prøvestørrelse

Prøvestørrelse is a term used in Norwegian statistics and quality management to denote the size of a sample drawn from a population for testing, measurement, or analysis. It represents how many units are selected for inspection or assessment and is a central parameter in planning experiments, surveys, and quality-control processes. In practice, it corresponds to the English concept of sample size.

In statistics and research, the prøvestørrelse influences the precision of estimates and the power of statistical tests. It is chosen based on the desired margin of error, confidence level, and the expected variability of the measured characteristic. Larger samples yield more precise estimates but require more time and resources.

In quality control and manufacturing, a prøvestørrelse refers to the number of items sampled from a batch for inspection under a sampling plan. The plan also specifies rules for accepting or rejecting the lot, often using criteria such as the number of defects allowed. The choice aims to balance risk for producers and customers and is affected by the lot size and the acceptable quality level (AQL).

Determination of the prøvestørrelse typically involves statistical considerations. For estimating a proportion, a common approximation is n ≈ Z^2 p(1-p) / E^2, with finite population correction n' ≈ n / (1 + (n-1)/N). When p is unknown, a conservative choice is p ≈ 0.5. Practical decisions also account for cost, time, and survey design, including stratification or sequential sampling.