Two questions hidden inside good results
Are the measurements consistent? Are they appropriately close to the reference for the property? These questions overlap in a complete evaluation but are not interchangeable. The NIST terminology note distinguishes measurement terms and cautions against using them loosely.
When a report says precise, ask whether it means a particular observed spread under defined conditions. A quality adjective without a calculation, sample basis or context is not enough to interpret the evidence.
A simple example of consistent but biased results
Illustrative values only: a fictional reference is assigned 10.00 on a defined measurement basis. Results of 9.10, 9.11 and 9.09 are tightly grouped, with a mean of 9.10. The small spread does not remove the difference from 10.00.
The example does not prove an actual method has bias, and it ignores uncertainty to keep the distinction easy to see. A real evaluation needs the reference value, its uncertainty and relevant method evidence. The arithmetic simply shows why consistency alone cannot establish correctness.
| Item | Value | Interpretation |
|---|---|---|
| Assigned reference | 10.00 | Reference context requires its own evidence |
| Three repeat results | 9.10 / 9.11 / 9.09 | Small observed spread |
| Mean | 9.10 | Agreement is not the same as closeness to 10.00 |
| Real decision | Method and uncertainty review | Example alone sets no acceptance limit |
Repeating one injection is not repeating the entire workflow
Several injections from one prepared vial may examine an instrumental part of the measurement. Independent preparations can include additional sources of variation. Results obtained on another day, instrument or operator introduce another comparison context.
Ask what was repeated and what stayed fixed. A report of injection repeatability should not be relabeled whole-process reproducibility. Conversely, a wider between-day spread should be interpreted with the actual changed conditions rather than immediately blamed on the supplied peptide.
Choose the evidence needed for your question
For an amount measurement, the reviewer may need calibration, reference-value information and preparation context in addition to repeat results. For a purchasing comparison, preserve the stated measurement basis on both sides.
Do not invent a universal acceptable variation or demand an exact match to a value without considering its meaning. The responsible laboratory should evaluate the required performance. This article clarifies the terms; it does not validate a method or certify any batch.
Questions buyers ask
Does a small standard deviation prove accuracy?
No. It describes spread on the stated basis, not every source of bias or correctness.
Are three injections three independent preparations?
Not necessarily. Ask which step was repeated before interpreting the scope of the result.
Your next step
Ask what was repeated, which reference was used and how the measurement basis was defined. Keep consistency and bias questions separate.
Related reading: measurement uncertainty; report review.
AI-assisted educational guide prepared from the cited sources. No testing of a supplied batch is reported here. For research-material purchasing, not medical guidance or instructions for human use. Confirm the actual offered material and available evidence before ordering.