Title: Stein-rule estimation in genetic carrier testing

Authors: Tong Zeng; R. Carter Hill

Addresses: Department of Applied Business Sciences and Economics, University of La Verne, La Verne, CA 91750, USA ' Department of Economics, Louisiana State University, Baton Rouge, LA 70803, USA

Abstract: In this paper, we apply the fully correlated random parameters logit (FCRPL) model to the genetic carrier testing data using shrinkage estimation. We show that shrinkage estimates with higher shrinkage constant improve the percentages of correct predicted choices by 2% and 10% respectively with Jewish and general population samples. The mean estimates of elasticity based on the shrinkage estimates are closer to those with the FCRPL model estimates and have smaller standard errors than the corresponding results based on the uncorrelated random parameters logit model estimates.

Keywords: pretest estimator; positive-part Stein-like estimator; likelihood ratio test; random parameters logit model.

DOI: 10.1504/IJCEE.2020.107369

International Journal of Computational Economics and Econometrics, 2020 Vol.10 No.2, pp.111 - 128

Received: 27 Feb 2017
Accepted: 27 Nov 2017
Published online: 21 May 2020 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article