Editverse Free Calculator

Linearity Assessment Calculator (Polynomial Test)

Check whether an assay’s response is linear across its measuring range: the calculator fits a straight line and a quadratic to your level means, tests the quadratic term and reports the largest deviation from the line, with the exact R command behind the result. A full guide follows below. Validated against R 4.3.3.

In Simple Words

Linearity is the ability of an assay to return results directly proportional to the true concentration across its measuring interval. This calculator uses the polynomial idea behind the earlier CLSI EP6-A (2003): it fits the measured-versus-expected data with a straight line and a second-order polynomial and tests whether the curved term adds anything beyond the line. The current EP06 2nd edition (2020) judges each level’s deviation from the straight line against an allowable nonlinearity, so compare the reported maximum deviation with your own limit.

If the non-linear terms are not significant, or the deviation from linearity stays within a pre-set allowable limit, the method is judged linear over the range tested. The point of the test is not statistical significance alone but whether any curvature is large enough to matter clinically — small but real non-linearity may still be acceptable.

When to Use a Linearity Assessment

Assess linearity during method validation and verification, and when establishing the analytical measuring range or reportable interval of a quantitative assay. It confirms that dilutions and high or low values are reported accurately.

Prepare samples spanning the claimed range, replicate at each level, and define the allowable deviation from linearity before analysing. Judge clinical importance, not just the p-value of the non-linear terms. Pair linearity with detection limits at the low end and method-comparison regression such as Passing-Bablok. Unsure how to set the allowable deviation? Our Test Selector can help.

Linearity (polynomial test) Test assay linearity across the range · validated vs R 4.3.3
User guide

When to use. Check that an assay responds linearly across its reporting range by testing for a quadratic component (the polynomial approach of the earlier CLSI EP6-A).

Assumptions. At least five equally informative levels; replicates averaged per level; constant variance across the range.

Common mistakes. A high R² alone does not prove linearity. Not significant does not prove linear with very few levels: inspect residuals.

Worked example

Click Load example data above, then Run Analysis to see a fully worked result with figure, formula and interpretation for this test.

References
  1. CLSI. Evaluation of the Linearity of Quantitative Measurement Procedures; EP06. 2nd ed. 2020.

Worked Examples

Example 1 — Linear across the range

Samples at several concentrations across the measuring interval are measured in replicate. Fitting a straight line and a second-order polynomial shows the quadratic term adds nothing significant, and the deviation from a straight line stays within the allowable limit. The method is linear over the range.

Example 2 — Significant but unimportant curvature

A second-order term is statistically significant, but the largest deviation from linearity is well within the allowable limit set in advance. Because the curvature is too small to affect clinical decisions, the method is still accepted as linear for practical use.

Assumptions to Check

Samples spanning the measuring range. Linearity must be tested across the claimed analytical interval, not just near the middle.

Replication at each level. Replicate measurements at each concentration are needed to separate real curvature from random scatter.

A pre-defined allowable deviation. The acceptable deviation from linearity must be set in advance on clinical grounds, so the decision does not hinge on statistical significance alone.

Common Mistakes

Equating significance with non-linearity

A significant higher-order term can reflect trivial curvature. Judge the size of the deviation against a clinical limit, not just the p-value.

Testing too narrow a range

Assessing linearity only near the midpoint misses curvature at the extremes, where dilutions and high values actually matter.

No replication

Without replicates at each level, random scatter is mistaken for curvature or hides it. Replicate at every concentration.

Setting the allowable deviation afterwards

Choosing the acceptable deviation after seeing the data biases the conclusion. Define it before analysis.

Frequently Asked Questions

What does the linearity test actually check?

It fits a straight line and a second-order polynomial, tests whether the quadratic term improves the fit, and reports the largest percentage deviation from the line so you can compare it with your pre-set allowable limit.

Is a significant non-linear term a failure?

Not necessarily; if the actual deviation from linearity stays within the clinically allowable limit, the method can still be accepted as linear for practical use.

How wide a range should I test?

Across the full claimed measuring interval, with samples at the extremes, since that is where non-linearity has the most clinical impact.

Does the result match R?

Yes. The calculator fits the same linear and quadratic models as R’s lm() and shows the R command so you can reproduce the fit and the quadratic-term test.

References

  1. Jhang JS, Chang CC, Fink DJ, Kroll MH. Evaluation of linearity in the clinical laboratory. Arch Pathol Lab Med. 2004;128(1):44–48. PMID 14692813.
  2. Hsieh E, Liu JP. On statistical evaluation of the linearity in assay validation. J Biopharm Stat. 2008;18(4):677–690. PMID 18607801.
  3. CLSI. Evaluation of the Linearity of Quantitative Measurement Procedures. EP06. 2nd ed. Wayne, PA: Clinical and Laboratory Standards Institute; 2020.

Need This Analysis Done for Your Manuscript?

Our team supports method validation — CLSI EP6 linearity assessment with polynomial evaluation and allowable-deviation criteria — with reproducible R output.

Get Expert Statistical Help