Search research articles

Search research articles

Related Experiment Videos

Predictive accuracy and explained variation.

Michael Schemper¹

¹Section of Clinical Biometrics, Department of Medical Computer Sciences, Vienna University, Spitalgasse 23, A-1090 Vienna, Austria. michael.scheper@akh-wien.ac.at

Statistics in Medicine

|July 11, 2003

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

Explained variation and degrees of necessity and of sufficiency for competing risks survival data.

Biometrical journal. Biometrische Zeitschrift·2024

Same author

Degrees of necessity and of sufficiency: Further results and extensions, with an application to covid-19 mortality in Austria.

Statistics in medicine·2021

Same author

Quantifying degrees of necessity and of sufficiency in cause-effect relationships with dichotomous and survival outcomes.

Statistics in medicine·2019

Same author

Explained variation in shared frailty models.

Statistics in medicine·2017

Same author

Asymptomatic Severe Aortic Stenosis in the Elderly.

JACC. Cardiovascular imaging·2016

Same author

Role of a heart valve clinic programme in the management of patients with aortic stenosis.

European heart journal. Cardiovascular Imaging·2016

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

This study introduces unified measures for predictive accuracy and explained variation in regression models. These metrics quantify how well covariates predict outcomes, even when significance is high.

Area of Science:

Statistics
Biostatistics
Epidemiology

Background:

Predictive accuracy quantifies how well covariates determine individual outcomes in regression models.
Explained variation measures the relative improvement in prediction accuracy when using covariates compared to unconditional prediction.

Purpose of the Study:

To present a unified concept of predictive accuracy and explained variation.
To provide measures applicable to various outcome types (continuous, binary, polytomous, survival).
To demonstrate applications using examples from different regression models.

Main Methods:

Development of a unified concept based on absolute prediction error.
Formulation of measures in both model-based and observed-vs-expected contrasts.

Related Experiment Videos

Application examples across continuous, binary, polytomous, and survival regression models.

Main Results:

A unified framework for assessing predictive accuracy and explained variation is presented.
Measures are applicable across diverse regression model types and outcome variables.
Demonstrated that predictive accuracy can be low even with significant covariates.

Conclusions:

The proposed unified measures offer a consistent approach to evaluating regression model performance.
Emphasizes that statistical significance and effect size of covariates do not always translate to high predictive accuracy.
Highlights the importance of assessing absolute and relative predictive accuracy in regression analysis.