Pemodelan Mixed Effects Logistic Regression (Studi Kasus pada Data Anak Putus Sekolah Jenjang Sekolah Dasar (SD) di Provinsi Jawa Timur)

Nurul Ainy, Khomaria (2023) Pemodelan Mixed Effects Logistic Regression (Studi Kasus pada Data Anak Putus Sekolah Jenjang Sekolah Dasar (SD) di Provinsi Jawa Timur). Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6003211012-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2023. Download (4MB) | Request a copy

Abstract

Angka putus sekolah jenjang SD tahun 2022 sebesar 38.716 anak, dan Provinsi Jawa Timur menyumbang 6,37% terhadap angka putus sekolah secara nasional. Permasalahan putus sekolah merupakan masalah serius mengingat Undang-Undang Sistem Pendidikan Nasional menyatakan bahwa setiap warga negara yang berusia tujuh sampai lima belas tahun wajib mengikuti pendidikan dasar. Penelitian ini akan melakukan pemodelan hierarki mixed effects logistic regression dalam analisis data anak putus sekolah dan melakukan kajian teori untuk penaksiran parameternya. Tujuan utama penelitian ini adalah untuk mengetahui faktor-faktor yang mempengaruhi putus sekolah SD di Jawa Timur dengan variabel respon berupa data kategorik menggunakan analisis regresi logistik biner dibandingkan dengan mixed effects logistic regression. Model mixed effects logistic regression memperhitungkan adanya potensi efek yang mungkin terjadi dari adanya siswa yang tersarang di dalam sekolah, yaitu pengaruh anak putus sekolah bukan hanya dari faktor individu siswa, tetapi bisa terjadi dari potensi efek sekolah. Data yang digunakan dalam penelitian ini merupakan data individu sebanyak 1.048.575 siswa SD di Provinsi Jawa Timur, yang terdiri atas 10 variabel prediktor level individu siswa dan 10 variabel prediktor level sekolah. Dari hasil analisis diperoleh bahwa model terbaik untuk status putus sekolah siswa SD di Provinsi Jawa Timur adalah mixed effects logistic regression, dengan nilai akurasi sebesar 99,4631%, AUC 50,4061%, dan sensitivitas 0,8159%. Variabel yang mempengaruhi status putus sekolah siswa SD di Provinsi Jawa Timur adalah tingkat kelas, umur, jarak dari rumah ke sekolah, status pernah mengulang, pekerjaan orang tua, status sekolah, rasio siswa SD per rombongan belajar, rasio siswa guru, jumlah siswa per sekolah, dan sekolah punya toilet terpisah.
====================================================================================================================================
The dropout rate for elementary school in 2022 is 38.716 children, and East Java Province contributes 6,37% to the dropout rate nationally. The problem of dropping out of school is a serious problem considering that the Law on the National Education System states that every citizen aged seven to fifteen is required to attend basic education. This study will perform hierarchical mixed effects logistic regression modeling in data analysis of dropout children and conduct theoretical studies for parameter estimation. The main objective of this study was to determine the factors that influence elementary school dropouts in East Java with the response variable in the form of categorical data using binary logistic regression analysis compared to mixed effects logistic regression. The mixed effects logistic regression model takes into account the potential effects that might occur from the presence of students who are nested in the school, that is, the effect of dropping out of school is not only from individual student factors, but can occur from the potential effects of school. The data used in this study were individual data of 1,048,575 elementary school students in East Java Province, consisting of 10 individual student level predictor variables and 10 school level predictor variables. From the results of the analysis it was found that the best model for dropout status of elementary school students in East Java Province was mixed effects logistic regression, with an classification accuracy value of 99,4631%, AUC 50,4061%, and sencitivity 0,8159%. The variables that affect the dropout status of elementary school students in East Java Province are grade level, age, distance from home to school, status of having repeated, parent’s occupation, school status, ratio of elementary school students per study group, student-teacher ratio, number of students per school, and the school has separate toilets.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	ketepatan klasifikasi, mixed effects logistic regression, putus sekolah, regresi logistik biner, sekolah dasar binary logistic regression, classification accuracy, elementary school, dropouts, mixed effects logistic regression
Subjects:	H Social Sciences > HA Statistics > HA31.7 Estimation Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49101-(S2) Master Thesis
Depositing User:	Khomaria Nurul Ainy
Date Deposited:	11 Aug 2023 01:06
Last Modified:	11 Aug 2023 01:06
URI:	http://repository.its.ac.id/id/eprint/104366

Actions (login required)

View Item