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 nonparametric changepoint model for stratifying continuous variables under order restrictions and binary outcome.

Georgia Salanti1, Ulm Kurt

  • 1Institute for Medical Statistics and Epidemiology, Klinikum Rechts der Isar, Munich, Germany. vorgia@web.de

Statistical Methods in Medical Research
|August 27, 2003
PubMed
Summary

Monotonic regression offers a stable alternative to parametric models for continuous predictors. This study introduces methods for model parsimony and handling multiple predictors, creating interpretable constant risk groups.

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

Psychotherapy, Antidepressants, and Combined Treatment for Depression: A Network Meta-analysis on Social Functioning Outcomes.

Psychotherapy and psychosomatics·2026
Same author

Comparative effectiveness and safety of pharmacological treatments for rapid tranquilisation in emergency settings: a systematic review and individual participant data network meta-analysis.

The lancet. Psychiatry·2026
Same author

The GALENOS approach to triangulating evidence: a structured approach for integrating information from human and animal studies.

BMC medical research methodology·2026
Same author

Reclassifying dropouts from antipsychotic trials due to study-reported adverse events: a systematic review and individual-participant-data meta-analyses.

European neuropsychopharmacology : the journal of the European College of Neuropsychopharmacology·2026
Same author

Pharmacological interventions for ADHD: a systematic review and dose-effect network meta-analysis.

The lancet. Psychiatry·2026
Same author

The Use of Baseline Score in Cost-Effectiveness Modeling of Competing Interventions.

Value in health : the journal of the International Society for Pharmacoeconomics and Outcomes Research·2026

Area of Science:

  • Statistics
  • Biostatistics
  • Regression Analysis

Background:

  • Parametric models may not be optimal for continuous predictors requiring stratification.
  • Monotonic regression provides an alternative framework for such scenarios.
  • Existing methods lack efficient procedures for model parsimony and multivariate predictor handling.

Purpose of the Study:

  • To propose a method for enhancing the parsimony of monotonic regression models.
  • To extend monotonic regression to handle multiple predictor variables, including interactions.
  • To develop a simple, interpretable model combining reduced monotonic regression and monotonic surface estimation.

Main Methods:

  • A reducing procedure using Fisher exact tests and bootstrap for model selection.

Related Experiment Videos

  • An iterative algorithm, an extension of the Pool Adjacent Violators Algorithm (PAVA), for multiple predictors.
  • Permutation-based approach for assessing p-values in multivariate monotonic likelihood ratio tests.

    • Main Results:

    • A method to select between full and reduced monotonic regression models was developed.
    • A monotonic surface model was proposed, handling predictor interactions effectively.
    • A combined approach resulted in a simple model with constant risk groups.

    Conclusions:

    • The proposed methods enhance model parsimony and extend monotonic regression to multivariate settings.
    • The monotonic surface model offers an alternative to additive models when interactions are present.
    • The developed approach provides a stable and interpretable modeling strategy, particularly beneficial for bivariate step functions.