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

The Problem with "Magnitude-based Inference".

Kristin L Sainani1

  • 1Division of Epidemiology, Department of Health Research and Policy, Stanford University, Stanford, CA.

Medicine and Science in Sports and Exercise
|April 24, 2018
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

Data and code (Un)availability in sports meta-analysis studies.

Annals of epidemiology·2026
Same author

Dual-energy X-ray absorptiometry per cent fat Z-score as a predictor of menstrual status in adolescent and young adult female athletes.

British journal of sports medicine·2026
Same author

A Descriptive Analysis of the Seasonal Patterns of Bone Stress Injury Incidence in Division I Collegiate Distance Runners.

The American journal of sports medicine·2025
Same author

The association between overuse and musculoskeletal injuries and the female athlete triad in Division I collegiate athletes.

PM & R : the journal of injury, function, and rehabilitation·2024
Same author

Response to "Homeopathy: A null field or effective psychotherapy?"

Journal of clinical epidemiology·2024
Same author

Exploratory analyses: How to meaningfully interpret and report them.

PM & R : the journal of injury, function, and rehabilitation·2023
Same journal

The Impact of Cardiorespiratory Fitness on Cytotoxic T Cell Metabolism and Function.

Medicine and science in sports and exercise·2026
Same journal

Female Athletes Through the Lifespan: Clinical Considerations and a Call for Comprehensive Sports Medicine Healthcare.

Medicine and science in sports and exercise·2026
Same journal

Artificial Intelligence in Exercise Science and Sports Medicine.

Medicine and science in sports and exercise·2026
Same journal

Non-Vigorous Physical Activity Associated with Reduced Hospitalization Risk with or without Diabetes or Peripheral Artery Disease: Study of Latinos.

Medicine and science in sports and exercise·2026
Same journal

One Step Further in Resistance Training Prescription: Do Recent Updates Require Reconsideration?

Medicine and science in sports and exercise·2026
Same journal

Response.

Medicine and science in sports and exercise·2026
See all related articles

Magnitude-based inference (MBI) does not offer superior error rates compared to standard hypothesis testing. MBI inflates type I errors, making it unreliable for sports science research.

Area of Science:

  • Sports Science
  • Statistical Methodology
  • Hypothesis Testing

Background:

  • Magnitude-based inference (MBI) is increasingly used in sports science.
  • Statisticians have raised concerns regarding MBI's statistical validity.
  • Proponents claim MBI has superior error rates over traditional methods.

Purpose of the Study:

  • To reanalyze and evaluate the claimed superior error rates of MBI.
  • To compare MBI's type I and type II error rates against standard null hypothesis testing.

Main Methods:

  • Simulations were conducted using MBI proponents' code.
  • Error rates were estimated across various effect sizes, sample sizes, and smallest important effects.
  • Mathematical analysis and corrected error rate definitions were employed.

Related Experiment Videos

Main Results:

  • MBI does not demonstrate superior type I and type II error rates.
  • A trade-off exists between type I and type II errors in MBI.
  • At optimal sample sizes, MBI significantly inflates type I error rates (2-6x standard testing).

Conclusions:

  • MBI exhibits unpredictable and often unacceptably high type I error rates.
  • MBI's error rates are dependent on sample size and the choice of smallest important effect.
  • The use of magnitude-based inference is not recommended in research.