Geographical and Place-based dependence in multilevel models
Guest Speaker Dr Levi Wolf, University of Bristol
Multilevel models have been applied to study many geographical processes in epidemiology, economics, political science, sociology, urban analytics, and transportation. They are most often used to express how the effect of a treatment or intervention may vary by geographical group, a form of geographical process heterogeneity.
|The University of Exeter Q-Step Centre seminar|
|Date||7 February 2020|
|Time||15:30 to 17:00|
|Place||Clayden Computational Lab|
In addition, these models provide a notion of "platial" dependence: observations that are within the same geographical place are modeled as similar to one another. Recent work has shown that geographical dependence can be introduced into multilevel models, and has examined the empirical properties of these models' estimates.
This talk examines a kind of multilevel model that includes both "platial" and "geographical" dependence. Through further analysis, we obtain the relationship between classic multilevel, spatial multilevel, and single-level models. This structure exposes a tension between a main benefit of multilevel models, estimate shrinkage, and the effects of geographical dependence. We show, both mathematically and empirically, that classic multilevel models may overstate estimate precision and understate estimate shrinkage when spatial dependence is present. This result extends long-standing results in single-level modeling to mutilevel models.
Booking is not required, however please arrive early as seating is limited.