TSE ISO/TR 8550-3

TSE ISO/TR 8550-3 is a guidance document on the selection and use of acceptance sampling systems for lots of discrete products when acceptance decisions are based on measurable characteristics. It focuses on items that can be individually identified and separated within a lot, so the sampling logic is built around countable units and continuous-scale measurements rather than simple pass/fail inspection.

What sampling by variables means in practice

Sampling by variables uses measured data such as dimensions, mass, strength, or other continuous characteristics. Compared with attribute sampling, it can make better use of information from each inspected item because the actual measurement value carries more detail than a binary conformity result. In general technical terms, this approach is most effective when the characteristic is approximately normally distributed, or can be transformed to a distribution that closely resembles normality, since many variables-based acceptance schemes rely on statistical distribution assumptions.

Where this guidance applies

• Lots made up of discrete, individually identifiable product items
• Products with measurable characteristics on a continuous scale
• Selection and use of acceptance sampling systems, rather than a specific single plan
• Characteristics that are approximately normal, or can be transformed to behave similarly

In broader quality-control practice, variables sampling is used to connect sample size, process variation, and characteristic averages in a way that can yield more information-rich decisions than attribute-based inspection. The core idea is not that every product must be inspected, but that a properly chosen sample can represent the lot well enough for acceptance purposes when the product and its measurements fit the assumptions of the method. This part provides guidance in that setting rather than prescribing one universal rule for all products.