E Annex E: Considerations regarding measurement uncertainty in the context of ISO/IEC 17025

What this heading covers

The three headings are specific: general, characteristics of software quality measurements, and applicability of measurement uncertainty. That sequence suggests the annex first describes what makes a software quality measurement different, then asks when an uncertainty statement applies. It matters because a laboratory accredited to ISO/IEC 17025 that reports a software quality measure is making a statement about the competence of its measurement, not only about the product. The annex is informative and it does not turn the draft into an accreditation standard. A team outside a laboratory can still use it to think about the confidence it has in a number.

Record position: annex 5 of 5. The heading is E in the public contents; its text is informative and not public.

What the heading asks of a team

What to put in place before publication

1. If a laboratory reports the measure, record the method, the population and the collection conditions as part of the result.
2. Write a short uncertainty note for each measure: what could vary and how much it would move the result.
3. Separate definitional uncertainty, which comes from the measure, from collection uncertainty, which comes from the data.
4. State whether a comparison between two products is robust to the uncertainty you have described.
5. Keep the annex on the reading list for the published edition, because the wording will matter to an accredited lab.

Where teams usually fall short

• A quality measure reported to three decimal places with no statement of how it was collected.
• Two products compared on a difference smaller than the variation in the collection method.

What happens next

That is the last annex. The model page returns to the nine characteristics; the what-changes page shows what moved between the 2016 edition and this draft.