XXI. MEZINÁRODNÍ KOLOKVIUM O REGIONÁLNÍCH VĚDÁCH. SBORNÍK
PŘÍSPĚVKŮ.
21ST
INTERNATIONAL COLLOQUIUM ON REGIONAL SCIENCES.
CONFERENCE PROCEEDINGS.
Place: Kurdějov (Czech Republic)
June 13-15, 2018
Publisher: Masarykova univerzita, Brno
Edited by:
Viktorie KLÍMOVÁ
Vladimír ŽÍTEK
(Masarykova univerzita / Masaryk University, Czech Republic)
Vzor citace / Citation example:
AUTOR, A. Název článku. In Klímová, V., Žítek, V. (eds.) XXI. mezinárodní kolokvium
o regionálních vědách. Sborník příspěvků. Brno: Masarykova univerzita, 2018. s. 1–5.
ISBN 978-80-210-8969-3.
AUTHOR, A. Title of paper. In Klímová, V., Žítek, V. (eds.) 21st
International Colloquium
on Regional Sciences. Conference Proceedings. Brno: Masarykova univerzita, 2018. pp. 1–
5. ISBN 978-80-210-8969-3.
Publikace neprošla jazykovou úpravou. / Publication is not a subject of language check.
Za správnost obsahu a originalitu výzkumu zodpovídají autoři. / Authors are fully responsible for the
content and originality of the articles.
© 2018 Masarykova univerzita
ISBN 978-80-210-8969-3
ISBN 978-80-210-8970-9 (online : pdf)
Sborník příspěvků XXI. mezinárodní kolokvium o regionálních vědách Kurdějov 13.–15. 6. 2018
714
DOI: 10.5817/CZ.MUNI.P210-8970-2018-94
SPATIAL CORRELATIONS IN DEMOGRAPHIC
PHENOMENA ON THE EXAMPLE OF THE REGION OF
WARMIAAND MAZURY IN POLAND
AGNIESZKA SZCZEPAŃSKA
Department of Real Estate Resources
Faculty of Geodesy, Geospatial and Civil Engineering
University of Warmia and Mazury in Olsztyn
 Prawocheńskiego 15, 10-724 Olsztyn, Poland
e-mail: aszczep@uwm.edu.pl
Annotation
Demographic potential is a significant determinant of the development of spatial units. Quantitative and
qualitative descriptions of analyses potential provide vital inputs for development and management at different
levels of governance. Methods that support the identification of spatial correlations play an important role in
analyses of demographic processes. Spatial statistics are used to evaluate and compare various phenomena based
on spatial correlations (autocorrelations). This article analyzes demographic changes (population growth) in the
Polish Region of Warmia and Mazury with the use of local Moran’s I. This statistic supported the identification
of clusters of homogeneous areas and outliers in the studied region. Two homogeneous areas were identified: the
Olsztyn Functional Urban Area with the city of Ełk and local instabilities in the proximity of Elbląg and Szczytno
urban areas. The analysis did not reveal significant spatial correlations in demographic growth between small
urban areas in the Region of Warmia and Mazury and their surroundings.
Key words
region, population growth, spatial correlation, Moran’s I
JEL classification: J1, C1
1. Introduction
Demographic potential is one of the prime movers behind development at the central, regional and local level. It
determines the development of spatial units and unleashes the productive potential of human resources.
Demographic potential is a measure of human resource development through the revival of the local population
or the inflow of migrants. Effective development policies at various levels of territorial governance cannot be
implemented without reliable demographic data (Sobczyk, 2015). The quantitative and qualitative aspects of
demographic trends have to be analyzed to promote local development, identify sources of untapped potential and
the main barriers to growth, and facilitate the implementation of administrative measures to prevent adverse social
and economic phenomena (Kołodziejczyk, 2002).
According to Weeks (2004, p. 381), “Demography is an inherently spatial science since it almost always deals
with human populations in a defined geographic region, but spatial analysis has thus far played only a small role
in the development and testing of demographic theory”. Our understanding of demographic phenomena has been
expanded by the progress in information technology, including spatial information systems. Spatial methods play
an increasingly important role in demographic analyses because demographic phenomena are characterized by
spatial trends and autocorrelation. These correlations are investigated with the use of statistical methods to analyze
the studied phenomena in view of spatial relationships (Zhang, Ling, 2007). Local and global Moran’s I are one
of the methods for visualizing demographic phenomena. These statistics are frequently used to identify the
presence of autocorrelations between the attribute values of spatial objects in various research areas, including in
demographic studies at different scales (Carrijo, da Silva, 2017; Miśkiewicz-Nawrocka, Zeug-Żebro, 2017;
Laskowska, Dańska-Borsiak, 2016; Lee, Li, 2017; D’Aubigny, 2016; Shen et al., 2016; Podgrodzka, 2014; Chen,
2013; Chi, Zhu, 2008; Tiefelsdorf, 2002; Assuncao, Reis, 1999; Tiefelsdorf , Boots, 1997).
