Penerapan Metode Weighted Least Square Untuk Menangani Massalah Heteroskedastisitas Pada Faktor-Faktor Yang Memengaruhi Indeks Pembangunan Manusia
Kode Repository :SKM06/NAT/22
NPM :064118021
Nama :Natalia Medya Lifa Ir Seno
Pembimbing 1 :-Dr.Ir Fitria Virgantari. M. Si.
Pembimbing 2 :-Maya Widyastiti ,M.Si
Abstrak :-Abstract: Human Development Index (IPM) is a measure of the achievement of the main dimensions of human development which is constructed by several components, namely life expectancy which represents the health dimension, the expected number of years of schooling and the average length of schooling which represents the dimension of knowledge, and per capita expenditure on food and non-food adjusted monthly represents a decent standard of living. HDI in Indonesia is still not completely evenly distributed in terms of quality, so estimating which factors have the most significant influence on HDI is something that can be done in the hope that the government can find the right solution regarding the problems that occur in seeking improvement human development. This problem can be analyzed using a multiple linear regression model, where when suspecting it there are several assumptions that must be considered, one of which is that the variance must be homogeneous (homoscedasticity) so that the estimation results obtained correct and effective. In detecting whether the variance is homogeneous or non-homogeneous (heteroscedasticity), a statistical test can be carried out, namely the Glejser Test. One method of estimating multiple linear regression parameters that can correct the heteroscedasticity problem is the Weighted Least Square (WLS) method, namely by minimizing the sum of the squared errors by adding a weight. The results of the analysis show that the deviations in the homoscedasticity test have been resolved, it means that the improvement of the model due to the occurrence of heteroscedasticity symptoms has been successfully carried out and the parameter estimation process will obtain results that can be trusted.
Keywords: multiple linear regression analysis, weighted least square, heteroscedasticity, glejser test, human development index.
Abstrak: Indeks Pembangunan Manusia (IPM) adalah ukuran capaian dimensi utama pembangunan manusia yang dibangun oleh beberapa komponen, yaitu angka harapan hidup yang mewakili dimensi kesehatan, angka harapan lama sekolah dan rata-rata lama sekolah mewakili dimensi pengetahuan, dan pengeluaran per kapita makanan dan non makanan perbulan yang disesuaikan mewakili standar hidup layak.. IPM di Indonesia masih belum sepenuhnya merata dalam tingkat kualitas, sehingga mengestimasikan faktor mana yang paling signifikan pengaruhnya terhadap IPM merupakan hal yang dapat dilakukan dengan harapan pemerintah dapat mencari solusi yang tepat terkait masalah yang terjadi dalam mengupayakan peningkatan pembangunan manusia. Permasalahn tersebut dapat dianalisis menggunakan model regresi linear berganda, dimana ketika menduganya ada beberapa asumsi yang harus diperhatikan, salah satunya adalah ragam harus bersifat homogen (homoskedastisitas) sehingga hasil estimasi yang diperoleh benar dan efektif. Dalam mendeteksi apakah ragam bersifat homogen atau tidak homogen