Menu Content/Inhalt
Seminars Print
previous year previous month next month next year
See by year See by month See by week See Today Search Jump to month
Valentin Patilea, INSA, Rennes Print
Thursday, 05 November 2009, 12:15 - 13:15


Testing conditional moment restrictions in the presence of right-censoring of the dependent variable.

Much of the recent statistics, biostatistics and econometric literature addressed the problem of parameters estimation in conditional moment restriction (CMR) models in the presence of censored observations. Parametric censored regression and censored quantile regression models are benchmark examples. The problem we consider in this paper is the specification testing of the CMR against nonparametric alternatives. The novelty is that the dependent variable could be right-censored. Starting from a classical kernel based test statistic and using suitable inverse probability of censoring weights, we show how to build a test statistic for general CMR in the presence of censoring. Depending on the identification assumption considered, two types of weights are proposed based on Kaplan-Meier (resp. smoothed Kaplan-Meier) estimator of the distribution (resp. conditional distribution) of the censoring variable.

Valentin Patilea, INSA, Université de Rennes

Location: S 12.227
Contact: Claude Adan, This e-mail address is being protected from spam bots, you need JavaScript enabled to view it