Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Two-stage adaptive designs with correlated test statistics.

Gerhard Hommel1, Verena Lindig, Andreas Faldum

  • 1Institute for Medical Biometry, Epidemiology and Informatics, University of Mainz, Mainz, Germany. hommel@imbei.uni-mainz.de

Journal of Biopharmaceutical Statistics
|July 19, 2005
PubMed
Summary

This study examines type I error in two-stage adaptive designs when data stages are dependent. Results show that statistical decisions can be conservative or anticonservative based on design parameters and correlation.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Neuropsychological late effects in young children with medulloblastoma: A longitudinal analysis after 5 and 15 years from the HIT-SKK-'87/'92 studies.

Neuro-oncology practice·2026
Same author

The HIT Network for Children and Adolescents With CNS Tumors Facilitates Improvements of Diagnostic Assessments, Multimodal Treatments, Individual Counseling, and Research in Germany, Austria, and Switzerland.

Pediatric blood & cancer·2026
Same author

The importance of comprehensive geriatric assessment in predicting the outcome of patients with proximal humerus fractures.

Aging clinical and experimental research·2026
Same author

Geriatric-Typical Characteristic Complexes Predict Short-Term Outcome of Proximal Humeral Fractures in Geriatric Patients.

Clinical interventions in aging·2026
Same author

Characterizing Overall Survival of Patients with Acute Myeloid Leukemia: A Competing Risk Analysis of SEER Data Covering 46 Years.

Cancers·2025
Same author

Centenarians with proximal humeral fracture.

BMC geriatrics·2025

Area of Science:

  • Biostatistics
  • Clinical Trial Design

Background:

  • Adaptive sequential designs typically assume independent data units per stage.
  • Dependency between data stages can occur, violating standard assumptions.
  • This dependency impacts the validity of statistical tests and p-values.

Purpose of the Study:

  • Investigate type I error rates in two-stage adaptive designs with dependent test statistics.
  • Analyze the impact of bivariate normality and correlation on design performance.
  • Evaluate the behavior of adaptive designs under relaxed independence assumptions.

Main Methods:

  • Assumed bivariate normality for stage test statistics.
  • Employed analytical methods under a constant conditional error function.
  • Calculated type I error rates for various design parameters and correlation scenarios.

Related Experiment Videos

Main Results:

  • Type I error rates can be conservative or anticonservative.
  • The direction of error rate deviation depends on design parameters.
  • The sign of the correlation coefficient between stage statistics is a critical factor.

Conclusions:

  • The assumption of independent data units is crucial for standard adaptive designs.
  • Dependent stages necessitate careful consideration of statistical properties.
  • The findings offer insights into robust adaptive design strategies and comparisons with existing methods like Bauer-Köhne.