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

Using the bootstrap to improve estimation and confidence intervals for regression coefficients selected using

Peter C Austin1

  • 1Institute for Clinical Evaluative Sciences, Toronto, Ont., Canada. peter.austin@ices.on.ca

Statistics in Medicine
|October 18, 2007
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

Ranking hospital performance based on individual indicators: can we increase reliability by creating composite indicators?

BMC medical research methodology·2019
Same author

Profiling Hospital Performance on the Basis of Readmission After Transcatheter Aortic Valve Replacement in Ontario, Canada.

Journal of the American Heart Association·2019
Same author

Incidence and Risk Factors of Keratinocyte Carcinoma After First Solid Organ Transplant in Ontario, Canada.

JAMA dermatology·2019
Same author

Days Alive and Out of Hospital: Validation of a Patient-centered Outcome for Perioperative Medicine.

Anesthesiology·2019
Same author

Derivation and External Validation of a Scoring System for Predicting Fracture Risk After Ischemic Stroke in a Canadian Cohort.

JAMA neurology·2019
Same author

Importance of Nonobstructive Coronary Artery Disease in the Prognosis of Patients With Heart Failure.

JACC. Heart failure·2019
Same journal

Interpretable Bayesian Modeling for Multireader Multicase Studies: Addressing Overdispersion and Limited Sample Size in Diagnostic Enhancement Evaluation.

Statistics in medicine·2026
Same journal

Adaptive Sequential Multiple Hypotheses Testing for Concomitant Vaccine Safety Surveillance.

Statistics in medicine·2026
Same journal

Novel Distance Regression for Repeated Outcomes With Missing Data: Applications to Longitudinal and Crossover Studies of Microbiome Beta-Diversity.

Statistics in medicine·2026
Same journal

Optimal Weighted Tests for Replication Studies and the 'Two-Trials Rule' With Multiple Hypotheses.

Statistics in medicine·2026
Same journal

Identifiable Copula-Double-Cox Models: A Fully Parametric Framework for Dependent Right-Censored Survival Data.

Statistics in medicine·2026
Same journal

Moving From Individualized Risk-Based Prevention to Benefit-Based Prevention: Estimating Individualized Life-Years Gained From Prevention Services as a Basis for Eligibility.

Statistics in medicine·2026
See all related articles

This study introduces a novel bootstrap method to improve regression coefficient estimation and confidence intervals after backwards variable elimination. The new approach offers better coverage rates for selected variables in statistical modeling.

Area of Science:

  • Statistics
  • Biostatistics
  • Regression Analysis

Background:

  • Automated model selection, like backwards variable elimination, is common in applied research for parsimonious regression models.
  • Concerns exist regarding biased regression coefficients and inaccurate confidence interval coverage rates with traditional methods.
  • There is a need for improved statistical techniques in regression modeling.

Purpose of the Study:

  • To develop and evaluate a bootstrap-based method to enhance the estimation of regression coefficients and confidence intervals.
  • To address the limitations of conventional backwards variable elimination in statistical modeling.
  • To provide more reliable statistical inference for selected predictors in regression analysis.

Main Methods:

  • A novel method employing backwards variable elimination within multiple bootstrap samples was developed.

Related Experiment Videos

  • Regression coefficients for unselected variables were set to zero in each bootstrap sample.
  • Averaging coefficients across samples and constructing non-parametric percentile bootstrap confidence intervals were performed.

    • Main Results:

    • Monte Carlo simulations demonstrated the method's performance in estimating regression coefficients and confidence intervals.
    • The proposed bootstrap method yielded confidence intervals with superior coverage compared to conventional approaches.
    • The method showed improved statistical accuracy for variables selected via backwards elimination.

    Conclusions:

    • The developed bootstrap technique offers a significant improvement over standard backwards variable elimination for regression analysis.
    • This method provides more accurate confidence intervals and less biased coefficient estimates.
    • The approach is applicable to real-world datasets, as illustrated with a heart attack patient cohort.