Testing independence of covariates and errors in nonparametric regression

Sankar, Subhra, Bergsma, Wicher ORCID: 0000-0002-2422-2359 and Dassios, Angelos ORCID: 0000-0002-3968-2366 (2017) Testing independence of covariates and errors in nonparametric regression. Scandinavian Journal of Statistics, 45 (3). pp. 421-443. ISSN 0303-6898

Preview

Text (Dassios_Testing independence_2017.pdf) - Accepted Version Download (1MB) | Preview

• Author
• Author
• Publisher

Abstract

Consider a nonparametric regression model Y = m(X)+✏, where m is an unknown regression function, Y is a real-valued response variable, X is a real co-variate, and ✏ is the error term. In this article, we extend the usual tests for homoscedasticity by developing consistent tests for independence between X and ✏. Further, we investigate the local power of the proposed tests using Le Cam’s contiguous alternatives. An asymptotic power study under local alternatives along with extensive finite sample simulation study shows the performance of the new tests is competitive with existing ones. Furthermore, the practicality of the new tests is shown using two real data sets.

Item Type:	Article
Official URL:	http://onlinelibrary.wiley.com/journal/10.1111/(IS...
Additional Information:	© 2017 Board of the Foundation of the Scandinavian Journal of Statistics
Divisions:	Statistics
Subjects:	H Social Sciences > HA Statistics
Date Deposited:	16 Aug 2017 14:27
Last Modified:	12 Dec 2024 01:31
URI:	http://eprints.lse.ac.uk/id/eprint/83780

Actions (login required)

View Item