•  
  •  
 

Corresponding Author

Xiaoxiao Wei

Abstract

Spatial autocorrelation in model residuals can have a significant impact on the results of spatial or space-time models. This can result in misleading estimates of the influence of different factors, potentially exaggerating or even reversing the perceived effects of these factors. This study also considers the potential implications of the Modifiable Areal Unit Problem (MAUP) in the context of spatial-temporal models. In this case study for southeastern Ghana, we examined whether and how spatial autocorrelation in model residuals might generate bias in regression coefficients when explaining women’s body mass index (BMI) across urban and rural areas. Eigenvector spatial filtering, with various settings of influential zones, was systematically tested in a latent trajectory model to detect the impacts of spatial autocorrelation. We found that spatial autocorrelation in model residuals did bias the coefficients of key independent variables such as land cover type, not only affecting their magnitude but also altering their sign or significance. This highlights the risk of significantly misinterpreting the relationships between variables. The bias was effectively mitigated or reduced in urban and rural subsets identified through a data-mining approach, while it persisted in other subsets. This distinction in bias mitigation underscores the necessity of customizing models to suit specific subset attributes. Such systematic testing also enabled our choice of appropriate size of influential zones, within which spatial autocorrelation in data and model residuals was prevalent and thus accounted for biased coefficients. Additionally, we found that BMI trajectories and the associated drivers in urban areas are quite different from those in rural areas, indicating the necessity for differentiating analytical approaches between these areas. This finding therefore justifies the construction of separate BMI models for rural and urban areas. Our methodology demonstrates the importance of managing both temporal variability and spatial autocorrelation simultaneously, improving the model's usefulness in handling other space-time data.

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.