A suggested test for spatial autocorrelation and/or heteroskedasticity and corresponding Monte Carlo results
Article Abstract:
A new specification test that determines whether the error terms of a spatial regression model are heteroskedastic and/or spatially correlated is evaluated. In testing the regression model, the analyzed test does not require the linearity of the model nor the normal distribution of the error terms. Based on the Monte Carlo results, the standard test based on Moran's I (MI) statistic was found to detect spatial autocorrelation best. When spatial autocorrelation is high, the 'disturbance problem' can be detected by all the considered tests, namely, MI, a Langrangian multiplier (LM) test, a generalized LM, and the analyzed specification test.
Publication Name: Regional Science & Urban Economics
Subject: Social sciences
ISSN: 0166-0462
Year: 1998
User Contributions:
Comment about this article or add new information about this topic:
Spatial autocorrelation: a new computationally simple test with an application to per capita county police expenditures
Article Abstract:
A large sample test for determining spatial autocorrelation in econometric models is illustrated. Both linear and nonlinearmodels are applicable and the test is simple in terms of computation. Based on large sample moments methods, the test can provide the added information neededby tests based on estimators. In an illustration using a regression model of the per capita county police expenditures of Mississippi, the simplified version of the test revealed autocorrelation that was absent in the full version.
Publication Name: Regional Science & Urban Economics
Subject: Social sciences
ISSN: 0166-0462
Year: 1992
User Contributions:
Comment about this article or add new information about this topic:
Located autocorrelation diagnostic statistic (LADS) for spatial models: conceptualization, utilization, and computation
Article Abstract:
An error diagnostic called the localized autocorrelation diagnostic statistic (LADS) is presented. It is differentiated from other diagnostics in that it focuses on localized occurrences of a potentially distinct problem in which localized problems are not predicted in advance. It evaluates cases of poorly functioning areas and determines whether the case is the exception or the rule.
Publication Name: Regional Science & Urban Economics
Subject: Social sciences
ISSN: 0166-0462
Year: 1992
User Contributions:
Comment about this article or add new information about this topic:
- Abstracts: A general equilibrium spatial model of housing quality and quantity. Factor taxes and evasion in general equilibrium
- Abstracts: An investigation into the social context of early adoption behavior. Pennies-a-day: the effect of temporal reframing on transaction evaluation
- Abstracts: Technology measurement and the appraisal of information technology. Shaping user-side innovation through standardisation: the example of ICT