Standards Development

Scope

This part of ISO 16269 gives guidelines for the statistical analysis of one normal sample, two normal samples and of paired samples leading to the paired t-test.In the body of the standard, all calculations are illustrated with worked examples based on real experimental data, The calculations are repeated in templates designed to facilitate manual calculations. Those templates are provided in Annexes C, D, E, and F. The examples are also analysed using statistical software. The output from the statistical software is shown in Annex G.For inferences about the means of normal distributions the following tests are considered:a)      the t-test for the hypothesis that a mean has a specified value;b)      the two-sample t-test for the hypothesis that the means of two normal distributions are identical;c)       the paired t-test.For inferences about the variances of normal distributions the following tests are considered:a)      the chi-squared test for the hypothesis that a variance has a specified value;b)      the F-test for the hypothesis that the variances of two normal distributions are identical.It is recommended to use confidence intervals of the relevant parameters in the summary of the statistical analyses. This standard gives the confidence intervals for the following parameters:a)      the mean of a normal distribution;b)      the variance of a normal distribution;c)       the standard deviation of a normal distribution;d)      the difference of means of two normal distributions;e)      the ratio of the variances of two normal distributions.Power and sample size calculations are essential for the planning of statistical investigations. They are considered both in the body of the standard and in the annexes.