Modeling Community Well-Being and Cross-Level Interactions Using Functional Data

• Activity Number: 417 - Recent advancement on life time data analysis
• Type: Contributed
• Date/Time: Thursday, August 12, 2021 : 2:00 PM to 3:50 PM
• Sponsor: International Statistical Institute
• Author(s): Wlodzimierz Okrasa* and Dominik Rozkrut
• Companies: Cardinal Stefan Wyszynski University in Warsaw and Statisitcs Poland
• Keywords: community well-being; Functional Data; multilevel modelling; geospatial analysis

Abstract

Analyzing the relationships between objective and subjective well-being at community level calls for multilevel modelling to assess the influence of ‘place’, ‘locality’, ‘neighborhood’, accounting for cross-level interaction effect. This paper aims to validate the adequacy and efficiency of alternative approaches to such a task – the static, using spatial cross-sectional data, and the dynamic, using spatio-temporal / panel data (2013-16) within the Functional Data framework. Three types of measures are employed: multidimensional index of local deprivation (all communes); household income variables; and individual well-being (different surveys). The paper is structured as follows: an overview of modeling strategies is followed by exploration of spatial patterns of community and individual well-being. This is done in parallel by multivariate Functional Data approach using data for both commune and household / individual level. Multilevel modeling for both types of approaches attempts to also assess the spatial interaction effect. An evaluative comparison of alternative methodologies made in conclusions allows for suggestions on strategies suitable for analytical and policy purposes.