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

Regression models for prognostic prediction: advantages, problems, and suggested solutions.

F E Harrell, K L Lee, D B Matchar

    Cancer Treatment Reports
    |October 1, 1985
    PubMed
    Summary

    Accurate patient outcome prediction using multiple regression models requires validating assumptions and avoiding overfitting. Data reduction methods improve models when sample sizes are small, ensuring reliable predictions.

    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

    Impact of the COVID-19 pandemic on a tertiary care public hospital in Singapore: resources and economic costs.

    The Journal of hospital infection·2021
    Same author

    Pre-discharge prognostic factors of physical function among older adults with hip fracture surgery: a systematic review.

    Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA·2019
    Same author

    Serum Cholesterol Levels within the High Normal Range Are Associated with Better Cognitive Performance among Chinese Elderly.

    The journal of nutrition, health & aging·2016
    Same author

    Chapter 6: assessing applicability of medical test studies in systematic reviews.

    Journal of general internal medicine·2012
    Same author

    The effect of obesity on outcome of unrelated cord blood transplant in children with malignant diseases.

    Bone marrow transplantation·2010
    Same author

    Unrelated umbilical cord blood transplantation in children with immune deficiency: results of a multicenter study.

    Bone marrow transplantation·2009

    Area of Science:

    • Biostatistics
    • Clinical Research Methodology
    • Predictive Modeling

    Background:

    • Multiple regression models are widely used for predicting patient outcomes across various diseases.
    • A common issue is the failure to validate model assumptions and the tendency to overfit data.
    • Overfitting, caused by too many predictors and small sample sizes, leads to unreliable models.

    Purpose of the Study:

    • To highlight the importance of validating assumptions in multiple regression models.
    • To address the problem of overfitting in predictive modeling.
    • To discuss methods for improving regression model performance and accuracy.

    Main Methods:

    • Review of multiple regression model development and validation practices.
    • Discussion of data reduction techniques for small sample sizes.

    Related Experiment Videos

  • Comparison of regression models with stratification and recursive partitioning.

    • Main Results:

    • Models developed without validating assumptions or with overfitting are unlikely to perform well on new data.
    • Data reduction methods can significantly enhance regression model performance when the ratio of patients to predictors is low.
    • Validated regression models outperform other methods when assumptions are met and overfitting is avoided.

    Conclusions:

    • Thorough validation of assumptions and avoidance of overfitting are critical for reliable multiple regression models.
    • Data reduction techniques are valuable for improving predictive accuracy in challenging sample size scenarios.
    • Regression models offer superior predictive accuracy when rigorously developed and validated.