Bland–Altman Plot Calculator: Bias, Limits of Agreement and Method Comparison
Paste paired measurements of the same samples by two methods (a new device and a reference, two raters, two labs). The calculator gives the mean bias, the 95% limits of agreement with their confidence intervals, a test for proportional bias, Lin’s concordance correlation coefficient, and the Bland–Altman plot, with differences on the original scale, as percentages or as ratios.
- PriceFree, no login or sign-up needed
- Your dataProcessed in your browser, never uploaded
- Works onPhone, tablet or computer, any modern browser
- Made byDr. Muhammad Aammar Tufail, Codanics
- Published17 November 2026
- Last updated16 November 2026
📚 How to cite this tool
This tool is free. If it helped your thesis, paper, report or assignment, please cite it. It keeps the tools free and helps other students find them.
Tufail, M. A. (2026). Bland–Altman plot calculator: bias and limits of agreement [Online tool]. Codanics. Retrieved 16 November 2026, from https://codanics.com/tools/bland-altman-calculator/
In-text citation: (Tufail, 2026) · Narrative: Tufail (2026)
The access date is today’s date, filled in automatically. Author: Muhammad Aammar Tufail. Published 17 November 2026.
How to use the Bland–Altman Plot Calculator
- Enter the paired measurements: method A in the first column and method B in the second, one sample per line (or paste them from Excel). A simulated comparison of a handheld chlorophyll meter with laboratory extraction on 40 leaves is loaded as an example.
- Name the two methods and choose how to express the differences: A − B when the spread of the differences is constant, percentage or ratio when it grows with the size of the measurement.
- Optionally enter the acceptable difference: the largest disagreement that would not matter in practice. Decide it before looking at the data.
- Click Compare methods. Read the bias and the limits of agreement with their confidence intervals, check the plot for patterns and the proportional bias test.
- Copy the results sentence and tables, download the differences as CSV and save the Bland–Altman plot as PNG or SVG.
What a Bland–Altman analysis shows
When a new method (a cheaper device, a field kit, a second rater) is compared with an established one, a high correlation is not enough: two methods can be perfectly correlated yet differ by a constant amount or a constant factor. Bland and Altman (1986) proposed plotting the difference between the two measurements against their mean and summarising agreement with the mean difference (bias) and the 95% limits of agreement: the range within which 95% of the differences between the methods are expected to lie.
How to read the plot
- The solid line is the bias. If its confidence interval excludes zero, one method reads systematically higher.
- The dashed lines are the limits of agreement. Whether they are acceptable is a practical, not a statistical, judgement: compare them with the largest difference that would not matter.
- A funnel shape (differences grow with the mean) calls for percentage or ratio differences.
- A sloping cloud indicates proportional bias: one method over-reads at high values and under-reads at low values.
Formulas
The bias, limits of agreement and their CIs, the proportional-bias regression and Lin’s CCC with its CI were checked against hand-coded R and the formulas of epiR::epi.ccc(). The CIs of the limits use the approximate method of Bland and Altman (1999); exact intervals (Carkeet, 2015) are slightly wider for small samples.
Designing a method-comparison study
- Measure at least 30–50 (ideally 100) samples covering the whole range of values you expect in practice.
- Decide the acceptable difference in advance, from clinical or practical requirements.
- Measure each sample with both methods at the same time, in random order, and blind to the other result.
- For agreement between raters on categories, use Cohen’s kappa instead; for reliability of continuous ratings, use the ICC.
Frequently asked questions
What are the limits of agreement?
They are the mean difference ± 1.96 standard deviations of the differences. About 95% of the differences between the two methods are expected to fall between them. If the limits are narrow enough for your purpose, the methods can be used interchangeably.
Why not just use correlation?
Correlation measures how closely the points follow any straight line, not whether the two methods give the same values. A method that always reads twice as high correlates perfectly but agrees badly. The Bland–Altman plot and Lin’s concordance coefficient measure agreement with the line of equality.
When should I use percentage or ratio differences?
When the spread of the differences grows with the size of the measurement (a funnel shape in the plot), as is common for concentrations and counts. Percentage differences express disagreement relative to the mean; ratio differences analyse log(A/B), so the limits become ratios such as 0.85 to 1.12.
What is proportional bias?
It means that the difference between the methods depends on the size of the measurement, for example a meter that reads too low on pale leaves and too high on dark ones. It is tested by regressing the differences on the means. If it is present, a single bias and fixed limits are misleading.
How many samples do I need?
Bland and Altman recommend at least 30–50 pairs, and 100 or more for precise limits. The confidence intervals of the limits show how precise they are; if they are wide, measure more samples.
What is a good Lin's CCC?
Lin’s concordance correlation coefficient ranges from −1 to 1, with 1 meaning perfect agreement on the line of equality. McBride (2005) suggests below 0.90 poor, 0.90–0.95 moderate, 0.95–0.99 substantial and above 0.99 almost perfect agreement for measurement methods.
Should the differences be normally distributed?
The 95% limits assume roughly normal differences. Check the histogram or use the normality test calculator on the differences (download them as CSV). With skewed differences, try the ratio option.
Do I need to log in or create an account to use the Bland–Altman Plot Calculator?
No. There is no login, no sign-up and no email required. Open the page and start using it straight away, for free, on any device.
How do I cite this tool in my thesis or paper?
Use the citation box on this page. In APA 7th edition: Tufail, M. A. (2026). Bland–Altman plot calculator: bias and limits of agreement [Online tool]. Codanics. Retrieved [date you used it], from https://codanics.com/tools/bland-altman-calculator/. BibTeX, RIS (EndNote, Mendeley, Zotero), MLA, Harvard, Chicago and Vancouver formats are also available.
Free guides and courses from Codanics
More free tools
See all free Codanics tools → · Codanics on YouTube · WhatsApp channel