 , abstract =

    work page doi:10.1093/pm/pnaf032 2025

  18. [19]

    Research in Nursing & Health , author =

    Reflexology and meditative practices for symptom management among people with cancer:. Research in Nursing & Health , author =. 2021 , note =. doi:10.1002/nur.22169 , abstract =

    work page doi:10.1002/nur.22169 2021

  19. [20]

    Contemporary Clinical Trials , author =

    A sequential multiple assignment randomized trial (. Contemporary Clinical Trials , author =. 2020 , keywords =. doi:10.1016/j.cct.2020.106089 , abstract =

    arXiv 2020

  20. [21]

    Nicotine & Tobacco Research , author =

    Prize-. Nicotine & Tobacco Research , author =. 2025 , pages =. doi:10.1093/ntr/ntae243 , abstract =

    work page doi:10.1093/ntr/ntae243 2025

  21. [22]

    2024 , keywords =

    Contemporary Clinical Trials , author =. 2024 , keywords =. doi:10.1016/j.cct.2024.107444 , abstract =

    arXiv 2024

  22. [23]

    Dynamic treatment regimes: practical design considerations , volume =

    Lavori, Philip W and Dawson, Ree , month = feb, year =. Dynamic treatment regimes: practical design considerations , volume =. Clinical Trials , publisher =. doi:10.1191/1740774504cn002oa , abstract =

    work page doi:10.1191/1740774504cn002oa

  23. [24]

    Statistics in Medicine , author =

    An experimental design for the development of adaptive treatment strategies , volume =. Statistics in Medicine , author =. 2005 , note =. doi:10.1002/sim.2022 , abstract =

    work page doi:10.1002/sim.2022 2005

  24. [25]

    and Hussain, Cory and Lee, Jennifer A

    Chaparro, Juan D. and Hussain, Cory and Lee, Jennifer A. and Hehmeyer, Jessica and Nguyen, Manjusri and Hoffman, Jeffrey , month = jan, year =. Reducing. Applied Clinical Informatics , publisher =. doi:10.1055/s-0039-3402757 , abstract =

    work page doi:10.1055/s-0039-3402757

  25. [26]

    Woo, Meghan and Bacon, Olivia , month = mar, year =. Alarm. Making

  26. [27]

    Journal of electrocardiology , author =

    A. Journal of electrocardiology , author =. 2018 , pages =. doi:10.1016/j.jelectrocard.2018.07.024 , abstract =

    work page doi:10.1016/j.jelectrocard.2018.07.024 2018

  27. [28]

    Lan, K. K. Gordon and DeMets, David L. , year =. Discrete. Biometrika , publisher =. doi:10.2307/2336502 , abstract =

    work page doi:10.2307/2336502

  28. [29]

    and Fleming, Thomas R

    O'Brien, Peter C. and Fleming, Thomas R. , year =. A. Biometrics , publisher =. doi:10.2307/2530245 , abstract =

    doi:10.2307/2530245

  29. [30]

    Pubblicazioni del R Istituto Superiore di Scienze Economiche e Commericiali di Firenze , author =

    Teoria statistica delle classi e calcolo delle probabilita , volume =. Pubblicazioni del R Istituto Superiore di Scienze Economiche e Commericiali di Firenze , author =. 1936 , pages =

    1936

  30. [31]

    Westfall, P. H. and Young, S. S. and Wright, S. Paul , year =. On. Biometrics , publisher =. doi:10.2307/2532216 , number =

    work page doi:10.2307/2532216

  31. [32]

    Knowledge and Information Systems , author =

    A survey of methods for time series change point detection , volume =. Knowledge and Information Systems , author =. 2017 , keywords =. doi:10.1007/s10115-016-0987-z , abstract =

    doi:10.1007/s10115-016-0987-z 2017

  32. [33]

