(Pontificia Universidad Católica del Perú, 2021-05-06) Mur, Jesús
We present a simple test of spatial autocorrelation based on the skedastic structure of the spatial series. Its distribution function is known for all sample sizes. Moreover, it is very simple to obtain, specially in a case of small samples where the new GQsp test has great power, higher than other alternatives existing in the literature.
(Pontificia Universidad Católica del Perú, 2021-05-06) López Hernández, Fernando A.; Mínguez Salidos, Román
This paper presents a test based on the principle of Lagrange Multipliers to identify spatial instability in the constant coefficient of regression models including substantive spatial dependence. The test has been adapted to the Scan methodology. Its main advantage is that it identifies areas with differential behavior without the need to provide information about their location, shape, or size. The study shows the utility of the test, reconsidering the results obtained by Mur et al.(2008) about instability in the distribution of per capita income in European regions.