A dynamic approach to relative taxonomy in the assessment of changes in the social cohesion of polish provinces in 2010-2018

The classic approach to relative taxonomy was proposed by S. Wydymus (2013), and its positional version developed by J. Lira (2015). Although they are applied to panel data, both methods of relative taxonomy are static. This article presents a dynamic version of relative taxonomy. In contrast to the static approach, the dynamic approach not only shows relations between objects in individual periods but also changes in the phenomenon of interest that occurred between the objects throughout the reference period. The static version, described by Wydymus and Lira, does not account for the presence of missing (NA) values in a data array, in contrast to the proposed dynamic modification, which does. The proposed dynamic approach to relative taxonomy was used to assess the social cohesion of Polish provinces (voivodeships) in the period 2010-2018. The analysis indicates that social cohesion across the provinces systematically increased, while differences between them decreased. The rate at which social cohesion improved was faster in those provinces in which it was lower at the start of the reference period, i.e. Kujawsko-Pomorskie, Lubuskie, Warminsko-Mazurskie and Zachodniopomorskie.