Application of non-parametric methods to enhance statistical accuracy within Uzbekistan’s national data system
Annotatsiya
This article investigates the role and practical significance of non-parametric tests in modern statistical analysis. The relevance of the research is determined by the necessity of enhancing the accuracy of data and the quality of hypothesis testing during the alignment of the national statistical system with international standards. The article analyzes the limitations of traditional parametric methods, specifically their shortcomings in cases where data does not follow a normal distribution. The advantages of non-parametric tests (Mann-Whitney-Wilcoxon test, Wilcoxon Signed-Rank test, Sign test, Kruskal-Wallis Test, Spearman rankcorrelation coefficient) are scientifically substantiated, highlighting their flexibility regarding sample size and distribution shape. Furthermore, the effectiveness of these methods in identifying intergroup differences based on real-world statistical data is demonstrated. The findings serve to improve analytical processes within national statistical practices and ensure the reliability of forecasts.