Nainiyah, Khoirotus (2019) Model Regresi Robust untuk Analisis Faktor-faktor yang Mempengaruhi Prestasi Belajar Mahasiswa Jalur Ujian Tulis. Other thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
06211745000025-Undergraduate_Theses.pdf Download (4MB) | Preview |
Abstract
Seleksi masuk PTN jalur ujian tulis memiliki 10 mata uji pada tahap seleksinya yang diharapkan menjadi prediksi tinggi untuk keberhasilan calon mahasiswa dalam mengikuti perkuliahan yang diukur dengan IPS dan IPK. ITS Surabaya merupakan salah satu PTN yang ikut melaksanakan jalur ujian tulis. Sehingga prodi yang ada di ITS akan digunakan sebagai acuan prodi nasional. Selain nilai mata uji, terdapat beberapa faktor yang dapat berpengaruh terhadap prestasi belajar mahasiswa diantaranya jenis kelamin, jenis sekolah, jurusan sekolah, bidikmisi dan asal daerah. Dari nilai mata uji dan karakteristik mahasiswa berdasarkan prodi diduga terdapat hubungan yang mempengaruhi indeks prestasi. Salah satu metode statistika untuk mengetahui hubungan tersebut adalah analisis regresi. Pada analisis regresi metode OLS memiliki asumsi yang harus terpenuhi untuk memperoleh model yang baik. Namun seringkali pada data analisis terdapat data outlier yang akan berpengaruh terhadap ketidakmampuan memenuhi asumsi, sehingga untuk mengatasi permasalahan outlier akan digunakan metode regresi robust. Terdapat beberapa metode robust yang telah banyak digunakan dan pada penelitian ini digunakan metode regresi robust LTS dan LWS. Penelitian ini menggunakan data sekunder hasil ujian tulis tahun 2014 yang pada kajian awal terdapat outlier dan akan di analisis hubungannya terhadap indeks prestasi. Hasil analisis didapatkan faktor yang berpengaruh terhadap prestasi belajar mahasiswa pada sebagian besar prodi yaitu nilai mata uji numerikal dan jenis kelamin, serta pemilihan model terbaik didapatkan hasil metode robust LTS. ==================================================================================================================================
The written test at public universities has 10 test subjects at the selection stage which are expected to be high predictions for the success of prospective students in attending lectures as measured by grade point. ITS Surabaya is one of the public universities that participated in the written test path. So that study programs at ITS will be used as a reference for national study programs. Besides on test subjects, there are several factors that can affect student learning achievement including gender, type of school, school majors, bidikmisi scholarship and regional origin. From the test scores and the characteristics of students based on study programs, it is assumed that there is a relationship that affects the achievement index. One of statistical method to find out the relationship is regression analysis. In OLS method regression analysis has assumptions that must be fulfilled to obtain a good model. However, often in the analysis data there are data outliers that will affect the inability to fulfill the assumptions, so the robust regression will be used to overcome the outlier problem. There are several robust methods that have been widely used and in this study, robust regression method of LTS and LWS are used. This study uses secondary data from the written exam results in 2014 which in the initial study there are outliers and will analyze the relationship to the achievement index. The results of the analysis found that the factors that affect student achievement in most study programs are the numerical value test and gender, and the best selection models obtained is the robust method of LTS.
| Item Type: | Thesis (Other) |
|---|---|
| Additional Information: | RSSt 519.536 Nai m-1 2019 |
| Uncontrolled Keywords: | Indeks Prestasi, OLS, Prodi ITS, Regresi Robust |
| Subjects: | L Education > L Education (General) |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | Khoirotus Nainiyah |
| Date Deposited: | 03 Oct 2024 08:18 |
| Last Modified: | 03 Oct 2024 08:18 |
| URI: | http://repository.its.ac.id/id/eprint/64538 |
Actions (login required)
 |
View Item |