The trajectories and challenges of the development of inner peripheries in the new conditions of cohesion post SARS-CoV-2

4. Research methodology (the way of research conducting, research methods, analysis methods, research equipment and apparatus)

In the research project, we plan to gather empirical material of primary and secondary nature and to make use of quantitative and qualitative analysis methods to ensure that in the course of the research, the theory and knowledge accumulated by other researchers and in qualitative surveys are used in quantitative research of descriptive, explanatory and verifying nature (Morse and Niehaus, 2009). Stage one and three described above will be based on quantitative research where secondary and primary sources of data will be used. At stage two, we will also employ the qualitative methods of collecting and analysing information. Below we have presented in detail the research methodology of three stages of the project:

Delimitation and typology of inner peripheries

At this stage of the analysis, we take into account the operating spatial units in the form of urban regions defined originally by means of a graph approach to identifying nodal regions (Nystuen and Dacey, 1961; Śleszyński, 2014) on the basis of a commune-based commuting matrix for 2016 created by the Central Statistical Office based on data from registers of personal income tax (GUS, 2019). For each spatial unit, we will define the values of the peripherality indicators; at this stage, selection thereof will be based on the theory and experiences of previous surveys.

Taking into consideration the various aspects of peripherality necessitates three sources of data: secondary data from latest public statistics provided by the Central Statistical Office (data from the Bank of Local Data on the demographic structure and migrations, entrepreneurship, education, building construction intensity and accommodation resources). The other source is aggregated data on labour and the locals’ remuneration, financial performance of companies and commune self-governments. These unpublished data, obtained and aggregated on the basis of taxpayers registers in the Ministry of Finance, will be made available to us by the Association of Polish Cities. The third source will be an original base of geospatial data based on open and public data (OpenStreetMaps) combined with statistical data provided by the Central Statistical Office and devoted to multi-faceted spatial accessibility of services and jobs, based on a methodology based on the source rather than the destination, e.g. the accumulated opportunities model (Páez, Darren, and Morency, 2012).

A detailed list of the used indicators will result from the selection procedures and verification of their theoretical and methodological suitability. Bearing in mind the multi-faceted nature of peripherality, selection of the indicators will be followed by a synthesis of their values by the methods of data dimensionality reduction (exploratory factor analysis) and both non-hierarchical and hierarchical methods of classification (k-means, Stevens, 2009).

Studies of selected inner peripheries

At this stage, primary data will be used on a larger scale coupled with qualitative data based on the case study methodology (Cresswell and Poth, 2017).

On top of the quantitative data obtained in the first part, we expect making use of three subsequent sources of information. The first source will include unpublished and topical financial data on employment (forms of employment, remuneration) and business activities and commune finance (including investment) obtained from the Association of Polish Cities. This information, considerably exceeding the data provided by the Central Statistical Office, will allow to evaluate the recent changes to the socio-economic situation by taking into account aspects of inequality, limited use of labour resources and economic safety the measurement of which is of special importance to the evaluation of the consequences of crises (Stiglitz, Fitoussi and Durand, 2018). The goal of this stage will be to provide a possibly most precise description of the dynamics of the socio-economic situation in a specific area rather than generalizing the knowledge of other areas. For this reason, statistical analysis methods will not be used except for descriptive statistics.

The contextual knowledge on the dynamics of issues related to the peripheral status of the regions under consideration and the response to them will be provided by representatives of various groups of local stakeholders: representatives of local authorities (mostly communal), entrepreneurs operating in a specific area and the local residents. In each of the surveyed regions, a focus group interview will be conducted with representatives of these groups selected for a specific purpose (Kvale, 2010). Transcribed recordings of the interviews will be coded. In order to describe quantitatively the significance of the topics identified in the interviews, the last method of collecting data in this part of the project will be used – online questionnaires distributed among entrepreneurs and representatives of commune self-governments and entrepreneurs in the region under survey. The research samples will be selected for a specific purpose and will not be representative; therefore their statistical presentation will be delivered in the form of descriptive statistics.

Dynamic research into the changes to inner peripheries in the time of crisis

In the last stage of the research we will draw on the data obtained at stage one, updated in the course of the research and supplemented by data from the Census of 2021 (provided that it will have been conducted and its results will have been published without delay against the plans). However, unlike at stage one, on top of the statistical methods used to describe the phenomena at hand (including the typology and delimitation), this stage will be of explanatory and prognostic nature. Therefore, we will use correlation and regression methods (including difference in differences), both aspatial and presented in a way considering spatial regularities i.e. global space autocorrelation – Moran’s index (Suchecki, 2010) and Local Indicators of Spatial Association (LISA) (Ansellin, 1995) as well as geographically weighted regression methods (Bivand, Pobesma and Gómez-Rubio, 2013). The features referring to the level and dynamics of socio-economic development will serve to compile a ranking list of the objects under scrutiny on the basis of the calculated synthetic measure by means of TOPSIS. This method is one of the so-called model methods of linear ordering of objects. Developing the Synthetic Measure of Growth will allow us, in particular, to analyse the developmental distance of specific units in relation with the model, sigma- and beta-type convergence processes in the analysed units and occurrence of spatial autocorrelation. An analysis of the relations between the variability coefficient of the Synthetic Measure of Growth and Moran’s I index will prove useful in identifying the type of development diffusion between cities and inner peripheries (Smętkowski, 2015).