PENERAPAN ANALISIS REGRESI RIDGE PADA DATA PASIEN HIPERTENSI DI RUMAH SAKIT UMUM DAERAH SIDIKALANG TAHUN 2014

medis pasaribu, Abdul Jalil Armi Arma, Ria Masniari Lubis

Abstract


ABSTRACT

Multicollinearity is the correlation between some independent variables which makes coefficient regression yielded by multiple regression analysis cannot be used in estimating. The problem of multicollinearity which is indicated by the value of Variance Inflation Factor (VIF) > 10 occurs when the influence of age, obesity (IMT), and cholesterol were modeled toward systole and diastole blood pressure.

The objective of the research was to obtain the regression model of age, obesity (IMT), and cholesterol with systole blood pressure and with diastole blood pressure of patients suffered from hypertension at the Inpatient Wards of the RSUD (Regional General Hospital) Sidikalang, in 2014. The samples consisted of 105 patients. The data were secondary data and using α=0,1. Analyzed by using ridge regression in order to obtain correct regression model.

The result of ANOVA test showed that p < 0.000 which indicated that, simultaneously, there was the influence of age, obesity (IMT), and cholesterol on systole and diastole blood pressure. The equation for modeling blood pressure was obtained at k = 0.07; they were Y systole = 21.2828 + 0.7169 Age + 1.7986 IMT + 0.2222 Cholesterol, and Y diastole = 24.5290 + 0.5025 Age + 0.6648 IMT + 0.0946 Cholesterol.

It is recommended that the next researchers solve the problems of multicollinearity during the modeling; one of them is by using ridge regression analysis.

Full Text: PDF

Refbacks

  • There are currently no refbacks.