Sborník příspěvků XXI. mezinárodní kolokvium o regionálních vědách Kurdějov 13.–15. 6. 2018
715
The goal of the study is to determine the spatial dependences in the region in the context of demographic
phenomena. In this study, local Moran’s I was used to determine the presence of correlations between the smallest
units of territorial division (municipalities) in the Region of Warmia and Mazury based on the adopted variable
of population growth. The areas characterized by spatial variations in terms of population growth were identified
in the examined region. Local Moran’s I was calculated to identify clusters of spatial units characterized by similar
values of the adopted variable and to determine the presence or absence of autocorrelations with the neighboring
units. The analyzed spatial units are situated in the Region of Warmia and Mazury in north-eastern Poland.
2. Demographic trends in Poland
Poland’s population has been declining steadily since 2012, and the country is presently facing a demographic
crisis (Sytuacja demograficzna Polski, 2016). The above is confirmed by the projections of negative population
growth developed for Poland by the United Nations World Population Prospects 2017 (Fig. 1). Negative
population growth is presently observed in most European countries.
Fig. 1: Population of Poland
Source: United Nations, Department of Economic and Social Affairs, Population Division (2017): World Population
Prospects: The 2017 Revision.
Changes in the proportions of urban and rural populations are yet another characteristic demographic trend in
Poland. The observed changes have intensified after Poland’s political transformation when vast numbers of rural
residents began migrating to urban areas. According to the World Urbanization Prospects 2014, the urban
population will continue to grow until 2050, which will lead to a steady decline in the rural population (Fig. 2).
The greatest population increase will take place in large urban centers, whereas the population of medium-sized
and small urban areas will decrease (Prognoza ludności na lata 2014-2050).
Fig. 2: Country profile – Poland. Urban and rural proportions
Source: United Nations, Department of Economic and Social Affairs, Population Division (2014): World Urbanization
Prospects: The 2014 Revision.
Sborník příspěvků XXI. mezinárodní kolokvium o regionálních vědách Kurdějov 13.–15. 6. 2018
716
2.1 Spatial variations in demographic processes in the Region of Warmia and Mazury
The demographic trends at the national level are similar to those noted at the regional and local level. The Region
of Warmia and Mazury, one of the 16 administrative regions in the country, is situated in north-eastern Poland
(Fig. 3). It spans an area of 24,173 km2
and accounts for 7.7% of Poland’s territory. The region has a population
of 1,436,367, which represents only 3.7% of the national population (Central Statistical Office, data for 2016).
Fig. 3: Location of the analyzed region
Source: own elaboration.
Warmia and Mazury is one of the poorest regions in Poland and the EU, and it is characterized by spatial and
developmental peripherality. The predominance of agriculture and high unemployment rates significantly
influence the region’s demographic processes, in particular population growth and internal and external migration.
According to demographic predictions (Fig. 4), a steep and steady population decline is expected in the region in
the coming decades.
Fig. 4: Region of Warmia and Mazury: population in 1995-2016 and predicted population in 2017-2050
Source: own elaboration based on the Demographic Database of the Central Statistical Office
The analyzed region has 49 urban areas, including 2 cities with a population higher than 100,000, and two cities
with a population of 50,000-100,000. The urban areas in the region form 16 urban municipalities, and the
remaining areas have mixed urban-rural status. Most towns in the region are small and very small urban centers
with a low development potential; nonetheless, they exert an influence on their surroundings. In 2005-2016, the
Region of Warmia and Mazury had a relatively high rate of urbanization at around 59-60% (Fig. 5).
