nismayil01/Analysing-Political-Boundaries-and-Population-Density-in-UGANDA-2006-2010

Modifiable Areal Unit Problem (MAUP) defines how the results change based on the chosen units for data collection. Spatial data aggregation, which involves data aggregation into various zones, can result in scale and zone configuration problems. # Analysing-Political-Boundaries-and-Population-Density-in-UGANDA-2006-2010 Modifiable Areal Unit Problem (MAUP) defines how the results change based on the chosen units for data collection. Spatial data aggregation, which involves data aggregation into various zones, can result in scale and zone configuration problems. In bivariate analysis, the correlation coefficient between two variables depends on the level of aggregation which can become stronger or weaker and lead to incorrect results if the data is not accurately aggregated. Thus, we can have a false sense of the relation between variables. This effect is even more complex within multivariate analysis. Gehlke and Biehl (1934) concluded that contiguous aggregation results in the increase of correlation coefficient for absolute variables, which is not the same scenario for the ratios or percentages. The impact of aggregation depends on several factors such as random grouping, grouping with independent or dependent variables or by proximity according to Blalock (1964). Random grouping was found to not affect the correlation coefficient. Independent variable aggregation results in the increase of the correlation coefficient with no change in slope parameter, but the dependent variable will increase the correlation coefficient as well as the slope parameters. The grouping by proximity is also slightly inflated e in terms of the correlation coefficient and slope. To solve this issue, there has been several solutions proposed one of which is spatial entropy. Here the model is trained by the model performance and spatial data information (areal size, level of resolution etc.) and parameter estimates.