    Journal of the American Statistical Association , author =

    Rectangular. Journal of the American Statistical Association , author =. 1967 , pages =. doi:10.1080/01621459.1967.10482935 , language =

    arXiv 1967

  33. [34]

    and Barnett, Ian J

    Martin, Melissa Lynne and Brook, Juliette and Rush, Sage and Satterthwaite, Theodore D. and Barnett, Ian J. , month = jan, year =. Variance component score test for multivariate change point detection with applications to mobile health , url =. doi:10.48550/arXiv.2601.22147 , abstract =

    work page doi:10.48550/arxiv.2601.22147

  34. [35]

    JMIR mHealth and uHealth , author =

    Correlations. JMIR mHealth and uHealth , author =. 2018 , pages =. doi:10.2196/mhealth.9691 , abstract =

    work page doi:10.2196/mhealth.9691 2018

  35. [36]

    JMIR mHealth and uHealth , author =

    Predicting. JMIR mHealth and uHealth , author =. 2021 , pages =. doi:10.2196/26540 , abstract =

    work page doi:10.2196/26540 2021

  36. [37]

    JAMA Network Open , author =

    Passive. JAMA Network Open , author =. 2025 , pages =. doi:10.1001/jamanetworkopen.2025.19047 , abstract =

    arXiv 2025

  37. [38]

    Depression and anxiety , author =

    The accuracy of passive phone sensors in predicting daily mood , volume =. Depression and anxiety , author =. 2019 , pages =. doi:10.1002/da.22822 , abstract =

    work page doi:10.1002/da.22822 2019

  38. [39]

    Biometrika , author =

    Variance component testing in generalised linear models with random effects , volume =. Biometrika , author =. 1997 , pages =. doi:10.1093/biomet/84.2.309 , abstract =

    work page doi:10.1093/biomet/84.2.309 1997

  39. [41]

    Biometrics , author =

    The. Biometrics , author =. 2003 , pages =. doi:10.1111/1541-0420.00032 , abstract =

    arXiv 2003

  40. [42]

    Journal of the American Academy of Child & Adolescent Psychiatry , author =

    The. Journal of the American Academy of Child & Adolescent Psychiatry , author =. 1997 , keywords =. doi:10.1097/00004583-199704000-00018 , abstract =

    work page doi:10.1097/00004583-199704000-00018 1997

  41. [47]

    Personality and Individual Differences , author =

    Reliability and validity of a brief measure of sensation seeking , volume =. Personality and Individual Differences , author =. 2002 , keywords =. doi:10.1016/S0191-8869(01)00032-0 , abstract =

    work page doi:10.1016/s0191-8869(01)00032-0 2002

  42. [48]

    Neuropsychopharmacology: Official Publication of the American College of Neuropsychopharmacology , author =

    Relapse prediction in schizophrenia through digital phenotyping: a pilot study , volume =. Neuropsychopharmacology: Official Publication of the American College of Neuropsychopharmacology , author =. 2018 , keywords =. doi:10.1038/s41386-018-0030-z , abstract =

    doi:10.1038/s41386-018-0030-z 2018

  43. [49]

    Current Psychiatry Reports , author =

    Realizing the. Current Psychiatry Reports , author =. 2015 , keywords =. doi:10.1007/s11920-015-0602-0 , abstract =

    work page doi:10.1007/s11920-015-0602-0 2015

  44. [55]

    Technometrics , author =

    Multivariate. Technometrics , author =. 1988 , pages =. doi:10.1080/00401706.1988.10488402 , language =

    arXiv 1988

  45. [56]

    Xia, C. H. and Barnett, I. and Tapera, Tinashe M. and Adebimpe, A. and Baker, J. T. and Bassett, D. S. and Brotman, M. A. and Calkins, M. E. and Cui, Z. and Leibenluft, E. and Linguiti, S. and Lydon-Staley, D. M. and Martin, M. L. and Moore, T. M. and Murtha, K. and Piiwaa, K. and Pines, A. and Roalf, D. R. and Rush-Goebel, S. and Wolf, D. H. and Ungar, L...

