THE STATISTICAL ANALYSIS OF POLISH FOOD ENTERPRISES: - NONPARAMETRIC APPROACH.

Statistical analysis of Polish food enterprises is done to present economic situation of agro-food industry in Poland. In analysis, nonparametric approach is chosen as effective and simple method of studying variables in populations. This approach is widely used, especially when additional information about regarded variable is not available (as often happens in economic researches). Two nonparametric estimation methods are taken into consideration: kernel density estimation and bootstrap confidence interval. The special emphasis is taken on choosing the proper values of parameters in kernel density estimation and choosing the most effective bootstrap interval among these presented in literature. The study concerns applying basic descriptive statistics and nonparametric estimation of number of employees and revenues total of Polish food enterprises, using kernel method for estimating the density function and bootstrap confidence interval for median of regarded variable. Results and conclusions from the study can be useful for the users of nonparametric methods in economic researches. The main research aim of the paper is to present and examine some statistical procedures that can be used in the analysis of economic situation of chosen enterprises connected strictly with food production. The good properties of regarded methods allow compering some regions of country to indicate these regions where there are friendly conditions for the food production enterprises, including the natural character of region (rural or urban area). [ABSTRACT FROM AUTHOR]