1 200 000
1 250 000
1 300 000
1 350 000
1 400 000
1 450 000
1 500 000
1995 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Totalpopulation[persons]
predictedpresent
Sborník příspěvků XXI. mezinárodní kolokvium o regionálních vědách Kurdějov 13.–15. 6. 2018
717
Fig. 5: Rate of urbanization in 2005-2016 and the predicted rate of urbanization in 2017-2050 in the Region of
Warmia and Mazury
Source: own elaboration based the Demographic Database of the Central Statistical Office and the Population Forecast for
2014-2050
Despite its predominantly agricultural character, the region had a high urbanization rate of 59.03% in 2016 relative
to the national average of 60.18%. Population growth in urban areas peaked in 2007, and it continued to decrease
in the following years. In the analyzed period, urban areas had a negative net migration rate until 2011, but this
trend has been reversed as of 2012. The spatial distribution of the region’s population is presented in Figure 6.
Fig. 6: Spatial distribution of the population in the Region of Warmia and Mazury in 2016.
Source: own elaboration based on the Local Data Bank of the Central Statistical Office
According to the report entitled “The demographic situation of Poland in 2016”, suburban areas in the vicinity of
large cities (regional capitals, largest sub-regional centers) are the only areas to experience steady population
increase, which contributes to growing levels of suburbanization. Population growth was also observed in the
areas surrounding Olsztyn, the capital city of the Region of Warmia and Mazury. In contrast, the vast majority of
municipalities in the region, including urban and urban-rural municipalities, were characterized by significant
population decline in 2002-2016 (Fig. 7). The highest population increase was noted around, rather than within
urban centers. The only exception was the city of Ełk which reported the highest population growth in the region.
The greatest population decrease in the urban core with a simultaneous population increase in suburban areas were
observed in the region’s second largest city, Elbląg, and in the city of Szczytno. The city of Olsztyn exerts the
greatest influence on the surrounding areas on account of its administrative functions and socioeconomic potential.
56,00
56,50
57,00
57,50
58,00
58,50
59,00
59,50
60,00
60,50
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
[%] present predicted
Sborník příspěvků XXI. mezinárodní kolokvium o regionálních vědách Kurdějov 13.–15. 6. 2018
718
Fig. 7: Population increase in the municipalities of the Region of Warmia and Mazury in 2002-2016
Source: own elaboration based on the Local Data Bank of the Central Statistical Office
3. Methodology and Results
The spatial correlations in demographic trends between the municipalities of the Region of Warmia and Mazury
were determined with the use of spatial statistics. This approach supports analyses of autocorrelations, namely the
spatial relationship of variable values in the evaluated area. Spatial autocorrelations exist when the presence of a
given phenomenon in one spatial unit increases or the decreases the probability of that phenomenon occurring in
the neighboring units (Janc 2006). Spatial statistics are used to identify the patterns of spatial correlation and
spatial heterogeneity. Global and local measures of spatial autocorrelation are applied for this purpose (Antczak,
Lewandowska-Gwarda, 2016; Janc, 2006). A spatially and temporary non-homogeneous system is characterized
by different values across locations or time intervals (Radło-Kulisiewicz, 2015).
Spatial statistics relies on measures of global and local autocorrelation. Global autocorrelation demonstrates
similarities between areas, whereas local autocorrelation provides information on the proximity of the analyzed
spatial unit and indicates whether the neighboring spatial units are characterized by high or low values of the
evaluated variable.
The local indicator of spatial association (LISA) statistic developed by Anselin (1995) is a measure of local
autocorrelation, and Local Moran’s I is one of such measures. Local Moran’s I is widely used to test global spatial
correlations. According to Anselin, “the local statistic should give an indication of outstanding clustering in either
positive or negative spatially distributed regression residuals, and it should point to significant spatial outliers in
the regression residuals that do not fit into their surrounding environment” (cited by Tiefelsdorf, Boots, 1997, p.
248). Positive values of local Moran’s I indicate similarities between clusters of neighboring spatial units. Spatial
units that are regularly distributed in space and do not form clusters as well as neighboring units with different
variable values produce negative values of local Moran’s I. In the absence of correlations between neighboring
units, the expected value of the discussed statistic approximates zero (Antczak, Lewandowska-Gwarda, 2016;
Janc, 2006).
