PEMODELAN REGRESI DOUBLE HURDLE POISSON UNTUK MENGATASI DATA OVERDISPERSI PADA KASUS ANGKA KEMATIAN IBU DI PROVINSI JAWA BARAT

PUTRI, AYU SETIYOWATI (2023) PEMODELAN REGRESI DOUBLE HURDLE POISSON UNTUK MENGATASI DATA OVERDISPERSI PADA KASUS ANGKA KEMATIAN IBU DI PROVINSI JAWA BARAT. Sarjana / Sarjana Terapan (S1/D4) thesis, Universitas Muhammadiyah Semarang.

[img]
Preview
Text
1. HALAMAN DEPAN.pdf

Download (517kB) | Preview
[img]
Preview
Text
5. ABSTRAK.pdf

Download (447kB) | Preview
[img]
Preview
Text
11. BAB 1.pdf

Download (645kB) | Preview
[img]
Preview
Text
12. BAB II.pdf

Download (910kB) | Preview
[img] Text
13. BAB III.pdf
Restricted to Repository staff only

Download (553kB) | Request a copy
[img] Text
14. BAB IV.pdf
Restricted to Repository staff only

Download (685kB) | Request a copy
[img]
Preview
Text
15. BAB V.pdf

Download (508kB) | Preview
[img]
Preview
Text
16. DAFTAR PUSTAKA.pdf

Download (530kB) | Preview
[img] Text
18. HASIL PENGOLAHAN DATA.pdf
Restricted to Repository staff only

Download (671kB) | Request a copy
[img] Text
19. FULL TEKS SKRIPSI.pdf
Restricted to Repository staff only

Download (1MB) | Request a copy
Official URL: http://repository.unimus.ac.id

Abstract

ABSTRAK Putri Ayu Setiyowati, B2A221005, Pemodelan Regresi Double Hurdle Poisson Untuk Mengatasi Data Overdispersi dan Multikolinieritas Pada Kasus Angka Kematian Ibu Di Provinsi Jawa Barat Tahun 2020, Program Studi Statistika, Universitas Muhammadiyah Semarang, Pembimbing: I. Dr. Rochdi Wasono, M.Si; II. Indah Manfaati Nur, M.Si. Angka kematian ibu di Provinsi Jawa Barat cenderung mengalami kenaikan dari 684 kasus pada tahun 2019 menjadi 745 kasus pada tahun 2020 atau meningkat sebesar 61 kasus atau 85,77% per 100.000 kelahiran hidup. Analisis regresi poisson merupakan regresi nonlinier untuk data diskrit dengan asumsi yang harus terpenuhi yaitu equidispersi. Namun pada praktiknya asumsi yang sering terjadi yaitu overdispersi dimana nilai varians lebih besar dari nilai rata-rata pada variabel respon. Overdispersi terjadi salah satunya disebabkan oleh excess zero. Metode regresi double hurdle poisson merupakan salah satu metode yang digunakan untuk data overdispersi. Regresi double hurdle poisson terbagi menjadi dua model yaitu model logit yang digunakan untuk menaksir data yang bernilai nol, model truncated poisson digunakan untuk menaksir data bernilai positif saja. Berdasarkan hasil analisis diketahui bahwa variabel persalinan oleh tenaga medis dan persentase perempuan hamil dengan usia kawin < 17 tahun berpengaruh terhadap angka kematian ibu di provinsi Jawa Barat tahun 2020 pada model truncated poisson. sedangkan variabel persentase penduduk miskin berpengaruh terhadap angka kematian ibu di provinsi Jawa Barat tahun 2020 pada model logit. Berdasarkan nilai AIC diketahui bahwa model regresi double hurdle poisson memiliki nilai AIC lebih kecil yaitu sebesar 77,61752 jika dibandingkan dengan model regresi poisson yaitu sebesar 80,19337 sehingga model regresi double hurdle poisson merupakan model yang baik digunakan untuk menangani data overdispersi. Kata kunci: Kematian Ibu, Overdispersi, Excess Zero, Regresi Double Hurdle Poisson ABSTRACT Putri Ayu Setiyowati, B2A221005, Double Hurdle Poisson Regression Modeling To Overcome Data Overdispersion In Cases Of Maternal Mortality Rate In West Java Province, Statistics Study Program, Muhammadiyah University Of Semarang, Supervisor: I. Dr. Rochdi Wasono, M.Si; II. Indah Manfaati Nur, M.Si. The maternal mortality rate in West Java Province tends to increase from 684 cases in 2019 to 745 cases in 2020 or an increase of 61 cases or 85.77% per 100,000 live births. Poisson regression analysis is a nonlinear regression for discrete data with the assumption that must be met, namely equidispersion. However, in practice the assumption that often occurs is overdispersion where the variance value is greater than the average value of the response variable. One of the causes of overdispersion is excess zero. The double hurdle Poisson regression method is one of the methods used for overdispersion data. Double hurdle Poisson regression is divided into two models, namely the logit model which is used to estimate zero-value data, the truncated Poisson model is used to estimate positive data only. Based on the results of the analysis, it is known that the variables of deliveries by medical personnel and the percentage of pregnant women with marriage age <17 years have an effect on the maternal mortality rate in West Java province in 2020 in the truncated Poisson model. while the variable percentage of poor people influences the maternal mortality rate in West Java province in 2020 in the logit model. Based on the AIC value, it is known that the double hurdle Poisson regression model has a smaller AIC value of 77.61752 when compared to the Poisson regression model which is equal to 80.19337 so that the double hurdle Poisson regression model is a good model used to handle overdispersion data. Keywords: Maternal Mortality, overdispersion, excess zero, double hurdle Poisson regression

Item Type: Thesis (Sarjana / Sarjana Terapan (S1/D4) )
Call Number: 006/Statistika/VII/2023
Subjects: L Education > Statistics
Divisions: Faculty of Agricultural Science and Technology > S1 Statistics
Depositing User: perpus unimus
Date Deposited: 17 Jul 2023 03:56
Last Modified: 17 Jul 2023 03:56
URI: http://repository.unimus.ac.id/id/eprint/7129

Actions (login required)

View Item View Item