In the econometric field, spatio-temporal data are often modeled by means of spatial dynamic panel data models (SDPD). In the last two decades, several versions of the SDPD model have been proposed, each based on different assumptions on the spatial parameters and on different properties of the estimators. In particular, the classic version of the SDPD model assumes that the spatial parameters are homogeneous over location. Another version, more recently proposed and called Generalized SDPD, assumes that spatial parameters are adaptive over location.
In this work we propose a strategy for testing a particular structure of the spatial dynamic panel data model on a given dataset, by means of a multiple testing procedure that allows to choose between the generalized version of the model and some specific versions derived from the general one by imposing particular constraints on the parameters. The theoretical derivations of the testing procedure are carried out in a high-dimensional setup, where the number of locations may grow to infinity with the time series length. This makes our proposal also suitable to make some kind of “location screening”, which means automatically identifying the locations that most influence the values of a given spatio-temporal time series.
Parrella Maria Lucia is an Associate Professor of Statistics at the University of Salerno, where she teaches Statistics, Probability, Inference and Data Mining in several courses at undergraduate and postgraduate level. Her research interests are: kernel and local polynomial estimation, bootstrap methods for dependent data; variable selection in high dimensional regression; econometric spatio-temporal models. She is a member of the Italian Statistical Society (SIS), the International Society of NonParametric Statistics (ISNPS) and the International Environmetrics Society (TIES).