Method Validation Made Simple: Can You Trust This Measurement?

A laboratory says its method is validated. That sounds reassuring, but validated to do what? A method that identifies a molecule is not automatically suitable for measuring a tiny impurity. Validation asks whether a procedure performs well enough for its stated purpose.

1. Define the question before judging the method

Write the intended decision in plain words: identify a peptide, quantify its content or assess a particular impurity near a limit. Then specify the material, sample conditions and range involved. ICH Q2(R2) organizes validation around fitness for that intended purpose.

If the laboratory’s demonstrated scope differs from your question, a validation summary may not answer your need. Ask for the applicable scope and version instead of requesting only a “validated” label.

2. Accuracy: does the answer sit near the accepted value?

For a teaching example, imagine a reference assigned a value of 100 units. Three results are 99, 100 and 101. Another set is 89, 90 and 91. The second set agrees internally, yet is shifted away from the reference. Internal agreement does not remove bias.

Accuracy concerns closeness to an accepted value. Assessing it requires an appropriate reference or other justified approach; you cannot establish it merely by repeating an unknown sample.

3. Precision: do repeated answers agree?

Precision concerns agreement among repeated measurements. In the example, both 99/100/101 and 89/90/91 are relatively close within their own sets. A set of 80/100/120 has a mean of 100 but much wider variation. An average alone can hide inconsistency.

Also ask what was repeated. Re-injecting one prepared sample differs from independently preparing samples. Measurements across days or analysts explore variation not captured by repeat injections alone.

4. Range: does the method work where you need it?

Suppose an invented quantitative method has demonstrated useful performance from 10 to 100 units. A result at 0.5 or 150 falls outside that example scope. A straight calibration line on paper is not sufficient evidence for reliable answers outside the tested range.

Detection and quantitation are different questions too. A weak signal may be detectable without being reliably quantifiable. For a result near a decision limit, ask which capability was actually demonstrated.

5. Specificity and changes deserve attention

Specificity or selectivity asks whether relevant interfering components prevent the intended measurement. A method that reads the sum of two overlapping components may not quantify the target separately. Ask how that possibility was assessed for the material of interest.

A changed column, preparation, instrument or material can affect applicability. This does not mean every change demands the same work. It means the laboratory should assess what verification, transfer work or revalidation is needed for its conditions.

6. Practice: judge the claim rather than the label

When asking about a result, request the intended purpose, current method version, demonstrated range, relevant performance evidence and actual acceptance criteria. This basic lesson does not prescribe regulatory compliance or generic acceptance percentages for NHD materials.

Use this guide in a product enquiry

For materials such as KPV、MOTS-c, identify the exact form, quantity and research question first. Then ask what records are available for the relevant lot. Matching product names do not make different specifications interchangeable.

Discuss products and document requirements with NHD. Continue learning: Reading a chromatogram / LC–MS identity.