Simultaneous hypothesis testing on geographically weighted nonparametric regression (A case study: Poverty data in Sulawesi Island)

In this study, the Geographically Weighted Nonparametric Regression (GWNR) model was examined. This model is an extension of nonparametric regression to the geographically weighted regression model. Finding test statistics and their distribution for the simultaneously hypothesis test in the GWNR model was the aim of this investigation. Test statistics were obtained based on the difference in the residual sum of squares under H₀ and H₁ with each having a spatial factor. The statistical distribution of the test was the F distribution. The application of modeling and hypothesis testing were done simultaneously using the percentage of poor people in 81 regencies/cities in Sulawesi Island. The results of the simultaneous hypothesis testing showed rejection, which means that the predictor variable used had a significant effect on poor people in Sulawesi Island.