01Identify what the correction is intended to estimate

Eurachem’s 2026 uncertainty guide distinguishes total-analyte measurements from quantities defined by a particular extraction procedure. It also discusses circumstances in which recovery is corrected or handled through a specified performance framework. A recovery below 100% does not, on its own, select the appropriate approach.Eurachem/CITAC — Evaluation of measurement uncertainty from in-house precision and recovery data (opens in a new tab)

Imagine that a report aims to estimate the total amount of a target compound in the sampled material. Its method model includes an independently supported recovery factor. That is different from a report intentionally measuring only the amount extracted under named conditions.

Before comparing their values, ask whether both reports intend to describe the same quantity. Correcting an operationally defined extractable amount towards an imagined total can change the question rather than improve its answer.

02Work through one explicit correction

In an original hypothetical case, the uncorrected estimate is 9.0 mg and a justified recovery factor is 0.90. The stated model divides the uncorrected estimate by that factor: 9.0 mg ÷ 0.90 = 10.0 mg. This is a simplified report-reading calculation, not a Novum test result.

Dividing by 90 instead of 0.90 introduces a factor-of-100 error. Multiplying 9.0 by 0.90 moves the result in the opposite direction from this specified correction. The equation matters more than an informal instruction to adjust for recovery.

The difference between 9.0 and 10.0 is explained here entirely by the reporting model. It does not demonstrate that the laboratory physically recovered an additional milligram.

03Check that the recovery evidence belongs to this result

A correction based on a different sample matrix, analyte form or concentration range may not describe the present measurement. Request the basis for its applicability rather than assuming that any recovery study supplies the required factor.

Eurachem includes recovery in measurement models and uncertainty evaluation. Correcting the central estimate therefore does not remove uncertainty about the factor itself.Eurachem/CITAC — Evaluation of measurement uncertainty from in-house precision and recovery data (opens in a new tab)

For the same illustrative 9.0 mg input, factors of 0.88 and 0.92 would give approximately 10.23 and 9.78 mg. These alternatives are a sensitivity illustration, not a confidence interval or a claim that those recoveries were observed.

A factor should not be selected because it makes a result pass. Nor can it automatically repair a saturated detector, an incorrect sample identity or a calculation using the wrong preparation volume.

04Compare reports at a common calculation stage

If one certificate lists 9.0 mg uncorrected and another lists 10.0 mg corrected, first establish whether they derive from the same measurement and factor. They may be two presentations of one observation rather than independent confirmation.

Check whether the calibration or internal-standard calculation already compensates for the relevant loss. An additional manual division may count the same effect twice even when the final certificate does not use the word corrected prominently.

A useful clarification identifies the original estimate, correction equation, factor, supporting evidence and final uncertainty. If those details are unavailable, keep the two reporting bases distinct instead of silently converting either value.