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

Estimate of standard deviation for a log-transformed variable using arithmetic means and standard deviations.

Hui Quan1, Ji Zhang

  • 1Clinical Biostatistics and Research Decision Sciences, RY 34-A320, Merck Research Laboratories, P.O. Box 2000, Rahway, NJ 07065, USA. hui_quan@merck.com

Statistics in Medicine
|August 27, 2003
PubMed
Summary

This study provides methods to estimate the standard deviation of log-transformed variables when only untransformed data means and standard deviations are available. These techniques are crucial for statistical power calculations in research utilizing log-transformed data.

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

Epidemiological characteristics of serum small dense low-density lipoprotein cholesterol: From 2017 to 2019 in Chengdu adults, China.

Medicine·2026
Same author

Development and validation of a nomogram model for predicting negative/indeterminate HIV-1 antibody confirmation: a strategy to streamline the diagnostic pathway.

BMC medical informatics and decision making·2026
Same author

Haplotype-resolved long-read sequencing reveals parent-of-origin effects of tandem-repeat variation in autism spectrum disorder.

Science bulletin·2026
Same author

Analytical techniques for laboratory testing of HIV: a systematic review.

Journal of microbiological methods·2026
Same author

Sequential Parallel Comparison Design for Assessing Induction, Maintenance, Long-Term, and Other Treatment Effects on a Binary Endpoint.

Statistics in medicine·2026
Same author

Epidemiological characteristics and antibiotic resistance of pathogens in urinary tract infection in Chengdu 2019-2023, China.

BMC microbiology·2026

Area of Science:

  • Biostatistics
  • Statistical Modeling
  • Data Analysis

Background:

  • Log transformations are common in analyzing study variables.
  • Estimating standard deviation on the log scale is essential for power calculations.
  • Literature often provides only untransformed means and standard deviations.

Purpose of the Study:

  • To develop methods for estimating the standard deviation of log-transformed variables from untransformed data.
  • To provide methods for constructing confidence intervals for these estimates.
  • To address the estimation of standard deviation of change from baseline in the log scale.

Main Methods:

  • Derivation of formulas to estimate standard deviation of log-transformed data.
  • Development of methods for confidence interval construction.

Related Experiment Videos

  • Utilizing means and standard deviations of untransformed variables.

    • Main Results:

    • Presented methods for estimating standard deviation of log-transformed variables.
    • Provided methods for estimating standard deviation of change from baseline in log scale.
    • Simulations and examples demonstrated the performance of the estimation methods.

    Conclusions:

    • The proposed methods enable accurate estimation of log-transformed standard deviations from available data.
    • These methods are valuable for researchers conducting power analyses and interpreting results involving log-transformed data.
    • The study offers practical solutions for common data analysis challenges in scientific research.