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

Some suggestions for measuring predictive performance

L B Sheiner, S L Beal

    Journal of Pharmacokinetics and Biopharmaceutics
    |August 1, 1981
    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

    Pharmacodynamic measurements of methyl nicotinate percutaneous absorption.

    Pharmaceutical research·2013
    Same author

    A new Bayesian method to forecast and fine tune individual hemodialysis dose.

    Hemodialysis international. International Symposium on Home Hemodialysis·2009
    Same author

    Ways to fit a PK model with some data below the quantification limit.

    Journal of pharmacokinetics and pharmacodynamics·2002
    Same author

    Selecting reliable pharmacokinetic data for explanatory analyses of clinical trials in the presence of possible noncompliance.

    Journal of pharmacokinetics and pharmacodynamics·2001
    Same author

    The need for mixed-effects modeling with population dichotomous data.

    Journal of pharmacokinetics and pharmacodynamics·2001
    Same author

    Evaluating pharmacokinetic/pharmacodynamic models using the posterior predictive check.

    Journal of pharmacokinetics and pharmacodynamics·2001
    Same journal

    Integrated equation to evaluate accumulation profiles of drugs eliminated by Michaelis-Menten kinetics.

    Journal of pharmacokinetics and biopharmaceutics·2010
    Same journal

    Applications of a recirculatory stochastic pharmacokinetic model: limitations of compartmental models.

    Journal of pharmacokinetics and biopharmaceutics·2010
    Same journal

    Effect of plasma protein and tissue binding on the time course of drug concentration in plasma.

    Journal of pharmacokinetics and biopharmaceutics·2010
    Same journal

    Pharmacokinetics of methotrexate in solid tumors.

    Journal of pharmacokinetics and biopharmaceutics·2010
    Same journal

    Single- and multiple-dose kinetics of oral lorazepam in humans: the predictability of accumulation.

    Journal of pharmacokinetics and biopharmaceutics·2010
    Same journal

    Comparison of the in vitro and in vivo release of digoxin from four different soft gelatin capsule formulations.

    Journal of pharmacokinetics and biopharmaceutics·2010
    See all related articles

    Traditional correlation coefficients poorly describe prediction model performance. Mean square prediction error (precision) and mean prediction error (bias) offer better insights but are unreliable with imprecise reference methods.

    Area of Science:

    • Statistics
    • Biostatistics
    • Measurement Science

    Background:

    • Model performance is frequently assessed using correlation coefficients or regression analysis.
    • These traditional metrics offer limited insight into the true predictive capabilities of a model.
    • A need exists for more robust evaluation methods.

    Purpose of the Study:

    • To highlight the limitations of correlation coefficients in evaluating predictive models.
    • To introduce mean square prediction error (precision) and mean prediction error (bias) as superior metrics.
    • To discuss the application and limitations of these improved measures, especially concerning reference method precision.

    Main Methods:

    • Analysis of prediction model evaluation techniques.
    • Comparison of correlation coefficients with mean square prediction error and mean prediction error.

    Related Experiment Videos

  • Discussion of the impact of reference method imprecision on evaluation metrics.

    • Main Results:

    • Correlation coefficients and regression provide an inadequate description of predictive performance.
    • Mean square prediction error (precision) and mean prediction error (bias) offer a more accurate assessment.
    • These improved measures are valuable for comparing prediction methods but require a precise reference standard.

    Conclusions:

    • Mean square prediction error and mean prediction error are recommended for a better evaluation of prediction models.
    • Caution is advised when using these metrics if the reference method is imprecise.
    • The study emphasizes the importance of selecting appropriate statistical measures for reliable model assessment.