Local autocorrelations are evaluated and interpreted with the use of scatter plots and maps presenting the
distribution of spatial units in scatter plot quartiles based on the values of the taxonomic indicator: H-H – high
value unit with high value neighbors (1st
quartile), L-H – low value unit with high value neighbors (2nd
quartile),
L-L – low value unit with low value neighbors (3rd
quartile), H-L – high value unit with low value neighbors (4th
quartile). Local Moran’s I was calculated based on formula 1 (Więckowska, 2015 citing Anselin, 1995):
Sborník příspěvků XXI. mezinárodní kolokvium o regionálních vědách Kurdějov 13.–15. 6. 2018
719
=
( ̅) ∑ ( ̅) (1)
where:
n – number of spatial units,
, – value of the variable for similar objects,
̅ – average value of the variable for all objects,
– elements of the spatial weights matrix,
– variance =
∑ ( ̅)
.
A spatial weights matrix with row standardization to 1 was developed based on the boundary lines shared by
spatial units.
The results of the analysis are presented in the cluster map of local Moran’s I statistic (Fig. 8). Spatial correlations
between the values of the analyzed variable were noted only in some municipalities in the evaluated region. Most
municipalities were not bound by spatial correlations. The municipalities where the local Moran’s I was equal to
0 are marked in white in the map.
Fig. 8: Cluster map of local Moran’s I statistic for the Region of Warmia and Mazury
– population growth in 2002–2016.
Source: own elaboration
In an analysis of population growth, the values of local Moran’s I reveal two H-H clusters in the analyzed region.
H-H clusters denote municipalities with high population growth, which are surrounded by similar spatial units
(marked in red). The city of Olsztyn and the surrounding areas (6 rural municipalities) are clearly predominant in
the map. The city of Ełk forms a similar cluster. H-L and L-H clusters (outliers) were identified in the rural
municipality of Szczytno which surrounds the city of Szczytno (high population growth in rural areas surrounding
the city with one of the highest negative rates of population growth in the city of Szczytno). The city of Elbląg
also forms an L-H cluster, but population growth in Elbląg’s suburban areas was smaller than in Szczytno, whereas
the city of Elbląg was also characterized by one of the highest negative rates of population growth. The discussed
clusters are statistically significant, and they contribute to local instabilities. Spatial autocorrelations were not
determined in the remaining municipalities in the region, which indicates that despite a high number of urban
municipalities (urban centers) and urban-rural municipalities, these spatial units does not influence the surrounding
areas in terms of their demographic potential.
Conclusions
The cluster map of Moran’s I statistic presents spatial autocorrelations in demographic trends between the
municipalities of the Region of Warmia and Mazury. Local instabilities were observed in the municipalities
situated in the proximity of the cities of Elbląg and Szczytno. This is a highly undesirable situation for Elbląg, the
second largest urban center in the region which should drive local growth. The Olsztyn Functional Urban Area is
Sborník příspěvků XXI. mezinárodní kolokvium o regionálních vědách Kurdějov 13.–15. 6. 2018
720
composed of a regional urban center and the surrounding municipalities, and it creates a homogeneous spatial unit
with considerable variations in demographic potential. Similar observations were made in Ełk, one of the largest
urban centers in the region. The remaining municipalities were not bound by spatial correlations with the
neighboring units in terms of population growth. The above applies particularly to urban municipalities which do
not have sufficient demographic potential to drive positive population change in their immediate surroundings.
For this reason, spatial correlations were not noted in their sphere of influence.
The results of the study could point to a non-continuous development of demographic potential in the analyzed
region, which increases the existing disproportions between the city of Olsztyn, its immediate surroundings
(accumulation of demographic potential) and other municipalities. Regional management solutions could benefit
from the identification of areas where clusters of municipalities are formed based on the values of local Moran’s
I. Clusters of correlated spatial units (municipalities) can be used to identify areas that are homogeneous in terms
of the analyzed criteria as well as areas of local instability (outliers) which require special attention. Analyses of
spatial trends and correlations provide valuable inputs for regional policy.
Accumulation of demographic potential surroundings the capital of the region and another large cities is observed
in relation to other Polish regions (Kłusek, 2017; Kozera, Głowicka-Wołoszyn, 2017; Gołata, Kuropka, 2016;
Bartkowiak-Bakun, Standar, 2014; Kołodziejczyk, 2002). Spatial relationships are the strongest in urban
functional areas.
Literature
[1] ANSELIN, L., (1995), Local Indicators of Spatial Association – LISA. Geographical Analysis, vol. 27, no.2,
pp. 93–115. ISSN 1538-4632. DOI: 10.1111/j.1538-4632.1995.tb00338.x.
[2] ANTCZAK, E., LEWANDOWSKA-GWARDA, K., (2016). Zależności przestrzenne w kształtowaniu się
procesów migracyjnych w wybranych miastach Europy. In Dańska-Borsiak, B., Laskowska, I. (eds.)
