01Follow the sign through the calculation

In an original simplified example, a sample produces 9.7 response units and the relevant blank estimate is 10.0. Subtraction gives 9.7 − 10.0 = −0.3 response units. Neither original reading is negative; the difference is.

This illustration assumes that subtraction is appropriate for the method and that both readings use the same response basis. A chromatographic peak-area percentage or a differently processed blank cannot simply be substituted into the equation.

Converting a valid net response through a calibration can likewise produce a negative concentration estimate near background. The mathematical estimate and the physically possible concentration remain different concepts.

02Interpret the estimate alongside its uncertainty

The multi-agency MARLAP manual explicitly discusses negative radioanalytical results despite the impossibility of negative radioactivity. It recommends retaining numerical results with uncertainty, while recognising that a markedly negative value may indicate a quality-control problem.EPA — MARLAP, Chapter 19: Measurement Uncertainty (opens in a new tab)

That is a radioanalytical reporting example, not a blanket reporting rule for every peptide assay. Its useful lesson is that a negative estimate cannot be interpreted sensibly without its measurement context.

For illustration, a net estimate of −0.3 response units with a standard uncertainty of 0.5 units is a different situation from −0.3 with a standard uncertainty of 0.01. The same displayed value has very different scale relative to the stated uncertainty.

03See what replacing negatives with zero does

Consider three hypothetical net estimates from a stable near-background series: −0.3, 0.0 and +0.3 units. Their arithmetic mean is zero. Replacing only the negative estimate with zero changes the mean to +0.1 units.

That change comes from the editing rule, not from new analytical evidence. The example shows why a data table used for statistical analysis may need to preserve signed numerical estimates even when a public summary uses a qualified reporting convention.

Conversely, retaining a negative estimate does not establish a negative physical amount or prove absence of analyte. A detection decision requires its own method and criterion; it cannot be inferred merely from whether a rounded value has a minus sign.

Keep the raw estimate, uncertainty, detection or quantification qualifier and any display transformation distinguishable. Those fields answer different questions and allow later readers to understand the processing.

04Check an unexpected negative result specifically

Review whether the blank matches the sample process, whether subtraction occurred once, whether units agree and whether the calibration was applied on its supported basis. These are concrete checks rather than explanations assumed in advance.

For example, subtracting the same 10-unit blank twice from the illustrative 9.7-unit reading would produce −10.3. That is a distinct calculation mistake, not another ordinary interpretation of −0.3.

A report that supplies only not detected may not expose the underlying signed estimate. Ask for the reporting convention when the numerical data are needed for a comparison; do not invent a value from the qualifier.

The defensible conclusion may be that the available report does not support a quantitative amount. State that limit directly while preserving any evidence that the measurement process itself needs investigation.