Section 01 · The Field E-05
Duplicate and spike
Run two. Add a known amount. Everything else in the report rests on these two habits.
What each check actually does
A duplicate is a second sampler run alongside the first under identical conditions — same location, same duration, same analytical sequence. If the method is working and the air is uniform, the two results should agree. How closely they agree is expressed as a relative percent difference: the gap between the two readings divided by their mean, multiplied by a hundred. A figure below ten or fifteen percent is generally acceptable for field air sampling; wider than that, and something was wrong with one sampler, the analysis, or the assumption that both samplers saw the same air.
A spike — sometimes called a spiked sample or a field spike — works differently. Here, a known mass of the target compound is added to a sampler before or during deployment, or to a sample extract before analysis, and then that sampler is taken through the full analytical procedure. The recovered amount is compared to the amount added. The ratio, expressed as a percentage, is the recovery for that occasion. If the recovery falls outside the acceptable window — commonly eighty to a hundred and twenty percent — the analytical run is flagged, and the results that came out of it carry a qualification.
Why both are necessary
The two checks answer different questions. A duplicate tells you about reproducibility: were the collection and handling consistent? A spike tells you about accuracy: did the method actually find what was there? A method can be perfectly reproducible and still systematically miss half the compound if the sorbent is saturated or the desorption step is inefficient. Conversely, a spike recovery can look fine while one of a duplicate pair has been contaminated or allowed to break through. Running only one check leaves the other question unanswered.
Neither check is optional in a defensible measurement programme. Regulators, accreditation bodies and the laboratories that carry method validation forward all require documented duplicate and spike data as part of a chain of evidence. When a result appears in a report, the reader is entitled to ask: what was the duplicate agreement on this occasion, and what was the field spike recovery? If those numbers are absent, the result is unsupported — not necessarily wrong, but unverified by the method itself.
Filed inSection 01 — The Field