Metode fast double bootstrap pada regresi spasial data panel dengan Spatial Fixed Effect (studi kasus : pemodelan penduduk miskin di Provinsi NTB) == Fast Double Bootstrap Method on Spatial Panel Data Regression with Spatial Fixed Effect (Case Study: Modelling of Poverty in NTB Province)

Muhtasib, Nora (2015) Metode fast double bootstrap pada regresi spasial data panel dengan Spatial Fixed Effect (studi kasus : pemodelan penduduk miskin di Provinsi NTB) == Fast Double Bootstrap Method on Spatial Panel Data Regression with Spatial Fixed Effect (Case Study: Modelling of Poverty in NTB Province). Masters thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 1313201709_Master_Thesis.pdf]
Preview
Text
1313201709_Master_Thesis.pdf - Published Version

Download (3MB) | Preview

Abstract

Analisis tentang kemiskinan seringkali memiliki keterkaitan spasial antara satu wilayah dan wilayah di sekitarnya. Analisis data yang mempertimbangkan unsur spasial tidak hanya diterapkan pada data crosssection, namun telah berkembang pada data panel. Penggunaan data panel dalam regresi spasial memiliki banyak keuntungan, namun pengujian spatial dependency dan estimasi parameter yang dihasilkan pada regresi spasial data panel akan menjadi tidak akurat ketika diterapkan pada wilayah dengan unit analisis yang kecil. Salah satu metode mengatasi permasalahan pada jumlah pengamatan yang kecil adalah metode resampling. Penelitian ini menggunakan metode resampling fast double bootstrap (FDB) dengan memodelkan presentase penduduk miskin di Provinsi NTB sebagai kasus. Hasil pengujian spatial dependency menyimpulkan bahwa terdapat dependensi spasial kemiskinan antar wilayah di NTB. Pemodelan Spatial Autoregressive data panel dengan pendekatan FDB mampu menjelaskan keragaman persentase penduduk miskin di NTB sebesar 99,46 persen dan memenuhi asumsi kenormalan residual. Variabel yang berpengaruh secara signifikan terhadap persentase penduduk miskin di NTB adalah PDRB perkapita, persentase penduduk berusia 10 tahun keatas yang tidak/belum pernah bersekolah dan persentase penduduk yang bekerja di sektor industri.
===========================================================================================
Poverty analysis often have a spatial relationship between the region and the surrounding territories. Data analysis that considering spatial element is not only applied to the crosssectional data, but has grown on panel data. The use of panel data in spatial regression has many advantages, but the spatial dependency testing and parameter estimation result will be inaccurate when applied to small unit of analysis. One method to overcome the problems in a small number of observations is a resampling method. This study uses the fast double bootstrap resampling method (FDB) by modeling the percentage of poor people in NTB province as a case. The test results suggest that there are spatial dependency on poverty between regions in NTB. Modeling spatial autoregressive panel data with FDB approach is able to explain the diversity of the percentage of poor people in NTB by 99.46 percent and satisfy the assumption of normality of the residuals. The variables that significantly affect the percentage of poor people in NTB is GDP per capita, the percentage of the population aged 10 years and older who do not / have not been schooling and the percentage of people who work in the industrial sector.

Item Type: Thesis (Masters)
Additional Information: RTSt 519.536 Muh m
Uncontrolled Keywords: Model kemiskinan, Regresi spasial data panel, Fast double bootstrap, Spatial fixed effect, Poverty modeling, Spatial panel data regression
Subjects: H Social Sciences > HA Statistics
Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression
Divisions: Faculty of Mathematics and Science > Statistics > 49101-(S2) Master Thesis
Depositing User: Eny Widiastuti -
Date Deposited: 12 Apr 2018 03:53
Last Modified: 24 Aug 2018 03:51
URI: http://repository.its.ac.id/id/eprint/51735

Actions (login required)

View Item View Item