Analisis Survival pada Data Kejadian Bersama Menggunakan Metode Exact Partial Likelihood

Studi Kasus: Kecelakaan Lalu Lintas di Kota Samarinda Tahun 2016

Abstract

Survival data is the data of survival time until the appearance of certain events. In the survival analysis, ties are sometimes found, that is the situation where there are two or more individuals who experience the same event at the same time. There are several methods in estimating the parameters in the case of a ties, one of which is by applying the exact partial likelihood method. The exact method is the most accurate method, which can be applied in estimating Cox regression parameters from traffic accident data of Samarinda City in 2016. Traffic accidents are one of the most deadly events. The four main factors that cause traffic accidents are human factors, vehicles, roads, and weather or environmental factors. The variables used in this study are age, gender, role of victim, driver’s license, vehicle type, time of incident, line of the road, and weather. The results of analysis with the help of Rstudio software showed that the factors whose affect the fatality rate of traffic accident victims of Samarinda City are age and gender. For the age variables concluded that each addition of one year of age of the accident victim, the risk of dying from a traffic accident will also increase 1,0258 times. As for the gender variables concluded that the victim of male sex has a risk of 0.4180 times greater to die due to traffic accidents compared with female victims.