A growing body of evidence from environmental health promotion research shows that the physical environment can affect health directly or through complex interactions with people's behavior. A holistic understanding of these effects requires that we study individuals' behavior in space. Nonetheless, finding the right geographical context and translating it into a working spatial unit of analysis is far from easy. As a result, researchers have used a variety of spatial units in empirical research. Among these units, activity spaces have been especially gaining interest. Activity space models have been implemented in different ways to represent the area within which individuals travel during the course of their daily activities. Nonetheless, despite the fundamental role of activity spaces in empirical analysis, they have rarely been at the center of scholarly focus. This has contributed to a lack of conceptual and methodological consensus related to how activity spaces are defined and used in environmental health research. Recent evidence suggests that this might have resulted in uncertainties that negatively affect the validity of empirical findings. However, little is known about the involved mechanisms, their implications, and how they should be addressed in research. This dissertation takes steps toward filling these gaps. Initially, this dissertation aims to disambiguate the concept of activity space through a literature review. Following this conceptual clarification, the dissertation pursues its methodological objectives by implementing two novel individual-based activity space models. The first model, dynamic home range, is a versatile polygon representing an improved estimation of individual residential area. The second, individualized residential exposure model (IREM), is a more advanced model of activity space that incorporates a place-based estimation of exposure to provide a more refined picture of individuals' spatial behavior. This dissertation goes further by presenting an empirical framework for the measurement and use of activity space in research. While this framework shows how the relationships between activity space and personal and environmental characteristics can be multidimensionally examined, it also reveals interesting empirical findings from the Helsinki metropolitan area. These studies are based on data collected though online public participation geographical information system (PPGIS) surveys. Another contribution of this dissertation is to provide empirical evidence on how the choice of spatial units can considerably affect the results regarding the associations between the environment and health. The findings elucidate the implicit mechanisms that can affect the results and help provide guidelines for the choice of spatial units in research. Finally, this dissertation presents a set of practical geographical information system (GIS) tools that can facilitate use of advanced activity space models in environmental health research.
|Translated title of the contribution||Spatial units of analysis: are there better ways? - An empirical framework for use of individualized activity space models in environmental health promotion research|
|Publication status||Published - 2019|
|MoE publication type||G5 Doctoral dissertation (article)|
- activity space
- built environment
- spatial analysis