A Broad Symmetry Criterion for Nonparametric Validity of Parametrically Based Tests in Randomized Trials

Research output: Contribution to journalArticle

Abstract

Pilot phases of a randomized clinical trial often suggest that a parametric model may be an accurate description of the trial's longitudinal trajectories. However, parametric models are often not used for fear that they may invalidate tests of null hypotheses of equality between the experimental groups. Existing work has shown that when, for some types of data, certain parametric models are used, the validity for testing the null is preserved even if the parametric models are incorrect. Here, we provide a broader and easier to check characterization of parametric models that can be used to (i) preserve nonparametric validity of testing the null hypothesis, i.e., even when the models are incorrect, and (ii) increase power compared to the non- or semiparametric bounds when the models are close to correct. We demonstrate our results in a clinical trial of depression in Alzheimer's patients.

Original languageEnglish (US)
Pages (from-to)85-91
Number of pages7
JournalBiometrics
Volume68
Issue number1
DOIs
StatePublished - Mar 2012

Fingerprint

Randomized Trial
Parametric Model
Fear
Randomized Controlled Trials
Clinical Trials
Depression
Symmetry
Null hypothesis
testing
Randomized Clinical Trial
Testing
Data Model
Null
randomized clinical trials
Equality
Power (Psychology)
fearfulness
Trajectory
trajectories
preserves

Keywords

  • Causal inference
  • Hypothesis test
  • Randomized clinical trial
  • Robustness
  • Superefficiency

ASJC Scopus subject areas

  • Applied Mathematics
  • Statistics and Probability
  • Agricultural and Biological Sciences(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Immunology and Microbiology(all)
  • Medicine(all)

Cite this

A Broad Symmetry Criterion for Nonparametric Validity of Parametrically Based Tests in Randomized Trials. / Shinohara, Russell T.; Frangakis, Constantine; Lyketsos, Constantine G.

In: Biometrics, Vol. 68, No. 1, 03.2012, p. 85-91.

Research output: Contribution to journalArticle

@article{5630b0eb2e7442d885a044f4df576078,
title = "A Broad Symmetry Criterion for Nonparametric Validity of Parametrically Based Tests in Randomized Trials",
abstract = "Pilot phases of a randomized clinical trial often suggest that a parametric model may be an accurate description of the trial's longitudinal trajectories. However, parametric models are often not used for fear that they may invalidate tests of null hypotheses of equality between the experimental groups. Existing work has shown that when, for some types of data, certain parametric models are used, the validity for testing the null is preserved even if the parametric models are incorrect. Here, we provide a broader and easier to check characterization of parametric models that can be used to (i) preserve nonparametric validity of testing the null hypothesis, i.e., even when the models are incorrect, and (ii) increase power compared to the non- or semiparametric bounds when the models are close to correct. We demonstrate our results in a clinical trial of depression in Alzheimer's patients.",
keywords = "Causal inference, Hypothesis test, Randomized clinical trial, Robustness, Superefficiency",
author = "Shinohara, {Russell T.} and Constantine Frangakis and Lyketsos, {Constantine G}",
year = "2012",
month = "3",
doi = "10.1111/j.1541-0420.2011.01642.x",
language = "English (US)",
volume = "68",
pages = "85--91",
journal = "Biometrics",
issn = "0006-341X",
publisher = "Wiley-Blackwell",
number = "1",

}

TY - JOUR

T1 - A Broad Symmetry Criterion for Nonparametric Validity of Parametrically Based Tests in Randomized Trials

AU - Shinohara, Russell T.

AU - Frangakis, Constantine

AU - Lyketsos, Constantine G

PY - 2012/3

Y1 - 2012/3

N2 - Pilot phases of a randomized clinical trial often suggest that a parametric model may be an accurate description of the trial's longitudinal trajectories. However, parametric models are often not used for fear that they may invalidate tests of null hypotheses of equality between the experimental groups. Existing work has shown that when, for some types of data, certain parametric models are used, the validity for testing the null is preserved even if the parametric models are incorrect. Here, we provide a broader and easier to check characterization of parametric models that can be used to (i) preserve nonparametric validity of testing the null hypothesis, i.e., even when the models are incorrect, and (ii) increase power compared to the non- or semiparametric bounds when the models are close to correct. We demonstrate our results in a clinical trial of depression in Alzheimer's patients.

AB - Pilot phases of a randomized clinical trial often suggest that a parametric model may be an accurate description of the trial's longitudinal trajectories. However, parametric models are often not used for fear that they may invalidate tests of null hypotheses of equality between the experimental groups. Existing work has shown that when, for some types of data, certain parametric models are used, the validity for testing the null is preserved even if the parametric models are incorrect. Here, we provide a broader and easier to check characterization of parametric models that can be used to (i) preserve nonparametric validity of testing the null hypothesis, i.e., even when the models are incorrect, and (ii) increase power compared to the non- or semiparametric bounds when the models are close to correct. We demonstrate our results in a clinical trial of depression in Alzheimer's patients.

KW - Causal inference

KW - Hypothesis test

KW - Randomized clinical trial

KW - Robustness

KW - Superefficiency

UR - http://www.scopus.com/inward/record.url?scp=84858864001&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84858864001&partnerID=8YFLogxK

U2 - 10.1111/j.1541-0420.2011.01642.x

DO - 10.1111/j.1541-0420.2011.01642.x

M3 - Article

C2 - 21762116

AN - SCOPUS:84858864001

VL - 68

SP - 85

EP - 91

JO - Biometrics

JF - Biometrics

SN - 0006-341X

IS - 1

ER -