A Youden plot draws each subject's or lab's two paired measurements against each other with median crosshairs, so the quadrant pattern separates systematic from random differences.
Youden's two-sample plot, designed for inter-laboratory studies. The crosshairs divide the plot into four quadrants; points clustered on one diagonal indicate systematic bias (labs that score high on test 1 also score high on test 2 — could be calibration drift); points scattered across all four quadrants indicate random measurement error.
A 45° pattern in the upper-right / lower-left quadrants ⇒ consistent high / low scorers; a scatter perpendicular to that ⇒ no systematic ranking.
Two paired numeric columns.
Scatter with x- and y-median crosshair lines.
Outliers far from origin in any single quadrant flag labs / runs to follow up. Clear 45° vs anti-45° patterns separate calibration drift from random measurement noise.