Сравнение на методи

	Пространствен модел на маргинална структура	Spatial Instrumental Variables
Област	Причинно-следствено заключение	Причинно-следствено заключение
Семейство	Regression model	Regression model
Година на възникване	2000s–2010s	1988-1998
Създател	Robins, Hernan & Brumback (MSM foundation, 2000); spatial extensions developed in spatial epidemiology literature	Kelejian & Prucha (generalized spatial 2SLS); Anselin (spatial econometrics framework)
Тип	Causal inference / spatial weighting	Quasi-experimental causal inference with spatial dependence
Основополагащ източник	Robins, J. M., Hernan, M. A., & Brumback, B. (2000). Marginal structural models and causal inference in epidemiology. Epidemiology, 11(5), 550-560.	Kelejian, H. H., & Prucha, I. R. (1998). A Generalized Spatial Two-Stage Least Squares Procedure for Estimating a Spatial Autoregressive Model with Autoregressive Disturbances. Journal of Real Estate Finance and Economics, 17(1), 99-121.
Други названия	Spatial MSM, Geospatial MSM, Spatial IPW-MSM, Space-time marginal structural model	Spatial IV, Spatial 2SLS, Spatial Two-Stage Least Squares, S-IV
Свързани	6	6
Резюме	The Spatial Marginal Structural Model (Spatial MSM) extends the classical marginal structural model to settings where units are geographically distributed and spatial dependencies — such as neighborhood spillovers, clustering, and spatial confounding — may bias causal estimates. It estimates causal effects of spatially varying exposures by constructing inverse probability weights that account for both individual covariates and spatial location, then fitting a weighted outcome model in the resulting pseudo-population.	Spatial Instrumental Variables (Spatial IV) is a causal inference method for settings where units — regions, firms, neighborhoods — are spatially interdependent, creating endogeneity that standard IV approaches ignore. It constructs instruments from the spatially lagged values of exogenous characteristics of neighboring units, then applies two-stage least squares to recover unbiased causal estimates in the presence of both endogenous regressors and spatial autocorrelation.