THE USE OF STATISTICAL CRITERIA FOR EVALUATION TEST OF DATA ON PROPERTIES OF INORGANIC SUBSTANCES

Abstract

Three statistical criteria were compared on the basis of the sampling of chemical data in the article. When the criteria were compared, the level of significance was taken as equal to 0.05, because this value is most often used in the technical calculations. Evaluation of data on the one-sided outliers of variational series was carried out by using statistical criteria. With the help of Grubbs criterion for checking variation series, outliers were found neither on the right nor on the left. Further examination of the data with the use of three-sigma rule also showed no outliers. The latest statistical criterion used to detect errors was Chauvenet's criterion. When checking the data with this criterion, one error on the right was detected. No outliers were found on the left. According to the generalized results, namely, the vote on the three most statistical criteria, it can be concluded that the variational series belongs to the same general totality. Criteria presented in the article can be applied for the analysis of any data and for making conclusions based on them, if errors were found in the sampling.