A Joint Spatial Factor Analysis Model to Accommodate Data from Misaligned Nested Areal Units with Application to Louisiana Social Vulnerability

• Activity Number: 283 - GSS/SSS/SRMS Student Paper Award Presentations
• Type: Topic Contributed
• Date/Time: Tuesday, August 1, 2017 : 8:30 AM to 10:20 AM
• Sponsor: Social Statistics Section
• Author(s): Rachel Nethery* and Dale P. Sandler and Shanshan Zhao and Lawrence S. Engel and Richard K. Kwok
• Companies: and National Institute of Environmental Health Sciences and National Institute of Environmental Health Sciences and University of North Carolina at Chapel Hill and National Institute of Environmental Health Sciences
• Keywords: Spatial factor analysis ; Spatial misalignment ; Social indicators ; Social vulnerability ; Climate change ; Louisiana

Abstract

With the threat of climate change looming, identification of communities at the highest risk of devastation based not only on geographic features but also on social characteristics is imperative. Indices of community social vulnerability can be created by summarizing large sets of social indicator data using spatial factor analysis, assuming all variables arise from a common set of spatial units; however, current spatial factor analysis methodology is ill-equipped to handle spatially misaligned data. In this paper, we introduce a joint spatial factor analysis model that can accommodate spatial data from multiple levels of nested areal units and identify a common set of latent factors arising from the smallest areal units represented in the data. This model provides a statistically sound method for summarizing spatially misaligned data and allows for results and inference at the highest possible spatial resolution. We demonstrate the superiority of this model over competing methods and use it to construct an index of community social vulnerability in Louisiana based on a set of misaligned social indicator data. This social vulnerability index is shown to improve on an existing one.