What question did this study set out to answer?

To determine how violations of timing assumptions and data-driven interim scheduling inflate Type I error rates across frequentist and Bayesian sequential monitoring methods.

December 1, 1992Biometrics

Effects of Assumption Violations on Type I Error Rate in Group Sequential Monitoring

View Full Paper

Ask AI

Bookmark

Authors

MPMichael A. ProschanNational Institute of Allergy and Infectious Diseases DFDean FollmannNational Institutes of Health MWMyron A. WaclawiwGeneral / Preventive / Lipids

Discussion

Member takes

Implication

Methodological evaluation demonstrates Type I error inflation from assumption violations across sequential monitoring designs, highlighting the risk of data-driven look times.

Key Points

To determine how violations of timing assumptions and data-driven interim scheduling inflate Type I error rates across frequentist and Bayesian sequential monitoring methods.
Assessed the effect of unequal interim look spacings on group sequential monitoring methods that assume equal time intervals.
Evaluated Type I error inflation occurring when future look times are adaptively chosen based on observed data trends.
Analyzed the frequentist Type I error properties of a Bayesian sequential monitoring approach.
Unequal time spacings inflate the Type I error probability in methods requiring fixed, equally spaced looks.
Choosing future interim inspection times based on observed data trajectories leads to substantial Type I error rate inflation.
Bayesian sequential monitoring methods exhibit inflated frequentist false-positive rates under adaptive or unprespecified look schedules.

Cite This Study

Proschan et al. (1992) studied this question.

synapsesocial.com/papers/6a16e9cab13aec50ea6ba1e6 https://doi.org/10.2307/2532704

View Full Paper

Ask AI

Bookmark

Implication

Methodological evaluation demonstrates Type I error inflation from assumption violations across sequential monitoring designs, highlighting the risk of data-driven look times.

Effects of Assumption Violations on Type I Error Rate in Group Sequential Monitoring

Discussion

Member takes

Implication

Key Points

Cite This Study

Effects of Assumption Violations on Type I Error Rate in Group Sequential Monitoring

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider