Row of volumetric flasks, an analytical balance and a rack of vials on a stainless lab bench

Analytical Method Validation Explained: Accuracy, Precision, LOD and LOQ

Every so often a report will mention that a result came from a “validated method”. It sounds reassuring, and usually it is, but it's worth knowing what it actually means. Analytical method validation is the documented evidence that a lab procedure is fit for the specific job it's being used for. A result is only as good as the method behind it, so a bit of understanding here goes a long way when you're comparing reports.

Row of volumetric flasks, an analytical balance and a rack of vials on a stainless lab bench

Analytical method validation starts with purpose

Validation always starts with a question: what's this method meant to measure, in what kind of sample, and how accurately does it need to do it? A method built to confirm identity has different requirements from one designed to measure trace impurities. Labs lean on international guidance such as the ICH Q2 guideline, which lists the characteristics to assess depending on what the method's for. Not all of them apply to every method.

Think about kitchen scales. The ones that are perfect for weighing flour for a sponge are useless for a pinch of saffron. Fit for purpose depends entirely on what you're asking them to do.

Pillars of method validation: specificity, accuracy, precision, linearity, LOD/LOQ and robustness, resting on a defined intended purpose
Illustrative diagram.

The terms you'll come across

Specificity (sometimes called selectivity) is whether the method can measure the target without interference from other things in the sample, like impurities, breakdown products or formulation ingredients. For chromatography, that usually means showing the main peak is cleanly separated from its neighbours. Without it, nothing else on the list counts for much.

Accuracy is how close a result lands to the true or accepted value. Labs normally test it by analysing a sample with a known amount of the target and seeing how much they get back. If a sample's prepared with a known quantity and the method reports 99% of it, that's 99% recovery.

Precision is about consistency, not correctness. Picture a dartboard at the pub: three darts clustered tightly in the outer ring are precise, but they're not accurate. In the lab it's usually expressed as a relative standard deviation and checked at a few levels. Repeatability is the same analyst, same instrument, same day, running the same sample several times. Intermediate precision looks at variation within one lab across different days, analysts or instruments. Reproducibility is agreement between different labs, and that's usually tested in collaborative studies rather than routine validation.

Linearity shows that the method's response rises in proportion to the amount of compound across a stated range. The range is the span between the lowest and highest amounts where the method has been shown to be accurate, precise and linear. A result outside it is less certain, even if the number looks very precise.

Then there's what labs call robustness: how well the method copes with small, deliberate changes to conditions like column temperature, solvent mix or flow rate. A method that holds up there gives dependable results even when day-to-day conditions drift a little, which matters when it's run often and by different hands.

Laboratory bench with volumetric flasks, a rack of capped sample vials, a notebook, calculator and nitrile gloves

LOD and LOQ aren't the same thing

These two get mixed up all the time. The limit of detection (LOD) is the smallest amount that can be reliably told apart from background noise, but not necessarily measured accurately. The limit of quantitation (LOQ) is the smallest amount that can be measured with acceptable accuracy and precision. It matters when you see “not detected” on a report. That means the substance was below the detection limit for that method, not that it's definitely absent. Our guide to understanding research documentation touches on this too.

One more distinction while we're here. Validation happens when a method is developed or changed, to show it's fit for purpose. System suitability is a routine check each time the method's run, to confirm the instrument's behaving on the day. Both matter, and one doesn't replace the other. You'll see system suitability come up in our guide on how to read an HPLC chromatogram.

Reading a report with all this in mind

Short certificates often summarise results without any validation detail. That isn't automatically a problem, but it's fair to ask. When we look over a report, we want to see the method named or referenced, and a method that suits what's being reported, whether that's identity, purity or quantity. Units and limits should be stated, a detection limit should sit alongside any “not detected”, and if the lab claims accreditation, it should actually cover that test. If something's unclear, you can always ask the issuing lab or the supplier. For a broader checklist, see how to verify a certificate of analysis.

We don't make validation claims ourselves, because the reports we share come from third-party labs, not from us. What we do is publish those COAs, each with a verification link, for a number of our products on our COAs page, so you can see who issued each one and check it. If you've got a question about one, flick us an email at support@ascendaus.com.

Frequently asked questions

What is analytical method validation?

It's documented evidence that a lab test method suits its intended purpose, covering things like specificity, accuracy, precision, linearity, range, detection limits and robustness.

What's the difference between accuracy and precision?

Accuracy is how close a result is to the true value. Precision is how closely repeat results agree with each other. A method can be precise and still be off target.

What's the difference between LOD and LOQ?

The limit of detection is the smallest amount that can be detected. The limit of quantitation is the smallest amount that can be measured with acceptable accuracy and precision.

Does “not detected” mean a substance is absent?

No. It means the amount was below that method's detection limit. A more sensitive method might still pick it up.

Is system suitability the same as method validation?

No. Validation shows a method is fit for purpose; system suitability checks the instrument is working properly each time the method's run.

Research use only. For laboratory research use only. Not for human or animal consumption, and not for therapeutic or diagnostic use.

Cheers,
The Ascend Aus team

Back to blog