Analytical Derivations of the Neighborhood Effect Averaging Problem ( NEAP ) and Its Two Variants: From Representative Samples to Unbiased Health Effect Sizes and Critical Insights Into Health Geography

ABSTRACT Accurate environmental exposure measurements that maximally mitigate contextual errors are necessary for reliable modeling of human health outcomes. Recent advances in environmental exposure measurements highlight a new manifestation of contextual errors called the neighborhood effect averaging problem (NEAP). This study conceptualizes the NEAP in terms of geostatistical sampling and provides rigorous derivations. We aimed to articulate that the NEAP can be an inevitable methodological issue for the population's environmental exposure distribution when ignoring people's daily mobility in exposure measurements. We also illustrated that using spatially clustered and unrepresentative samples of the population may lead to the counter manifestation of this methodological issue called the neighborhood effect polarization problem (NEPP). Finally, we demonstrated how mitigating the NEAP may mitigate the underestimation of health effect sizes of environmental exposures, while the NEPP may inversely lead to the double underestimation of the health effect sizes. This study provides the theoretical foundation and a baseline analytical framework for the recent debate on the NEAP and may also provide essential insights into a range of studies in health geography, environmental epidemiology, and public health.