Pemodelan log linier dan regresi logistik biner bivariat pada medical check-up pegawai negeri sipil (PNS) Institut Teknologi Sepuluh Nopember (ITS)

Viestri, Dinar Ariana (2015) Pemodelan log linier dan regresi logistik biner bivariat pada medical check-up pegawai negeri sipil (PNS) Institut Teknologi Sepuluh Nopember (ITS). Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 1309100077_Undergraduate_Theses.pdf]
Preview
Text
1309100077_Undergraduate_Theses.pdf - Published Version

Download (2MB) | Preview

Abstract

Dengan melakukan cek kesehatan secara periodik, bisa diketahui kondisi kesehatan pribadi secara detail dan apabila terjadi kelainan,
akan ada tindakan antisipatif sedini mungkin sebelum kelainan tersebut menjadi semakin parah, sehingga menjadi penting untuk
melakukan analisis mengenai hasil medical check-up. Data yang
digunakan adalah data medical check-up PNS ITS Tahun 2013/2014. Pada penelitian ini dilakukan pemodelan menggunakan analisis log linier dan regresi logistik biner bivariat. PNS ITS yang melakukan medical check-up terdiri dari laki-laki sebesar 76%, dan perempuan sebesar 24%. Banyaknya dosen sebesar 56,1% sedangkan karyawan
43,9%. Pada latar belakang pendidikannya, pendidikan terakhir S2 menempati urutan terbanyak sebesar 31,5%. Pada analisis log linier menunjukkan adanya hubungan antara variabel fungsi ginjal dengan glukosa darah, hematologi dengan profil lemak, hematologi dengan fungsi ginjal, urin dengan glukosa darah, fungsi liver dengan fungsi ginjal, fungsi liver dengan glukosa darah, profil lemak dengan fungsi ginjal, profil lemak dengan glukosa darah, fungsi ginjal dengan glukosa darah. Sehingga model log linier yang terbentuk merupakan model jenuh (saturated). Pada analisis regresi logistik biner bivariat, variabel asam urat mempengaruhi probabilitas kesehatan PNS ITS dengan Fungsi Ginjal normal dan Glukosa Darah normal, Fungsi Ginjal tidak normal dan Glukosa Darah normal, Fungsi Ginjal normal dan Glukosa Darah tidak normal, Fungsi Ginjal tidak normal dan Glukosa Darah tidak normal.
=============================================================================================
By doing periodic health check-ups, personal health condition
can be known in detail and in case of abnormality, there will be an anticipatory action as early as possible before the disorder becomes more severe, so it is important to perform an analysis of the results of medical check-ups. The data used is the medical check-up PNS ITS year 2013/2014. In this research, modeling using a log linear analysis and bivariate binary logistic regression. ITS civil servants who do medical check-up consists of men by 76%, and 24% women. The number of lecturers of 56.1% while 43.9% of employees. In the educational background, S2 highest ranks of 31.5%. In the log linear analysis showed an association between kidney function variables with blood glucose, lipid profile with hematology, hematology with kidney function, urine with blood glucose, liver function with kidney function, liver function with blood glucose, lipid profile with kidney function, lipid profile with blood glucose. So the log linear models that form a saturated models. In bivariate binary logistic regression analysis, the uric acid variables affecting the probability of health of PNS ITS, with normal kidney function and normal blood glucose, abnormal kidney function and normal blood glucose, normal kidney function and abnormal blood glucose, abnormal kidney function and abnormal blood glucose.

Item Type: Thesis (Undergraduate)
Additional Information: RSSt 519.536 Vie p
Uncontrolled Keywords: log linear, bivariate binary logistic regression, medical check-up, PNS ITS, regresi logistik biner bivariat
Subjects: H Social Sciences > HA Statistics
Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression
Divisions: Faculty of Mathematics and Science > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User: Eny Widiastuti -
Date Deposited: 06 Apr 2018 08:50
Last Modified: 06 Apr 2018 08:50
URI: http://repository.its.ac.id/id/eprint/51712

Actions (login required)

View Item View Item