Bland-Altman Agreement Analysis
Assess the agreement of two measurement methods (or two measurements). Draw the Bland-Altman plot, compute the mean bias and 95% limits of agreement (LoA = bias ± 1.96SD), and judge whether the two methods are interchangeable. It reflects systematic error and agreement between methods better than a correlation coefficient.
① Input data
One paired measurement per row: method 1's value + method 2's value, space/Tab/comma-separated. You can paste two columns from Excel.
How to use & methodology
What is Bland-Altman analysis used for?
To assess the agreement of two measurement methods (or two measurements by the same method) and judge whether they are interchangeable. Common in comparing a new method with the gold standard, or measurements from different devices/sequences. It focuses on the 'between-method difference', not correlation.
Why not use a correlation coefficient to judge agreement?
A high correlation only means the two methods 'change together', not that the values agree — even if one systematically overestimates the other, correlation can be high. Bland-Altman looks directly at the bias and spread of the differences, revealing systematic error that correlation hides, and is the correct method for assessing agreement.
How do I interpret the bias and limits of agreement?
The mean bias reflects the systematic difference between the methods (near 0 means no systematic bias); the 95% limits of agreement (bias ± 1.96SD) contain 95% of the differences. The narrower the limits, the better the agreement.
How do I judge whether agreement is acceptable?
There is no universal statistical threshold; it depends on the clinical context. If the limits of agreement fall within clinically acceptable error (e.g. no material effect on diagnostic decisions), the methods are interchangeable; otherwise, even if they look statistically close, they should not be substituted.