Spatial Autocorrelation

Video 1 : Spatial Autocorrelation

Module Description

“The first law of geography: Everything is related to everything else, but near things are more related than distant things.”[1]. A question often asked is whether or not features with similar values are clustered, randomly distributed or dispersed. Spatial autocorrelation measures the degree of correlation on space [2]. Tests of spatial autocorrelation examine the independency of observed value in relation to values of that variable at neighboring locations. In this module, the value of the Moran’s I index and both the z-score and the p-value are calculated to assess the importance of that index. P-values are numerical approximations of the area under the curve for a known distribution, limited by the test statistic. We use the spdep package [3]

Argument

Variable: a numeric vector the same length as the neighbours list in listw.

zero.policy: default NULL, use global option value; if TRUE assign zero to the lagged value of zones without neighbours, if FALSE assign NA.

style: starting from a binary neighbours list, in which regions are either listed as neighbours or are absent (thus not in the set of neighbours for some definition), the function adds a weights list with values given by the coding scheme style chosen. B is the basic binary coding, W is row standardised (sums over all links to n), C is globally standardised (sums over all links to n), U is equal to C divided by the number of neighbours (sums over all links to unity), while S is the variance-stabilizing coding scheme proposed by Tiefelsdorf et al. [4] (sums over all links to n). nsim: number of permutations.

alternative: a character string specifying the alternative hypothesis, must be one of “greater” (default), “less” or “two.sided”.

Value

A dataframe containing the following components: Moran I test under randomisation, Monte-Carlo simulation of Moran I, Local Moran’s I statistic summary, and Monte-Carlo simulation of Local Moran’s I statistic summary. Also the value of the standard deviation of Moran’s I, p.value the p-value of the test, the value of the observed Moran’s I, its expectation and variance under the method assumption. The Spatial Autocorrelation (Global Moran’s I) tool is an inferential statistic, which means that the results of the analysis are always interpreted within the context of its null hypothesis. * When the p-value returned by this tool is statistically significant, you can reject the null hypothesis.

References

[1] Tobler, W. R. 1970. A computer movie simulating urban growth in the Detroit region. Economic Geography, 46: 234–40.

[2] Cliff, A. D., & Ord, J. K. (1981). Spatial processes: models & applications. Taylor & Francis.

[3] Bivand, Roger S. and Wong, David W. S. (2018) Comparing implementations of global and local indicators of spatial association TEST, 27(3), 716-748. URL https://doi.org/10.1007/s11749-018-0599-x

[4] Tiefelsdorf, M., Griffith, D. A., Boots, B. (1999). A variance-stabilizing coding scheme for spatial link matrices, Environment and Planning A, 31, pp. 165–180; Kelejian, H. H., and I. R. Prucha. 2010. Specification and estimation of spatial autoregressive models with autoregressive and heteroskedastic disturbances. Journal of Econometrics, 157: pp. 53–67.