A nonparametric test for serial independence of regression errors

Authors: M. delgado¹; J. Mora²

Source: Biometrika, Volume 87, Number 1, 1 March 2000 , pp. 228-234(7)

Buy & download fulltext article:

Price: $35.78 plus tax (Refund Policy)

Abstract:

A test for serial independence of regression errors is proposed that is consistent in the direction of serial dependence alternatives of first order. The test statistic is a function of a Hoeffding-Blum-Kiefer-Rosenblatt type of empirical process, based on residuals. The resultant statistic converges, surprisingly, to the same limiting distribution as the corresponding statistic based on true errors.

Keywords: Empirical process based on residuals; Hoeffding-Bl

Document Type: Original article

Affiliations: 1: Departamento de Estadistica y Econometria, Universidad Carlos III de Madrid, Calle Madrid 126, Getafe, E-28903 Madrid, Spain 2: Departamento de Fundamentos del Análisis Económico, Universidad de Alicante, Apartado de Correos 99, E-03080 Alicante, Spain

Publication date: 2000-03-01