Optimal designs for testing the functional form of a regression via nonparametric estimation techniques

For the problem of checking linearity in a heteroscedastic nonparametric regression model under a fixed design assumption we study maximin designs which maximize the minimum power of a nonparametric test over a broad class of alternatives from the assumed linear regression model. It is demonstrated that the optimal design depends sensitively on the used estimation technique (i.e. weighted or ordinary least squares) and on an inner product used in the definiton of the class of alternatives. Our results extend and put recent finndings of Wiens (1991) in a new light, who established the maximin optimality of the uniform design for lack-of-fit tests in homoscedastic multiple linear regression models.

[1]  H. Dette.  On a nonparametric test for linear relationships , 2000 .

[2]  Holger Dette,et al.  A consistent test for the functional form of a regression based on a difference of variance estimators , 1999 .

[3]  W. González-Manteiga,et al.  Goodness-of-fit test for linear models based on local polynomials , 1999 .

[4]  J. Zheng,et al.  A consistent test of functional form via nonparametric estimation techniques , 1996 .

[5]  E. Mammen,et al.  Comparing Nonparametric Versus Parametric Regression Fits , 1993 .

[6]  W. González-Manteiga,et al.  Testing the hypothesis of a general linear model using nonparametric regression estimation , 1993 .

[7]  H. Müller,et al.  On variance function estimation with quadratic forms , 1993 .

[8]  Raymond H. Myers,et al.  A new look at experimental design robustness , 1992 .

[9]  D. Wiens.  Designs for approximately linear regression: two optimality properties of uniform designs , 1991 .

[10]  P. Hall,et al.  Asymptotically optimal difference-based estimation of variance in nonparametric regression , 1990 .

[11]  J. Marron,et al.  On variance estimation in nonparametric regression , 1990 .

[12]  Raymond J. Carroll,et al.  Variance Function Estimation in Regression: the Effect of Estimating the Mean , 1988 .

[13]  James W. Neill,et al.  Testing Linear Regression Function Adequacy without Replication , 1985 .

[14]  W. J. Studden,et al.  Theory Of Optimal Experiments , 1972 .

[15]  J. Sacks,et al.  Designs for Regression Problems with Correlated Errors III , 1966 .

[16]  J. Cooper,et al.  Theory of Approximation , 1960, The Mathematical Gazette.

[17]  J. Kiefer,et al.  Optimum Designs in Regression Problems , 1959 .

[18]  H. Chernoff.  Locally Optimal Designs for Estimating Parameters , 1953 .

[19]  Radakovič.  The theory of approximation , 1932 .

[20]  Michael Jackson,et al.  Optimal Design of Experiments , 1994 .

[21]  Adrian Bowman,et al.  On the Use of Nonparametric Regression for Checking Linear Relationships , 1993 .

[22]  E. R. Shillington,et al.  Testing lack of fit in regression without replication , 1979 .

[23]  E. Nadaraya.  On Non-Parametric Estimates of Density Functions and Regression Curves , 1965 .

[24]  E. Nadaraya.  On Estimating Regression , 1964 .