Journal of the Korean Data & http://dx.doi.org/10.7465/jkdi.2018.29.6.1457 Information Science Society 한국데이터정보과학회지 2018, 29(6), 1457–1468 A comparative simulation study for estimating accelerated failure time models † Sangbum Choi 1 1 Department of Statistics, Korea University Received 1 October 2018, revised 14 November 2018, accepted 15 November 2018 Abstract Semiparametric accelerated failure time (AFT) models directly relate the predicted failure times to covariates and are a useful alternative to Cox’s propor- tional hazards models that work on the hazard function or the survival function. In this paper, we briefly review different approaches to estimate the AFT model and evaluate their performance with finite samples via extensive simulation studies. Specifically, we compared (i) inverse probability of censoring weighted (IPCW) least squares, (ii) log-rank estimator, (iii) Gehan-type log-rank estima- tor, (iv) Buckley-James estimator, and (v) nonparametric maximum likelihood estimator (NPMLE). Overall, rank-based estimators and Buckley-James esti- mator are efficient and relatively more robust to distributions of residual and censoring variables, whereas the IPCW estimator is very sensitive to distribu- tion and amount of censoring. The NPMLE is asymptotically efficient and useful as it allows for hazard-based formulation, and thus can be used to analyze more structured survival data. Keywords: Linear model, rank regression, relative efficiency, survival analysis. 1. Introduction Consider a random sample of n subjects. Let T i be an uncensored dependent variable of interest, such as the survival time, and let X i be an observable p × 1 vector of covariates, where i =1, ..., n. In many applications, especially in biomedical studies, T i cannot be completely observed due to possible censoring, for instance, end of the study, withdrawal of patients, or death from a cause unrelated to the † This research was supported by Basic Science Research Program through the National Research Foundation of Korea (2017R1C1B1004817) and the grant from Korea University (K1822621). 1 Assistant professor, Department of Statistics, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul 02841, Korea. Email: choisang@korea.ac.kr