PENGECEKAN ASUMSI PROPORTIONAL HAZARD PADA MODEL REGRESI COX; CHECKING PROPORTIONAL HAZARD ASSUMPTION IN COX REGRESSION MODEL

Ismalia, Nadia, Danardono

2015 | Skripsi | FMIPA UGM

Cox proportional hazard is a statistical method used to analyze the relationship between the dependent variable in the form survival time and event, with one or more covariates were measured at the time of the study. When the Cox proportional hazard regression model applied to the survival data, assumption proportional hazard should be checked whether the assumptions are met. Nonproportional hazards can cause the absence of differences in the interpretation of data, especially for large sample sizes. There are three general approaches to checking the proportional hazards assumption, approach using graph, testing the goodness of fit (GoF) and approach by adding variables that change over time (time independent variable/ covariates). There are two ways through approach using graph, that uses log-log survival curves and comparison of observed with expected survival curves.

Kata Kunci : Survival; Hazard; Cox Proportional Hazard Regression