Pemodelan Status Angkatan Kerja Usia Muda Di Provinsi Jawa Tengah Menggunakan Regresi Logistik Ordinal Multilevel

Utami, Karina Wahyu Sri (2026) Pemodelan Status Angkatan Kerja Usia Muda Di Provinsi Jawa Tengah Menggunakan Regresi Logistik Ordinal Multilevel. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5003221058_Undergraduate_Thesis.pdf Restricted to Repository staff only Download (7MB) | Request a copy

Abstract

Permasalahan ketenagakerjaan usia muda masih menjadi isu strategis di Indonesia, khususnya di Provinsi Jawa Tengah yang ditandai oleh tingginya pengangguran. Status angkatan kerja usia muda bersifat berjenjang, meliputi pengangguran, pekerja tidak penuh, dan pekerja penuh, sehingga pemodelan yang hanya membedakan bekerja dan tidak bekerja berpotensi menghilangkan informasi penting. Selain itu, data ketenagakerjaan memiliki struktur hierarki karena individu berkelompok dalam wilayah administrasi, yang jika diabaikan dapat menghasilkan inferensi yang kurang tepat. Oleh karena itu, penelitian ini bertujuan untuk memodelkan status angkatan kerja usia muda di Provinsi Jawa Tengah tahun 2024 menggunakan pendekatan regresi logistik ordinal multilevel. Data yang digunakan merupakan data mikro Survei Angkatan Kerja Nasional (Sakernas) Agustus 2024, mencakup 5.332 individu yang dipadukan dengan data agregat wilayah, dengan unit analisis penduduk usia 15-24 tahun yang tersebar di 35 kabupaten/kota. Pemodelan dilakukan dengan membangun model regresi logistik ordinal satu level dan multilevel, disertai pengujian signifikansi parameter dan evaluasi kesesuaian model. Pada model diterapkan pembobotan dengan balanced weight untuk menangani ketidakseimbangan data. Hasil penelitian menunjukkan bahwa variabel individu seperti usia, jenis kelamin, status dalam rumah tangga, penyelenggara pendidikan, pengalaman magang, pengalaman kerja, dan partisipasi sosial berpengaruh signifikan terhadap status angkatan kerja kelompok usia muda. Sedangkan untuk faktor wilayah yang berpengaruh meliputi laju PDRB, TPAK, dan rata-rata lama sekolah. Kemudian, variabel-variabel individu seperti usia, jenis kelamin, status perkawinan, status dalam rumah tangga, penyelenggara pendidikan, pengalaman magang, pengalaman kerja, partisipasi sosial, dan klasifikasi wilayah memiliki pengaruh yang berbeda-beda di setiap wilayah sehingga menjadi efek acak (random effect). Didapatkan nilai ICC 0,024 yang artinya sebesar 2,4% variasi dalam variabel respons dijelaskan oleh perbedaan antar kelompok. Pada regresi logistik ordinal multilevel dilakukan evaluasi performa model dengan akurasi sebesar 45,39%. ===================================================================================================================================
Youth employment remain a strategic issues in Indonesia, particularly in Central Java Province, which is characterized by a relatively high unemployment rate. Youth labor force status is inherently ordered, consisting of unemployed, underemployed, and fully employed categories. Therefore, modeling approaches that only distinguish between employed and unemployed may result in the loss of important information. In addition, labor force data have a hierarchical structure because individuals are nested within administrative regions, which, if ignored, may lead to biased or less accurate inferences. Accordingly, this study aims to model youth labor force status in Central Java Province in 2024 using a multilevel ordinal logistic regression approach. The study utilizes microdata from the August 2024 National Labor Force Survey (Sakernas) combined with regional-level data with the unit of analysis consisting of 5.332 individuals aged 15-24 years clustered within 35 regrencies/municipalities. The modeling process involves the development of both single-level and multilevel ordinal logistic regression models, followed by parameter significance testing and model evaluation. A balanced weighting approach was applied during model estimation to handle class imbalance. The results indicate that individual-level variables, including age, sex, household status, type of educational institution, internship experience, work experience, and social participation, have a significant effect on the labor force status of young individuals. At the regional level, gross regional domestic product (GRDP) growth rate, labor force participation rate (LFPR), and mean years of schooling were found to significantly influence the outcome. Furthermore, the effects of several individual-level variables include age, sex, marital status, household status, type of educational institution, internship experience, work experience, social participation, and area classification vary across regions, indicating the presence of random effects in the multilevel ordinal logistic regression model. The estimated ICC was 0.024, suggesting that 2.4% of the total variation in the response variable is attributable to differences between regions. Finally, the predictive performance of the multilevel ordinal logistic regression model was evaluated using classification accuracy, yielding an overall accuracy of 45,39%.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Angkatan Kerja Usia Muda, Ketenagakerjaan, Provinsi Jawa Tengah, Regresi Logistik Ordinal Multilevel, Central Java Province, Labour, Multilevel Ordinal Logistic Regression, Youth Labour Force
Subjects:	H Social Sciences > HA Statistics H Social Sciences > HA Statistics > HA31.3 Regression. Correlation. Logistic regression analysis. H Social Sciences > HA Statistics > HA31.7 Estimation Q Science Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Karina Wahyu Sri Utami
Date Deposited:	31 Jul 2026 07:42
Last Modified:	31 Jul 2026 07:42
URI:	http://repository.its.ac.id/id/eprint/140013

Actions (login required)

View Item