  46. [57]

    Harvard Review of Psychiatry , author =

    Methodology and. Harvard Review of Psychiatry , author =. 2017 , pages =. doi:10.1097/HRP.0000000000000133 , abstract =

    work page doi:10.1097/hrp.0000000000000133 2017

  47. [58]

    Our World in Data , author =

    Mental. Our World in Data , author =

  48. [59]

    Biostatistics , author =

    Inferring mobility measures from. Biostatistics , author =. 2020 , pages =. doi:10.1093/biostatistics/kxy059 , abstract =

    work page doi:10.1093/biostatistics/kxy059 2020

  49. [61]

    Journal of multivariate analysis , author =

    High. Journal of multivariate analysis , author =. 2022 , pages =. doi:10.1016/j.jmva.2021.104833 , abstract =

    arXiv 2022

  50. [63]

    The American Journal of Human Genetics , author =

    Haplotype-. The American Journal of Human Genetics , author =. 2007 , pages =. doi:10.1086/521558 , language =

    work page doi:10.1086/521558 2007

  51. [64]

    Applied Intelligence , author =

    Finder:. Applied Intelligence , author =. 2022 , keywords =. doi:10.1007/s10489-021-02532-x , abstract =

    work page doi:10.1007/s10489-021-02532-x 2022

  52. [65]

    JMIR mental health , author =

    Detection of. JMIR mental health , author =. 2020 , keywords =. doi:10.2196/21378 , abstract =

    work page doi:10.2196/21378 2020

  53. [67]

    Computers & Industrial Engineering , author =

    Change-point detection for shifts in control charts using fuzzy shift change-point algorithms , volume =. Computers & Industrial Engineering , author =. 2016 , keywords =. doi:10.1016/j.cie.2015.12.002 , abstract =

    work page doi:10.1016/j.cie.2015.12.002 2016

  54. [68]

    Personality and Individual Differences , author =

    Brief assessment of schizotypy:. Personality and Individual Differences , author =. 2011 , keywords =. doi:10.1016/j.paid.2011.07.027 , abstract =

    work page doi:10.1016/j.paid.2011.07.027 2011

  55. [69]

    Financial Internet Quarterly , author =

    Bayesian. Financial Internet Quarterly , author =. 2022 , pages =. doi:10.2478/fiqf-2021-0025 , abstract =

    work page doi:10.2478/fiqf-2021-0025 2022

  56. [73]

    Archives of General Psychiatry , author =

    An. Archives of General Psychiatry , author =. 1961 , pages =. doi:10.1001/archpsyc.1961.01710120031004 , abstract =

    arXiv 1961

  57. [77]

    , year =

    Hotelling, H. , year =. Multivariate. Techniques of Statistical Analysis , publisher =

  58. [78]

    IEEE Access , author =

    Multi-. IEEE Access , author =. 2020 , pages =. doi:10.1109/ACCESS.2020.2984571 , abstract =

    arXiv 2020

  59. [79]

    Journal of Applied Statistics , author =

    Modeling the effects of multiple exposures with unknown group memberships: a. Journal of Applied Statistics , author =. doi:10.1080/02664763.2020.1843611 , abstract =

    arXiv 2020

  60. [88]

    Academic radiology , author =

    Multi-. Academic radiology , author =. 2013 , pages =. doi:10.1016/j.acra.2013.09.010 , abstract =

    work page doi:10.1016/j.acra.2013.09.010 2013

  61. [89]

    2012 , pages =

    Radiology , author =. 2012 , pages =. doi:10.1148/radiol.12120208 , language =

    work page doi:10.1148/radiol.12120208 2012

  62. [90]

    2018 , pages =

    Journal of Neuroimaging , author =. 2018 , pages =. doi:10.1111/jon.12506 , abstract =

    work page doi:10.1111/jon.12506 2018

  63. [93]

    Brain , author =

    The pathogenesis of lesions and normal-appearing white matter changes in multiple sclerosis:. Brain , author =. 2000 , pages =. doi:10.1093/brain/123.8.1667 , language =

    work page doi:10.1093/brain/123.8.1667 2000

  64. [94]

