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

Choosing among generalized linear models applied to medical data

J K Lindsey1, B Jones

  • 1Department of Medical Statistics, School of Computing Sciences, De Montfort University, Leicester, U.K.

Statistics in Medicine
|February 17, 1998
PubMed
Summary

Choosing the right statistical model is crucial for accurate research findings. This study highlights generalized linear models and model selection criteria, like the Akaike information criterion (AIC), for robust analysis of treatment effects.

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

Effects of an antiperspirant with emollients on foot-sweat accumulation and blister formation while walking in the heat.

Journal of the American Academy of Dermatology·1995
Same author

The achievement of isoeffective bronchial mucosal dose during endobronchial brachytherapy.

International journal of radiation oncology, biology, physics·1995
Same author

Entry of microbes into the host: using M cells to break the mucosal barrier.

Current opinion in immunology·1995
Same author

Derivation of the optimum dose per fraction from the linear quadratic model.

The British journal of radiology·1995
Same author

On the nature of the mutation in the nude rat.

Trends in genetics : TIG·1995
Same author

Radiotherapy and chemotherapy for inoperable non-small cell lung cancer.

Postgraduate medical journal·1995

Area of Science:

  • Biostatistics
  • Statistical Modeling
  • Clinical Research Methodology

Background:

  • Distributional assumptions in statistical testing significantly influence study conclusions.
  • Response variable transformations can lead to controversy in data analysis.
  • Generalized linear models offer an alternative but require appropriate member selection.

Purpose of the Study:

  • To address the challenge of selecting appropriate generalized linear models for hypothesis testing.
  • To apply model selection criteria for comparing non-nested hypotheses.
  • To evaluate differences in T4 cell counts between disease groups using robust statistical methods.

Main Methods:

  • Utilized generalized linear models as an alternative to response variable transformations.

Related Experiment Videos

  • Employed standard model selection criteria, specifically the Akaike information criterion (AIC).
  • Applied these procedures to analyze T4 cell count differences in two disease groups.

    • Main Results:

    • Demonstrated the application of AIC for selecting optimal generalized linear models.
    • Successfully identified differences in T4 cell counts between the studied disease groups.
    • Highlighted the importance of pre-specified model selection criteria in study protocols.

    Conclusions:

    • Appropriate model selection criteria are essential for drawing valid inferences in research.
    • Specifying model selection criteria in study protocols, including clinical trials, ensures optimal analysis.
    • Generalized linear models with proper selection criteria provide a reliable framework for assessing treatment effects.