Nemours

“Stratification versus Regression Analysis: When and when not to use Stratified model in controlling for confounding”

Laurens Holmes, Jr, PhD
Board Certified Public Health
Director, Health Disparities Science Research, OHEI, Nemours/AIDHC
Adjunct Professor of Clinical Trials & Molecular Epidemiology,
Biological Sciences Department, University of Delaware
Editor-in-Chief, Pediatric Health, Medicine & Therapeutics

Host: Wilmington

Highlights!!!

  • Statistical test are based on assumptions regarding the variables used in the analysis, with the violation of such assumptions resulting in biased estimates.
  • But few researchers ever test the assumptions that their test is based on, implying potentials for biased conclusion and inference from the data!
  • Biomedical and clinical research must be more accountable by testing assumptions prior to evidence discovery from the data
  • When and when not to use regression models!
  •  

    Shirley Morrison invites you to an online meeting using WebEx.

    Meeting Number: 731 067 765
    Meeting Password: This meeting does not require a password.

    Location:
    Wilmington Lecture Hall & RC1 via WebEx
    1600 Rockland Road
    Wilmington, DE 19803

    Hosting Organization:
    Nemours

    Date:
    April 4, 2016

    Time:
    12:00 - 1:00 pm