WORKING WITH DATA IN STATISTICAL RESEARCH

Анотація

Statistical data in most cases are described as an arithmetic mean value and a median depending on themselves, in the absence or presence of extreme values, respectively. Also, when using the classical hypothesis testing, errors of the type I (erroneous rejection of the null hypothesis of the absence of differences) and type II (erroneous acceptance of the null hypothesis) inevitably occur. Sometimes in the data, you can make the wrong choice when choosing the dependent and independent variable. The classical check does not give a complete picture of the data of the general population. Confidence intervals for a range of values for selection within a certain probability (95%). One of the modern methods for assessing the mean, mean and other measures is resampling, in particular, bootstrap (multiple generation of samples).