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

A cautionary note on applying scores in stratified data

M J Podgor1, J L Gastwirth

  • 1National Eye Institute, Division of Biometry and Epidemiology, National Institutes of Health, Bethesda, Maryland 20892.

Biometrics
|December 1, 1994
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

Extension of the Peters-Belson method to estimate health disparities among multiple groups using logistic regression with survey data.

Statistics in medicine·2014
Same author

Power and related statistical properties of conditional likelihood score tests for association studies in nuclear families with parental genotypes.

Annals of human genetics·2005
Same author

On the power of affected relative pair designs for linkage studies.

Annals of human genetics·2004
Same author

Robust TDT-type candidate-gene association tests.

Annals of human genetics·2002
Same author

Statistical properties of Teng and Risch's sibship type tests for detecting an association between disease and a candidate allele.

Human heredity·2002
Same author

Trend tests for case-control studies of genetic markers: power, sample size and robustness.

Human heredity·2002

For stratified data analysis using rank tests, three scoring methods exist. Methods S and A are recommended for moderate to large strata, while method P shows sensitivity to score functions in smaller strata.

Area of Science:

  • Statistics
  • Nonparametric Statistics
  • Data Analysis

Background:

  • Rank tests are utilized for analyzing stratified data.
  • Existing methods for assigning scores include independent scoring within strata (S), aligned and pooled scoring (A), and pooled scoring without alignment (P).

Purpose of the Study:

  • To evaluate the performance of three scoring methods for rank tests in stratified data analysis.
  • To identify potential sensitivities and provide recommendations for method selection based on stratum size.

Main Methods:

  • Comparison of three scoring methods (S, A, P) for linear rank tests on stratified data.
  • Analysis of test statistics formed by combining stratum-specific linear rank tests.
  • Investigation of score function sensitivity for method P with varying stratum sizes.

Related Experiment Videos

Main Results:

  • Method P demonstrates sensitivity to the chosen score function when analyzing two moderately sized strata.
  • Methods S and A are generally more robust across different stratum sizes.

Conclusions:

  • Methods S (independent scoring) and A (aligned and pooled scoring) are recommended for analyzing moderate to large stratified datasets.
  • Method P (pooled scoring without alignment) should be used with caution, particularly with smaller strata, due to score function sensitivity.