Sinaga, Fieldri (2018) Small Area Estimation pada Kasus Respon Multinomial dengan Pendekatan Hierarchical Bayes (Studi Kasus : Proporsi Lama Sekolah di Kabupaten Keerom). Masters thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
06211650010021-Master_Thesis.pdf - Accepted Version Download (1MB) | Preview |
Abstract
Penelitian ini bertujuan untuk mengestimasi proporsi lama sekolah tingkat kecamatan di Kabupaten Keerom menggunakan pendekatan Hierarchical Bayes berdasarkan model Multinomial Logit, dan membandingkannya dengan estimasi langsung. Otonomi daerah mendorong pemerintah daerah untuk memiliki informasi yang dapat diandalkan dan detail di daerah kecil. Penghitungan lama sekolah level kecamatan tidak dapat dilakukan secara langsung karena kurang sampel. Salah satu cara untuk mengatasi hal tersebut adalah melalui pendekatan estimasi tidak langsung Small Area Estimation (SAE). Salah satu model SAE yang dapat digunakan adalah model Hierarchical Bayes (HB). Model Multinomial Logit digunakan sebagai model penghubung. Model HB memiliki kelebihan dibandingkan dengan model SAE lainnya karena menghasilkan MSE lebih rendah, varian posterior terkecil, dan dapat menangani variabel respon diskrit. Estimasi proporsi lama sekolah menggunakan model HB menunjukkan beberapa karakteristik, yaitu konvergen distribusi posterior, koefisien variasi lebih rendah dari estimasi langsung.
======================================================================================================
The purpose of this study is to estimate the proportion of years schooling at sub-district level in Keerom District using Hierarchical Bayes approach based on Multinomial Logit model and compare it with direct estimation. Regional autonomy encourages local governments to have reliable and detailed information in small areas. Estimation of years schooling for sub-district level can’t be done directly because of lack of sample. One of the way to overcome this is through the indirect estimation approach of Small Area Estimation (SAE). One of the SAE models that can be used is the Hierarchical Bayes (HB) model. The Multinomial Logit model is used as the connecting model. The HB model has advantages over other SAE models, as it produces a smaller MSE, produces the smallest posterior variance, and can handle discrete response variables. The proportion of years schooling estimation using HB model shows some characteristic, that is convergent posterior distribution and the coefficient of variance is smaller than the direct estimate.
| Item Type: | Thesis (Masters) |
|---|---|
| Additional Information: | RTSt 519.542 Sin s-1 3100018078370 |
| Uncontrolled Keywords: | Hierarchical Bayesian, Small Area Estimation, Gibbs Sampling, Multinomial Response |
| Subjects: | H Social Sciences > HA Statistics > HA31.7 Estimation Q Science > QA Mathematics > QA279.5 Bayesian statistical decision theory. |
| Divisions: | Faculty of Mathematics, Computation, and Data Science > Statistics > 49101-(S2) Master Thesis |
| Depositing User: | Fieldri Sinaga |
| Date Deposited: | 18 Jan 2019 04:01 |
| Last Modified: | 08 Feb 2021 07:22 |
| URI: | http://repository.its.ac.id/id/eprint/58610 |
Actions (login required)
 |
View Item |