Related Experiment Video
Updated: May 14, 2026

An R-Based Landscape Validation of a Competing Risk Model
Published on: September 16, 2022
A description of mixed group validation
Paul A Jewsbury1, Stephen C Bowden2
1University of Melbourne, Parkville, Victoria, Australia jewsbury@unimelb.edu.au.
Abstract:
Mixed group validation (MGV) is a statistical model for estimating the diagnostic accuracy of tests. Unlike the more common approach to estimating criterion-related validity, known group validation (KGV), MGV does not require a perfect external validity criterion. The present article describes MGV by (a) specifying both the standard error associated with MGV validity estimates and the effect of assumption violation, (b) recommending required sample sizes under various study conditions, (c) evaluating whether assumption violation can be identified, and (d) providing a simulated example of an MGV with imperfect base rate estimates. It is concluded that MGV will always have a wider margin of error than KGV, MGV performs best when the research design approximates a KGV design, the effect of assumption violation depends on the severity of the assumption violation and also the value of the base rates, and that assumption violation may only be detected in severe cases.
Related Concept Videos
Data Validation
Nursing assessment guides are generally based on holistic models rather than medical...
Data Validation
Key parameters for method validation include:
Group Design
Test for Homogeneity
Impact of Groups on Groups
In- and Out-Groups
