This standard specifies the methods necessary for the following statistical treatment:a) Determining the confidence interval for the mean (expected value) or variance of a random variable.b) Hypothesizing about the value of the mean or variance of a random variable based on a sample. These methods are valid only when the measurements are considered to have been obtained from independent observations. In the case of finite populations, they can be applied when the data can be considered independent. This holds true if the sample size is sufficiently small compared to the finite population. As a rule of thumb, it is desirable for the sample size to be less than 1/10 of the population size. In the application of these methods, it is assumed that the random variable follows a normal distribution. If this assumption can be verified from extensive experience with the target data group, the random variable can be considered to follow a normal distribution. The method of verifying that a random variable follows a normal distribution follows KS Q 5002 or KS Q ISO 5479. In addition, there are cases where the hypothesis of normality is acknowledged based on external information rather than a sample. If the hypothesis of normality is rejected, it is recommended not to use this standard method. However, even if the hypothesis of normality is not rejected, there is no guarantee that the random variable follows a normal distribution. Using probability paper, visual (graphical) examination of normality is possible, and if the deviation from the normal distribution is large, non-normality is clear.
KS A ISO 2854:2013 history
2023KS A ISO 2854-2023 Statistical interpretation of data, estimation and testing methods of mean and variance
0000 KS A ISO 2854-2013(2018)
2013KS A ISO 2854:2013 Statistical interpretation of data－Techniques of estimation and tests relating to means and variances
2008KS A ISO 2854:2008 Statistical interpretation of data-Techniques of estimation and tests relating to means and variances