Mirosław Belej, Radosław Cellmer

Methodology for identification of homogeneous group of housing markets in urban development

View full text in PDF format

Abstract

Knowledge of the real estate market is crucial for good land management. The question of similarities between local markets is connected with the question of internal variability of the individual markets. We present the results of a project to compare a similarities between local markets in the quantitative and qualitative aspects by creating appropriate integrated indicators for these markets. Such indicators will make it possible to find a similarity of different local markets and present them hierarchically. From the large number of methods available, techniques originating from two spheres of statistical inference were selected: distribution free methods and multivariate analysis methods. For analyses within the distribution free methods, Friedman's test (Friedman's non-parametric variance analysis) and Kendall's test (W. Kendall's compatibility ratio) were selected. For analyses within the multivariate analysis methods, factor analysis was selected, which constitutes a set of statistical methods and procedures allowing for conversion of a large number of variables tested to a significantly smaller number of independent factors or main components which preserve a considerable amount of information conveyed in the primary variables. If this statistical method is applied, it is possible to reduce the number of attributes describing local real estate markets down to a few integrated indicators, which preserve their descriptive qualities, despite the reduction of variables that has taken place.

Keywords:similarity; housing market; land management.