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

On the use of a pilot sample for sample size determination

R H Browne1

  • 1Texas Scottish Rite Hospital for Children, Research Department, Dallas 75219, USA.

Statistics in Medicine
|September 15, 1995
PubMed
Summary

Estimating sample size for t-tests requires population standard deviation (sigma). Using pilot study data for sigma may lead to underpowered studies. A confidence limit approach ensures adequate sample size for planned statistical power.

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

Mechanical evaluation of a tapered thread-run-out half-pin designed for external skeletal fixation in small animals.

Veterinary and comparative orthopaedics and traumatology : V.C.O.T·2011
Same author

Mutations responsible for Larsen syndrome cluster in the FLNB protein.

Journal of medical genetics·2006
Same author

Adolescents undergoing surgery for idiopathic scoliosis: how physical and psychological characteristics relate to patient satisfaction with the cosmetic result.

Spine·2001
Same author

The sinister side of PA supply and demand.

JAAPA : official journal of the American Academy of Physician Assistants·2001
Same author

Treatment of femoral fractures in children by pediatric orthopedists: results of a 1998 survey.

Journal of pediatric orthopedics·2001
Same author

The Littler line method and the area under a Gaussian curve: a new method of assessing digital range of motion.

The Journal of hand surgery·2001

Area of Science:

  • Biostatistics
  • Statistical Inference

Background:

  • Accurate sample size calculation is crucial for achieving desired statistical power in hypothesis testing.
  • Estimating population standard deviation (sigma) is a key component of sample size determination for t-tests.
  • Using sample standard deviation from pilot studies can lead to underestimation of true variance and reduced study power.

Purpose of the Study:

  • To investigate methods for calculating sample sizes that reliably achieve planned statistical power for t-tests.
  • To address the issue of underpowered studies resulting from inaccurate standard deviation estimates.
  • To propose a robust approach for sample size computation using pilot data.

Main Methods:

  • Monte Carlo simulations were employed to evaluate different sample size estimation strategies.

Related Experiment Videos

  • The performance of using a point estimate of sigma versus an upper confidence limit was compared.
  • The study focused on the context of t-tests where population standard deviation is unknown.

    • Main Results:

    • Using the sample standard deviation from a pilot study often results in insufficient sample sizes and lower-than-planned power.
    • Employing a 100(1-gamma) per cent upper one-sided confidence limit for sigma provides a sample size that achieves planned power in at least 100(1-gamma) per cent of trials.
    • This confidence limit approach offers a more conservative and reliable method for sample size calculation.

    Conclusions:

    • A confidence limit approach for estimating population standard deviation is recommended for sample size calculations.
    • This method enhances the likelihood of achieving the intended statistical power in research studies.
    • Researchers should consider using upper confidence limits on pilot study standard deviations to avoid underpowered trials.