Regionalne analizy ekonomiczne, Wydawnictwo Uniwersytetu Łódzkiego, Łódź, pp. 27-42. DOI:
10.18778/8088-052-8.02. ISBN 978-83-8088-051-1.
[3] ASSUNCAO, R. M., REIS, E. A., (1999). A new proposal to adjust Moran's I for population density.
Statistics in medicine, vol. 18, no. 16, pp. 2147-2162. ISSN 1097-0258. DOI: 10.1002/(SICI)1097-
0258(19990830)18:16<2147::AID-SIM179>3.0.CO;2-I.
[4] BARTKOWIAK-BAKUN, N., STANDAR, A., (2014). Zróżnicowanie sytuacji demograficznej obszarów
wiejskich pogranicza zachodniego. Roczniki Naukowe Stowarzyszenia Ekonomistów Rolnictwa i
Agrobiznesu, vol. 16, no. 6, pp. 27-31. ISSN 1508-3535.
[5] LOCAL DATA BANK OF THE CENTRAL STATISTICAL OFFICE [online]. [cit. 2018-02-11].
Accessible: https://bdl.stat.gov.pl/BDL/start.
[6] DEMOGRAPHIC DATABASE OF THE CENTRAL STATISTICAL OFFICE [online]. [cit. 2018-02-11].
Accessible: http://demografia.stat.gov.pl/bazademografia/.
[7] CARRIJO, T. B., DA SILVA, A. R., (2017). Modified Moran's I for Small Samples. Geographical Analysis,
vol. 49, no. 4, pp. 451-467. ISSN 1538-4632 . DOI: 10.1111/gean.12130.
[8] CHEN, Y., (2013). New approaches for calculating Moran’s index of spatial autocorrelation. Plos One, vol.
8, no. 7, e68336. DOI:10.1371/journal.pone.0068336.
[9] CHI, G., ZHU, J., (2008). Spatial regression models for demographic analysis. Population Research and
Policy Review, vol. 27, no. 1, pp. 17-49. ISSN 0167-5923. DOI: 10.1007/s11113-007-9051-8.
[10]D’AUBIGNY, G. (2016). A Statistical Toolbox For Mining And Modeling Spatial Data. Comparative
Economic Research, vol. 19, no. 5, pp. 5-24. ISSN 2082-6737 . DOI: 10.1515/cer-2016-0035.
[11]GOŁATA, E., KUROPKA, I., (2016). Large cities in Poland in face of demographic changes. Bulletin of
Geography. Socio-economic Series, vol. 34, no. 34, pp. 17-31.
[12]JANC K., (2006). Zjawisko autokorelacji przestrzennej na przykładzie statystyki I Morana oraz lokalnych
wskaźników zależności przestrzennej (LISA) – wybrane zagadnienia metodyczne. In Komornicki ,T.,
Podgórski, Z. (eds.), Idee i praktyczny uniwersalizm geografii, Dokumentacja Geograficzna, no. 33, IGiPZ
PAN, Warszawa, pp. 76-83. ISBN 83-87954-73-X.
[13]KŁUSEK, M.. (2017). Potencjał demograficzny a poziom rozwoju gospodarczego podregionów w Polsce w
okresie 2004–2012 Folia Oeconomica Acta Universitas Lodzensis, vol. 327, no. 1, pp. 231-244. ISSN 0208-
6018. DOI: 10.18778/0208-6018.327.14.
[14]KOŁODZIEJCZYK, D.. (2002). Ocena potencjału demograficznego gmin w Polsce (w aspekcie ilościowym
i jakościowym. Studia Obszarów Wiejskich, vol. 2, pp. 19-30. ISSN 1642-4689.
[15]KOZERA, A. B., GŁOWICKA-WOŁOSZYN, R.., (2017). Spatial Autocorrelation of Communes’ Income
Potential in Selected Metropolitan Areas. Folia Oeconomica Acta Universitas Lodzensis, vol. 4, pp. 155-168.
ISSN 0208-6018. DOI: 10.18778/0208-6018.330.11.
Sborník příspěvků XXI. mezinárodní kolokvium o regionálních vědách Kurdějov 13.–15. 6. 2018
721
[16]LASKOWSKA, I., DAŃSKA-BORSIAK, B., (2016). The Importance Of Human Capital For The Economic
Development Of EU Regions. Comparative Economic Research, vol. 19, no. 5, pp. 63-79. ISSN 2082-6737.
