One of the most important factors influencing price and its effect on the value of real estate is location, which can be considered on both global and local scales. The result is that prices in a given location should be similar and they should show spatial auto-correlation. This paper deals with the essence of spatial autocorrelation and introduces a methodology of measuring spatial dependences. The results of the research concern housing prices in the vicinity of Olsztyn. For research purposes, global and local Morans and Getis-Ord indexes were used along with an analysis of semivariograms. In this way, both the spatial autocorrelation and spatial differentiation of dependencies between housing prices were determined.

51-63