The aim of the article is to present a non-parametric kernel density estimation method as a tool used for empirical verification of the regional
convergence hypothesis, including convergence of clubs. It is explained how kernel density estimation complements other methods applied to verify the phenomenon of convergence. The empirical part shows an application of the non-parametric density estimation to the analysis of regional convergence of educational achievements of Polish pupils, measured by the average results of the mathematical and natural science part of the lower-secondary school leaving exams on the level of municipalities in years 2002—2013. The results of the analysis indicate the existence of regional convergence of exam results for Polish municipalities. In case of the analysis for three-yearly periods convergence of clubs was observed — the municipalities with lowest exam results constitute a separate club of convergence.
A Non-Parametric Method to Determine Basic Probability Assignment Based on Kernel Density Estimation