| dc.description.abstract | Welfare is an important benchmark in assessing the success of development in a region. One of the indicators that can be used to assess it is the Human Development Index (HDI). The HDI in a region is influenced by the HDI of other regions in close proximity, so it is assumed that there is a spatial dependence effect on HDI data. Therefore, the spatial regression method is considered suitable for use in modeling the welfare level of districts/cities in North Sumatra Province. In this study, three regression models were used, namely multiple regression models with Ordinary Least Square estimation, Spatial Autoregressive Model (SAR) and Spatial Error Model (SEM). Based on the selection of the best model with the coefficient of determination and Akaike Information Criterion (AIC) criteria, the best model is the Spatial Autoregressive Model (SAR) with four predictor variables that significantly affect the Human Development Index, namely the variable Life Expectancy at birth (X1), Mean Years of Schooling (X2), Expected Years of Schooling (X3), and Percapita Income (X4). The model obtained is yi = 6,99201 + 0,0668048 Eaai = 1,I = j Wij Yi + 0,391336 X1 + 1,32639 X2 + 0,774933 Xa + 0,0009667 X4 | en_US |
