TS 2631

TS 2631 addresses how a statistical tolerance interval is determined in the statistical interpretation of data. In this context, the interval may be one-sided or two-sided; its end points are called statistical tolerance limits and are also referred to as the natural limits of the process. The standard frames this terminology around a practical statistical view of data spread and process behaviour.

What a statistical tolerance interval means

In general technical use, a tolerance interval is a statistical tool that indicates how reliably a sample-based result can represent the expected spread of a population. One-sided intervals emphasise either a lower or an upper bound, while two-sided intervals describe variation at both ends of the range. In measurement and production contexts, such intervals help interpret process behaviour, distinguish unusual departures from normal variation, and communicate variability in a disciplined way.

The conceptual distinction highlighted by the standard

The value of this standard lies in treating the limits as more than simple numerical bounds; they are statistical interpretation tools. The relationship between tolerance limits and natural limits is important when analysing a process, because both terms point to the spread of data and to process stability from related but distinct perspectives. TS 2631 brings this terminology into a common framework used in engineering and quality practice.

• Covers one-sided and two-sided statistical tolerance intervals.
• Names the interval boundaries as statistical tolerance limits.
• Also refers to these limits as the natural limits of the process.
• Provides a general framework for the statistical interpretation of data.