SEMIPARAMETRIC ESTIMATION OF CENSORED TRANSFORMATION MODELS

Author: TUE G empty RGENS

Source: Journal of Nonparametric Statistics, Volume 15, Number 3, June 2003 , pp. 377-393(17)

Buy & download fulltext article:

Price: $56.94 plus tax (Refund Policy)

Abstract:

Many widely used models, including proportional hazards models with unobserved heterogeneity, can be written in the form $Lambda (Y) = min [betaˆ{prime} X + U, C]$</FORMULA>, where Lambda

is an unknown increasing function, the error term U has unknown distribution function Psi

and is independent of X, C is a random censoring threshold and U and C are conditionally independent given X. This paper develops new $nˆ{1/2}$</FORMULA>-consistent and asymptotically normal semiparametric estimators of Lambda

and

which are easier to use than previous estimators. Moreover, Monte Carlo results suggest that the mean integrated squared error of predictions based on the new estimators is lower than for previous estimators.

Keywords: Semiparametric estimation; Kernel regression; Transformation model; Unobserved heterogeneity; Duration analysis; Censoring

Document Type: Research article

DOI: http://dx.doi.org/10.1080/1048525031000120224

Affiliations: 1: Economics RSSS, Australian National University, Canberra ACT 0200, Australia

Publication date: 2003-06-01