The Annals of Statistics

Testing Goodness-of-Fit in Regression Via Order Selection Criteria

R. L. Eubank and Jeffrey D. Hart

Full-text: Open access

Abstract

A new test is derived for the hypothesis that a regression function has a prescribed parametric form. Unlike many recent proposals, this test does not depend on arbitrarily chosen smoothing parameters. In fact, the test statistic is itself a smoothing parameter which is selected to minimize an estimated risk function. The exact distribution of the test statistic is obtained when the error terms in the regression model are Gaussian, while the large sample distribution is derived for more general settings. It is shown that the proposed test is consistent against fixed alternatives and can detect local alternatives that converge to the null hypothesis at the rate $1/\sqrt n$, where $n$ is the sample size. More importantly, the test is shown by example to have an ability to adapt to the alternative at hand.

Article information

Source
Ann. Statist., Volume 20, Number 3 (1992), 1412-1425.

Dates
First available in Project Euclid: 12 April 2007

Permanent link to this document
https://projecteuclid.org/euclid.aos/1176348775

Digital Object Identifier
doi:10.1214/aos/1176348775

Mathematical Reviews number (MathSciNet)
MR1186256

Zentralblatt MATH identifier
0776.62045

JSTOR
links.jstor.org

Subjects
Primary: 62G10: Hypothesis testing
Secondary: 62E20: Asymptotic distribution theory 62J99: None of the above, but in this section

Keywords
Nonparametric regression smoothing parameter selection risk estimation

Citation

Eubank, R. L.; Hart, Jeffrey D. Testing Goodness-of-Fit in Regression Via Order Selection Criteria. Ann. Statist. 20 (1992), no. 3, 1412--1425. doi:10.1214/aos/1176348775. https://projecteuclid.org/euclid.aos/1176348775


Export citation