Spatial Logistic Regression Modeling with Inverse Weighting Distance for open unemployment in Districts/Cities on the Island of Sulawesi

Abstract

Unemployment is a condition where a person does not have a job, but is looking for a job. To see the unemployment situation in an area, logistic regression analysis can be used. Logistic regression is an analysis used to see the relationship between the response variable (Y) which is binary and the explanatory variable (X) which is categorical or continuous. The application of logistic regression often has a spatial influence on the model. In this study to model the open unemployment rate the spatial logistic regression method is used. Spatial logistic regression is logistic regression analysis by incorporating spatial influences into the model. Spatial dependency testing is used by Moran’s I Test. The weighting matrix used is the distance inverse weighting matrix. The results obtained, the value of Moran's I Test with a p-value of 2.14 x 10-12 <α (0.05), meaning that there is a spatial influence on the level of open unemployment on the island of Sulawesi. So the spatial logistic regression model is obtained as follows :g(x)    = 4,848 0,000002885(X₁) 0,0473(X₂) 0,006669(X₃) 0,04263(X₄) 0,269(X₅) 0,1642(X₆) 1,531(X₇) 0,1581(X₈) 0,2208(X₉) 0,009732(X₁₀) 0,01871(Z). Spatial factors affect the level of open unemployment based on the significance value <α (0.05)