DOI: 10.1515/cer-2016-0038.
[17]LEE, J., LI, S., (2017). Extending Moran's index for measuring spatiotemporal clustering of geographic
events. Geographical Analysis, vol. 49, no. 1, pp. 36-57. ISSN 1538-4632 . DOI: 10.1111/gean.12106.
[18]MIŚKIEWICZ-NAWROCKA, M., ZEUG-ŻEBRO, K.. (2017). Badanie zróżnicowania sytuacji
demograficznej w Polsce w latach 2005-2014. Studia Ekonomiczne, vol. 318, pp. 69-83. ISSN 2083-8611.
[19]PODOGRODZKA, M., (2014). Przestrzenne zróżnicowanie ludności według wieku w Polsce w latach 1991-
2010. In RĄCZASZEK, A., KOCZUR, W. (eds..), Polityka społeczna wobec przemian demograficznych,
Studia Ekonomiczne, vol. 167, pp. 62-76. ISSN 2083-8611.
[20]PROGNOZA LUDNOŚCI NA LATA 2014-2050. 2014. Główny Urząd Statystyczny, Warszawa.
[21]RADŁO-KULISIEWICZ, M., (2015). Przegląd wybranych podejść w zakresie prognozowania rozwoju
obszarów miast. Archiwum Fotogrametrii, Kartografii i Teledetekcji, vol. 27, pp. 109-122. ISSN 2083-2214.
DOI: 10.14681/afkit.2015.008.
[22]SHEN, C., LI, C., SI, Y., (2016). Spatio-temporal autocorrelation measures for nonstationary series: A new
temporally detrended spatio-temporal Moran's index. Physics Letters A, vol. 380, no.1-2, pp. 106-116. ISSN
0375-9601. DOI: 10.1016/j.physleta.2015.09.039.
[23]SOBCZYK, A., (2015). Ocena potencjału demograficznego szczecińskiego obszaru metropolitalnego z
zastosowaniem zmiennej syntetycznej. Folia Pomeranae Universitatis Technologiae Stetinensis.
Oeconomica, vol. 317, no. 78, pp. 81-92. ISSN 2081-0644. DOI: 10.21005/oe.2016.85.4.15.
[24]Sytuacja demograficzna Polski. Raport 2015-2016. (2016). Zakład Wydawnictw Statystycznych, Warszawa.
[25]TIEFELSDORF, M., (1998). Some practical applications of Moran's I's exact conditional distribution. Papers
in Regional Science, vol. 77, no. 2, pp. 101-129. ISSN 1435-5957. DOI: 10.1111/j.1435-
5597.1998.tb00710.x.
[26]TIEFELSDORF, M., BOOTS, B., (1997). A note on the extremities of local Moran's I(i)s and their impact on
global Moran's I. Geographical Analysis, vol. 29, no. 3, pp. 248-257. ISSN 1538-4632 . DOI: 10.1111/j.1538-
4632.1997.tb00960.x.
[27]UNITED NATIONS, DEPARTMENT OF ECONOMIC AND SOCIAL AFFAIRS, Population Division
(2017): World Population Prospects: The 2017 Revision. [online]. [cit. 2018-02-10]. Accessible:
https://esa.un.org/unpd/wpp/.
[28]UNITED NATIONS, DEPARTMENT OF ECONOMIC AND SOCIAL AFFAIRS, Population Division
(2014): World Urbanization Prospects: The 2014 Revision. [online]. [cit. 2018-02-10]. Accessible:
https://esa.un.org/unpd/wup/.
[29]WIĘCKOWSKA, B., (2018), Podręcznik Użytkownika PQStat. Analiza przestrzenna. PQSTAT
SOFTWARE. [online]. [cit. 2018-01-15]. Accessible: http://download.pqstat.pl/DokumentacjaGeo.pdf .
[30]WEEKS, J. R.. (2004). The role of spatial analysis in demographic research. In Gootchild M. F., Janelle D.G.
(eds.) Spatially integrated social science. Oxford University Press, New York, USA, pp. 381-399. ISBN 0-
19-515270-0.
[31]ZHANG, T., LIN, G., (2007). A decomposition of Moran's I for clustering detection. Computational Statistics
& Data Analysis, vol. 51, no. 12, pp. 6123-6137. ISSN 0167-9473. DOI: 10.1016/j.csda.2006.12.032.