Journal of Statistical Research ISSN 0256 - 422 X 2013, Vol. 47, No. 1, pp. 51-61 QUANTILE REGRESSION MODELS WITH PARTIALLY FUNCTIONAL EFFECTS FOR RANDOMLY RIGHT CENSORED DATA: A SIMULATION STUDY AMENA SULTANA Centre for Equity and Health Systems (CEHS), ICDDR,B, Dhaka 1212, Bangladesh Email: amena@icddrb.org JAHIDA GULSHAN Institute of Statistical Research and Training (ISRT) University of Dhaka, Dhaka 1000, Bangladesh Email: gulshan@isrt.ac.bd SUMMARY Quantile regression presents a flexible approach to the analysis of survival data, allowing for modeling quantile-specific covariate effect. Qian and Peng (2010) proposed profile es- timating equations and a readily and stably implemented iterative algorithm for censored quantile regression tailored to the partially functional effect setting with a mixture of vary- ing and constant effects and demonstrated improved efficiency of estimation over a naive two stage procedure. The aim of this study is to use the same algorithm on a quantile regres- sion setting where some covariate effects follow general parametric pattern (e.g. normal, gamma or logistic distribution) rather than a constant function or value and to determine the strength of using the algorithm in such regression settings through simulation. Simulation studies demonstrate that the method works well, for moderately censored data, if the para- metric pattern g(.) is a known function with unknown parameter(s). A sensitivity analysis is performed to check the consequences of misspecification of such parametric pattern. Keywords and phrases: Quantile regression; Censored data; Parametric pattern; Efficiency; Sensitivity analysis. 1 Introduction Regression quantiles, a new class of statistics is a simple minimization problem yielding the ordi- nary sample quantiles in the location model (Koenker and Bassett, 1978). Quantile regression, first proposed by Koenker and Bassett (1978), has emerged as a significant extension of classic linear regression by seminally using the concept of conditional quantiles. Quantile regression has great flexibility and straightforward interpretation in assessing covariate effects on event times, resulting in growing interests in its applications in survival analysis. The quantile regression is more flex- ible because the effect of covariates is not restricted to be constant in contrast to the accelerated c  Institute of Statistical Research and Training (ISRT), University of Dhaka, Dhaka 1000, Bangladesh.