Analisis Regresi Cox terhadap Kelulusan Mahasiswa Program Sarjana FMIPA IPB

Abstract

Kebijakan Merdeka Belajar-Kampus Merdeka (MBKM) memberikan fleksibilitas bagi mahasiswa dalam menyelesaikan studi, termasuk melalui Kurikulum Multistrata (K2020) di IPB University. Tingkat kelulusan tepat waktu menjadi salah satu indikator keberhasilan perguruan tinggi. Data menunjukkan bahwa tingkat kelulusan tepat waktu mahasiswa FMIPA IPB angkatan 2017–2019 belum memenuhi target, sehingga diperlukan analisis lebih lanjut pada angkatan 2020. Penelitian ini bertujuan membangun model regresi Cox untuk mengidentifikasi faktor yang memengaruhi kelulusan tepat waktu mahasiswa sarjana FMIPA IPB angkatan 2020. Model ini mengestimasi hazard ratio untuk membandingkan peluang kelulusan antar kelompok serta menangani data tersensor, yaitu mahasiswa yang belum lulus selama periode pengamatan. Data terdiri dari 748 mahasiswa FMIPA IPB angkatan 2020. Pelanggaran asumsi proportional hazard diatasi dengan stratifikasi berdasarkan kombinasi program studi dan kelompok IPK. Hasil analisis menunjukkan bahwa risiko kelulusan tepat waktu berbeda antar strata. SKS konversi EC dan KIPK berpengaruh signifikan terhadap kelulusan tepat waktu. Mahasiswa yang mengambil lebih dari 10 SKS EC memiliki peluang lulus tepat waktu 64,3% lebih tinggi, sedangkan mahasiswa bukan penerima KIPK memiliki peluang 25,6% lebih rendah. Selain itu, mahasiswa dengan 3–10 SKS EC, perempuan, dan jalur mandiri cenderung memiliki peluang kelulusan tepat waktu lebih tinggi.

The Merdeka Belajar-Kampus Merdeka (MBKM) policy provides flexibility for students in completing their studies, including through the Multistrata Curriculum (K2020) at IPB University. One indicator of the success of higher education is the on-time graduation rate. The data shows that the on-time graduation rate of FMIPA IPB students in the 2017-2019 batch has not met the target, so further analysis needs to be done on the 2020 batch. This study aims to build a Cox regression model of on-time graduation and identify the influence of several factors on the graduation of undergraduate students of FMIPA IPB class of 2020. This model estimates the hazard ratio to compare the chances of graduation between groups, and is able to handle censored data, namely students who have not graduated during the observation period. The data consisted of 748 students of FMIPA IPB class of 2020. Violation of the proportional hazard assumption was overcome by stratification based on a combination of study programs and GPA groups. The analysis results show that the risk of graduating on time is different between strata. EC credit conversion and KIPK have a significant effect on timely graduation. Students who take more than 10 EC credits have a 64.3% higher chance of graduating on time, while non-KIPK students have a 25.6% lower chance. In addition, students with 3– 10 credits of EC, female, and independent admission track tend to have a higher chance of graduating on time.