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

Issues in biomedical statistics: comparing means by computer-intensive tests

J Ludbrook1

  • 1Cardiovascular Research Laboratory, University of Melbourne Department of Surgery, Royal Melbourne Hospital, Parkville, Victoria, Australia.

The Australian and New Zealand Journal of Surgery
|November 1, 1995
PubMed
Summary
This summary is machine-generated.

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

Federated Learning Survival Model and Potential Radiotherapy Decision Support Impact Assessment for Non-small Cell Lung Cancer Using Real-World Data.

Clinical oncology (Royal College of Radiologists (Great Britain))·2024
Same author

The Problems of "Benign" Polypoid Lesions of the Stomach.

The Australian and New Zealand journal of surgery·2017
Same author

Comparison of telemetric and tail-cuff blood pressure monitoring in adrenocorticotrophic hormone-treated rats.

Clinical and experimental pharmacology & physiology·2001
Same author

Neural mechanisms in the cardiovascular responses to acute central hypovolaemia.

Clinical and experimental pharmacology & physiology·2001
Same author

Statistics in physiology and pharmacology: a slow and erratic learning curve.

Clinical and experimental pharmacology & physiology·2001
Same author

Computer-intensive statistical procedures.

Critical reviews in biochemistry and molecular biology·2000
Same journal

The Experimental Application of Microsurgical Techniques to Internal Mammary to Coronary Artery Anastomosis.

The Australian and New Zealand journal of surgery·2018
Same journal

The Use of Trimethoprim-Sulphamethoxazole in the Treatment of Complicated Urinary Tract Infection.

The Australian and New Zealand journal of surgery·2018
Same journal

The Bairnsdale Ulcer.

The Australian and New Zealand journal of surgery·2018
Same journal

Aorto-Caval Fistula: Successful Management of Two Cases.

The Australian and New Zealand journal of surgery·2018
Same journal

Subdural Empyema.

The Australian and New Zealand journal of surgery·2018
Same journal

Massive Ascites Due to Pancreatic Stones.

The Australian and New Zealand journal of surgery·2018
See all related articles

Computer-intensive methods like permutation tests and bootstrapping offer alternatives to classical statistical tests for comparing means. Permutation tests suit small, randomized datasets, while bootstrapping is better for large, randomly sampled populations.

Area of Science:

  • Statistics
  • Data Analysis

Background:

  • Classical statistical tests (t-tests, F-tests) for comparing means rely on specific assumptions.
  • Computer-intensive methods provide alternatives when these assumptions are not met or for enhanced analytical power.

Purpose of the Study:

  • To review and compare computer-intensive statistical procedures as alternatives to classical tests for distinguishing means.
  • To evaluate the applicability and advantages of permutation procedures and bootstrap methods.

Main Methods:

  • Review of permutation procedures, emphasizing their use with small, randomized experimental data.
  • Review of non-parametric bootstrap procedures, highlighting their suitability for large, randomly sampled populations.
  • Comparison of these methods with classical t and F tests under different inference models (randomization vs. population).

Related Experiment Videos

Main Results:

  • Permutation tests are advantageous for small datasets under a randomization model, requiring no assumptions beyond randomization.
  • Bootstrap procedures are suited for large datasets under a population model, without prior distributional assumptions.
  • Classical tests may be outperformed by permutation tests (randomization) or bootstrap techniques (random sampling).

Conclusions:

  • Permutation tests are recommended to replace classical tests when randomization is employed.
  • Non-parametric bootstrapping shows potential for superiority in constructing population confidence intervals or hypothesis testing with random sampling.
  • Further research is needed to establish the accuracy of bootstrap methods, especially for hypothesis testing with small samples, and broader software support is required.