    IEEE transactions on pattern analysis and machine intelligence , author =

    Multi-. IEEE transactions on pattern analysis and machine intelligence , author =. 2013 , pages =. doi:10.1109/TPAMI.2012.143 , abstract =

    work page doi:10.1109/tpami.2012.143 2013

  65. [95]

    NeuroImage : Clinical , author =

    Statistical normalization techniques for magnetic resonance imaging , volume =. NeuroImage : Clinical , author =. 2014 , pages =. doi:10.1016/j.nicl.2014.08.008 , abstract =

    work page doi:10.1016/j.nicl.2014.08.008 2014

  66. [97]

    2014 , pages =

    BioMed Research International , author =. 2014 , pages =. doi:10.1155/2014/312142 , abstract =

    doi:10.1155/2014/312142 2014

  67. [100]

    B, A. B. and Nick, T. and Hans, J. , pages =. Advanced

  68. [101]

    New England Journal of Medicine , author =

    Multiple. New England Journal of Medicine , author =. 2018 , pages =. doi:10.1056/NEJMra1401483 , language =

    work page doi:10.1056/nejmra1401483 2018

  69. [102]

    Current Neurology and Neuroscience Reports , author =

    Misdiagnosis of. Current Neurology and Neuroscience Reports , author =. 2013 , pages =. doi:10.1007/s11910-013-0403-y , abstract =

    work page doi:10.1007/s11910-013-0403-y 2013

  70. [103]

    Nature Reviews Neurology , author =

    The tension between early diagnosis and misdiagnosis of multiple sclerosis , volume =. Nature Reviews Neurology , author =. 2017 , pages =. doi:10.1038/nrneurol.2017.106 , abstract =

    work page doi:10.1038/nrneurol.2017.106 2017

  71. [104]

    Journal of the Neurological Sciences , author =

    The onset and progression of the lesion in multiple sclerosis , volume =. Journal of the Neurological Sciences , author =. 1975 , pages =. doi:10.1016/0022-510X(75)90138-0 , abstract =

    work page doi:10.1016/0022-510x(75)90138-0 1975

  72. [107]

    2017 , keywords =

    European Radiology , author =. 2017 , keywords =. doi:10.1007/s00330-017-4822-z , abstract =

    work page doi:10.1007/s00330-017-4822-z 2017

  73. [108]

    central vein sign

    Preventing multiple sclerosis misdiagnosis using the “central vein sign”:. Multiple Sclerosis and Related Disorders , author =. 2021 , keywords =. doi:10.1016/j.msard.2020.102671 , abstract =

    arXiv 2021

  74. [111]

    The R journal , author =

    fslr:. The R journal , author =. 2015 , pages =

    2015

  75. [115]

    and Matt, H

    Kyle, W. and Matt, H. , year =. tidycensus:

  76. [116]

    Neurology - Neuroimmunology Neuroinflammation , author =

    Neuromyelitis optica spectrum disorders:. Neurology - Neuroimmunology Neuroinflammation , author =. doi:10.1212/NXI.0000000000000225 , abstract =

    work page doi:10.1212/nxi.0000000000000225

  77. [117]

    Neuroepidemiology , author =

    Multiple. Neuroepidemiology , author =. 2018 , pages =. doi:10.1159/000484213 , abstract =

    work page doi:10.1159/000484213 2018

  78. [118]

    JAMA Neurology , author =

    Health. JAMA Neurology , author =. 2021 , pages =. doi:10.1001/jamaneurol.2021.3416 , abstract =

    arXiv 2021

  79. [121]

    Journal of Neurology , author =

    Early use of high-efficacy disease‑modifying therapies makes the difference in people with multiple sclerosis: an expert opinion , volume =. Journal of Neurology , author =. 2022 , keywords =. doi:10.1007/s00415-022-11193-w , abstract =

    work page doi:10.1007/s00415-022-11193-w 2022

  80. [124]

    , year =

    Vivitrol, V. , year =

Showing first 80 references.