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

Is statistical significance testing useful in interpreting data?

D A Savitz1

  • 1Department of Epidemiology, School of Public Health, University of North Carolina, Chapel Hill 27599-7400.

Reproductive Toxicology (Elmsford, N.Y.)
|January 1, 1993
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

Pregnancy and the risk of severe coronavirus disease 2019 infection: methodological challenges and research recommendations.

BJOG : an international journal of obstetrics and gynaecology·2021
Same author

Cardiovascular severe maternal morbidity in pregnant and postpartum women: development and internal validation of risk prediction models.

BJOG : an international journal of obstetrics and gynaecology·2020
Same author

Only some questions of cause and effect can be evaluated in highly selected populations.

BJOG : an international journal of obstetrics and gynaecology·2017
Same author

Maternal influenza and birth outcomes: systematic review of comparative studies.

BJOG : an international journal of obstetrics and gynaecology·2016
Same author

The effect of labour induction on the risk of caesarean delivery: using propensity scores to control confounding by indication.

BJOG : an international journal of obstetrics and gynaecology·2015
Same author

Ethnic enclaves and gestational diabetes among immigrant women in New York City.

Social science & medicine (1982)·2014
Same journal

Retraction notice to "Trehalose restores functional autophagy suppressed by high glucose" [Reprod. Toxicol. 85 (2019) 51-58].

Reproductive toxicology (Elmsford, N.Y.)·2026
Same journal

Exposure to toxic metals/metalloids in the environment and in vitro fertilization outcomes in a population group from Romania.

Reproductive toxicology (Elmsford, N.Y.)·2026
Same journal

Gadolinium induces Sertoli cell damage: cellular and molecular mechanisms.

Reproductive toxicology (Elmsford, N.Y.)·2026
Same journal

Influence of radiofrequency electromagnetic radiation on spermatogenesis and sperm function in rodent models: A systematic review.

Reproductive toxicology (Elmsford, N.Y.)·2026
Same journal

Microplastics as trojan horses: A new perspective on bisphenol toxicity in male infertility and assisted reproduction.

Reproductive toxicology (Elmsford, N.Y.)·2026
Same journal

Effects of prenatal pyrethroid pesticides exposure on neurodevelopment of 3-year-old children: A birth cohort study in rural Southwest China.

Reproductive toxicology (Elmsford, N.Y.)·2026
See all related articles

Statistical significance testing and P values in biomedical research may be detrimental. Confidence intervals offer an alternative, promoting a broader consideration of research accuracy beyond random error.

Area of Science:

  • Biostatistics
  • Epidemiology
  • Medical Research Methodology

Background:

  • P values and statistical significance testing are standard in biomedical research.
  • Their limitations and potential drawbacks warrant reconsideration, especially in observational epidemiology.

Purpose of the Study:

  • To critically evaluate the utility of P values and statistical significance testing.
  • To propose confidence intervals as a more informative alternative for research evaluation.

Main Methods:

  • Review of the foundational principles of statistical significance testing and null hypothesis.
  • Comparison of information conveyed by P values versus confidence intervals.

Main Results:

  • The reliance on null hypothesis and dichotomized results (positive/negative) can hinder proper research design and evaluation.

Related Experiment Videos

  • Confidence intervals provide information on random variation with fewer limitations than P values.

    • Conclusions:

    • Eliminating statistical significance testing as a strict decision rule is recommended.
    • This shift would encourage a more comprehensive assessment of methodological factors influencing research accuracy, moving beyond solely random error.