NEW & EXPANDED in 2020!* Workshop:

Spatial Econometrics: Empirical Analysis of Geospatial Association and Cross-Unit Interdependence

Robert J. Franzese, Jr. (This email address is being protected from spambots. You need JavaScript enabled to view it.)

Do the outcomes in the units or individuals of your research analyses cluster geospatially or within networks? Do some units’ or individuals’ outcomes depend on outcomes in other units/individuals? That is, are the outcomes of interest in your studies likely contagious from units to neighboring or otherwise proximate or connected units? Do the processes you study diffuse across units in some manner? Are there spillovers across subjects? If you study anything in the social sciences, and likely most things beyond, almost certainly they are/do.

This NEW & EXPANDED!* workshop (July 20-24 in Ann Arbor) teaches empirical methods for modeling, for estimating, and for the interpretation of such spatial or cross-unit clustering or interdependence (a.k.a., contagion/diffusion/spillover/network-dependence...).

*Now with more intro to models and methods for spatial or geospatial clustering (as opposed to contagion/interdependence) and distinguishing between the alternate sources of spatial association (common exposure, contagion, network selection).

Applied (computer-lab) sessions and exercises are bilingual, i.e. with lab scripts in Stata and R both available, and students are of course welcome to use other software as they prefer.

Register through the ICPSR Summer School portal, linked here:

http://www.icpsr.umich.edu/icpsrweb/sumprog

The course description from the ICPSR website is linked here & copied below:

http://www.icpsr.umich.edu/icpsrweb/sumprog/courses/0129

Spatial Econometrics: Empirical Analysis of Geospatial Association and Cross-Unit Interdependence

20-24 July 2020

PLEASE DO EMAIL (This email address is being protected from spambots. You need JavaScript enabled to view it.) with any & all questions

Description: Spatial (i.e., geospatial or otherwise cross-unit) association and interdependence are ubiquitous throughout the social sciences, and beyond. That is, events or outcomes in one observational unit are almost always related to similar occurrences in other observational units. This is so for such diverse phenomena as disturbances and conflicts within and among nations; crime, health, and environmental outcomes; economic and other policies in political jurisdictions; consumer, investor, and producer choices in markets; individuals’ opinions and behavior in societies; and voting by citizens in elections or by legislators in legislatures. In contexts where this omnipresent cross-unit association (or correlation) arises from interdependence (or contagion), "standard" statistical methods (which assume independent observations) are inappropriate, and design-based methods of "nonparametric causal-inference" are (at best) inadequate. This workshop introduces strategies appropriate for distinguishing spatial association from spatial interdependence and for proper estimation of processes involving interdependent observations, emphasizing spatial and spatiotemporal models of interdependent continuous and limited outcomes.

The main objective of the workshop is to demonstrate how such spatial, i.e. geo-spatial or otherwise cross-unit, interdependence can be incorporated into empirical analysis most productively. Course participants will learn how to: diagnose spatial-correlation patterns; estimate spatial-regression models; distinguish between different sources of spatial correlation (common exposure, contagion, and selection); and calculate and present the spatial and spatiotemporal effects that empirical models which incorporate interdependence imply. Methods to be covered include: measures of spatial association; models and methods for (exogenous) spatial correlation; instrumental-variable and maximum-likelihood estimators for models with (endogenous) spatial interdependence; multiple-spatial-lag models; spatial interdependence in models with limited and qualitative dependent-variables; and models for coevolutionary processes (i.e., processes with both spatial-cum-network interdependence and endogenous-connectivity/network-selection).

Prerequisites: None; in particular, although participants should be familiar with linear regression and models for qualitative/limited dependent variables (e.g., logit, probit, etc.), this workshop does not assume any prior knowledge of, or experience with, spatial statistics. Indeed, all necessary mathematical, statistical, geospatial-analytic, and spatial-econometric background will be reviewed as needed, albeit (obviously) very quickly.