Spatial Autocorrelation in Mumps Data


  • Maya Zhelyazkova
  • Maroussia Nikiforova Slavtchova-Bojkova* Faculty of Mathematics and Informatics, Sofia University



This paper aims to ascertain the presence or absence of spatial auto-correlation in the dissemination of the disease parotitis. For the solutionof similar problems the Moran's I and Geary's c tests (see for example [1]and [3]) are applied over the real data set from Bulgaria.In order to apply the Moran's I and Geary's c tests we need to nda geographic center in accordance with which to orientate the rest of thecounties. In our case it is convenient to put in this center the beginningof the coordinate system whose axes coincide with geographical directionseast, west (axis x) and north, south (axis y). The center of the coordinatesystem is selected as the county center with the highest dissemination ofmumps. This way every county center is uniquely determined with itscoordinates x and y.Although, with the reduction of the distances the p-value diminishes, wecan not reject the null hypothesis that there is zero spatial autocorrelationin the variable morbidity.



[1] How to calculate Moran's I in R

[2] M. Mitova-Bobcheva,M. Slavtchova-Bojkova,В  M. Kojouharova, A.Kurchatova, Analysis and Monitoring Surveillance Data of Mumps inBulgaria, Pliska Studia Mathematica Bulgaria, vol. 20, 149-154,(2011).

[3] M. Rosenberg, R. Socal, N. Oden, D. DiGiovanni, Spatial Autocorrela-tion of Cancer in Western Europe, European Journal of Epidemil-ogy, vol. 15, 15-22, (1999).






Conference Contributions