Early stopping to accept H(o) based on conditional power

Approximations and comparisons

Rebecca Betensky

Research output: Contribution to journalArticle

Abstract

It is intuitively appealing to clinicians to stop a trial early to accept the null hypothesis H0 if it appears that this will be the likely outcome at the planned end of the trial. We consider procedures that calculate at each time point the conditional probability of rejecting H0 at the end of the trial given the current data and some value of the parameter of interest. Lan, Simon, and Halperin (1982, Communications in Statistics C1, 207-219) calculate this probability under the design alternative, and Pepe and Anderson (1992, Applied Statistics 41, 181-190) use an alternative based solely on the current data. We investigate a modification to Pepe and Anderson's (1992) procedure that has a more satisfying interpretation. We define all of these procedures as formal sequential tests with lower stopping boundaries and study them in this context. This facilitates an improved understanding of the interplay of parameters by introducing visual displays, and it leads to an approximation for power by treating it as a boundary crossing probability. We use these tools to compare the performances of the different designs under a variety of parameter configurations.

Original languageEnglish (US)
Pages (from-to)794-806
Number of pages13
JournalBiometrics
Volume53
Issue number3
DOIs
StatePublished - Sep 1 1997

Fingerprint

Conditional Power
Early Stopping
Boundary Crossing Probability
statistics
Approximation
Statistics
Sequential Test
Calculate
Alternatives
Conditional probability
Null hypothesis
Display
Likely
Display devices
Configuration
Communication
testing
Design

Keywords

  • Boundary crossing
  • Sequential design
  • Stochastic curtailment

ASJC Scopus subject areas

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

Cite this

Early stopping to accept H(o) based on conditional power : Approximations and comparisons. / Betensky, Rebecca.

In: Biometrics, Vol. 53, No. 3, 01.09.1997, p. 794-806.

Research output: Contribution to journalArticle

@article{fbc8a54a2ca443a0b6e4d26167dfcfba,
title = "Early stopping to accept H(o) based on conditional power: Approximations and comparisons",
abstract = "It is intuitively appealing to clinicians to stop a trial early to accept the null hypothesis H0 if it appears that this will be the likely outcome at the planned end of the trial. We consider procedures that calculate at each time point the conditional probability of rejecting H0 at the end of the trial given the current data and some value of the parameter of interest. Lan, Simon, and Halperin (1982, Communications in Statistics C1, 207-219) calculate this probability under the design alternative, and Pepe and Anderson (1992, Applied Statistics 41, 181-190) use an alternative based solely on the current data. We investigate a modification to Pepe and Anderson's (1992) procedure that has a more satisfying interpretation. We define all of these procedures as formal sequential tests with lower stopping boundaries and study them in this context. This facilitates an improved understanding of the interplay of parameters by introducing visual displays, and it leads to an approximation for power by treating it as a boundary crossing probability. We use these tools to compare the performances of the different designs under a variety of parameter configurations.",
keywords = "Boundary crossing, Sequential design, Stochastic curtailment",
author = "Rebecca Betensky",
year = "1997",
month = "9",
day = "1",
doi = "10.2307/2533543",
language = "English (US)",
volume = "53",
pages = "794--806",
journal = "Biometrics",
issn = "0006-341X",
publisher = "Wiley-Blackwell",
number = "3",

}

TY - JOUR

T1 - Early stopping to accept H(o) based on conditional power

T2 - Approximations and comparisons

AU - Betensky, Rebecca

PY - 1997/9/1

Y1 - 1997/9/1

N2 - It is intuitively appealing to clinicians to stop a trial early to accept the null hypothesis H0 if it appears that this will be the likely outcome at the planned end of the trial. We consider procedures that calculate at each time point the conditional probability of rejecting H0 at the end of the trial given the current data and some value of the parameter of interest. Lan, Simon, and Halperin (1982, Communications in Statistics C1, 207-219) calculate this probability under the design alternative, and Pepe and Anderson (1992, Applied Statistics 41, 181-190) use an alternative based solely on the current data. We investigate a modification to Pepe and Anderson's (1992) procedure that has a more satisfying interpretation. We define all of these procedures as formal sequential tests with lower stopping boundaries and study them in this context. This facilitates an improved understanding of the interplay of parameters by introducing visual displays, and it leads to an approximation for power by treating it as a boundary crossing probability. We use these tools to compare the performances of the different designs under a variety of parameter configurations.

AB - It is intuitively appealing to clinicians to stop a trial early to accept the null hypothesis H0 if it appears that this will be the likely outcome at the planned end of the trial. We consider procedures that calculate at each time point the conditional probability of rejecting H0 at the end of the trial given the current data and some value of the parameter of interest. Lan, Simon, and Halperin (1982, Communications in Statistics C1, 207-219) calculate this probability under the design alternative, and Pepe and Anderson (1992, Applied Statistics 41, 181-190) use an alternative based solely on the current data. We investigate a modification to Pepe and Anderson's (1992) procedure that has a more satisfying interpretation. We define all of these procedures as formal sequential tests with lower stopping boundaries and study them in this context. This facilitates an improved understanding of the interplay of parameters by introducing visual displays, and it leads to an approximation for power by treating it as a boundary crossing probability. We use these tools to compare the performances of the different designs under a variety of parameter configurations.

KW - Boundary crossing

KW - Sequential design

KW - Stochastic curtailment

UR - http://www.scopus.com/inward/record.url?scp=0030821530&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0030821530&partnerID=8YFLogxK

U2 - 10.2307/2533543

DO - 10.2307/2533543

M3 - Article

VL - 53

SP - 794

EP - 806

JO - Biometrics

JF - Biometrics

SN - 0006-341X

IS - 3

ER -