Some Methodological Aspects of Validation of Models in Nonparametric Regression

In this paper we describe some general methods for constructing goodness of fit tests in nonparametric regression models. Our main concern is the development of statisticial methodology for the assessment (validation) of specific parametric models ℳ as they arise in various fields of applications. The fundamental idea which underlies all these methods is the investigation of certain goodness of fit statistics (which may depend on the particular problem and may be driven by different criteria) under the assumption that a specified model (which has to be validated) holds true as well as under a broad range of scenaria, where this assumption is violated. This is motivated by the fact that outcomes of tests for the classical hypothesis: “The model ℳ holds true” (and their associated p values) bear various methodological flaws. Hence, our suggestion is always to accompany such a test by an analysis of the type II error, which is in goodness of fit problems often the more serious one. We give a careful description of the methodological aspects, the required asymptotic theory, and illustrate the main principles in the problem of testing model assumptions such as a specific parametric form or homoscedasticity in nonparametric regression models.

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

[2]  Joseph Berkson,et al.  Some Difficulties of Interpretation Encountered in the Application of the Chi-Square Test , 1938 .

[3]  J. Neumann.  Distribution of the Ratio of the Mean Square Successive Difference to the Variance , 1941 .

[4]  John von Neumann,et al.  A Further Remark Concerning the Distribution of the Ratio of the Mean Square Successive Difference to the Variance , 1942 .

[5]  Joseph Berkson,et al.  Tests of significance considered as evidence , 1942 .

[6]  Experience with Tests of Significance: A Reply to Professor R. A. Fisher , 1943 .

[7]  J. L. Hodges,et al.  Testing the Approximate Validity of Statistical Hypotheses , 1954 .

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

[9]  Theo Gasser,et al.  Smoothing Techniques for Curve Estimation , 1979 .

[10]  T. Breurch,et al.  A simple test for heteroscedasticity and random coefficient variation (econometrica vol 47 , 1979 .

[11]  M. J. Harrison,et al.  A Test for Heteroscedasticity Based on Ordinary Least Squares Residuals , 1979 .

[12]  H. Bunke,et al.  Asymptotic results on nonlinear approximation of regression functions and weighted least squares , 1980 .

[13]  A note on parameter estimation in inadequate nonlinear regression models , 1981 .

[14]  R. Koenker,et al.  Robust Tests for Heteroscedasticity Based on Regression Quantiles , 1982 .

[15]  S. Weisberg,et al.  Diagnostics for heteroscedasticity in regression , 1983 .

[16]  J. Rice.  Bandwidth Choice for Nonparametric Regression , 1984 .

[17]  H. Müller,et al.  Kernels for Nonparametric Curve Estimation , 1985 .

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

[19]  T. Gasser,et al.  Residual variance and residual pattern in nonlinear regression , 1986 .

[20]  J. Berger,et al.  Testing Precise Hypotheses , 1987 .

[21]  Ulrich Stadtmüller,et al.  Estimation of Heteroscedasticity in Regression Analysis , 1987 .

[22]  D. Ruppert,et al.  Transformation and Weighting in Regression , 1988 .

[23]  Grace Wahba,et al.  Testing the (Parametric) Null Model Hypothesis in (Semiparametric) Partial and Generalized Spline Models , 1988 .

[24]  Herman J. Bierens,et al.  A consistent conditional moment test of functional form , 1990 .

[25]  Clifford H. Spiegelman,et al.  Testing the Goodness of Fit of a Linear Model via Nonparametric Regression Techniques , 1990 .

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

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

[28]  David Firth,et al.  Model checking with nonparametric curves , 1991 .

[29]  A. Kozek.  A nonparametric test of fit of a parametric model , 1991 .

[30]  P. Hall,et al.  On estimation of noise variance in two-dimensional signal processing , 1991, Advances in Applied Probability.

[31]  Terry L King.  Smooth Tests of Goodness of Fit , 1991 .

[32]  Thomas A. Severini,et al.  Diagnostics for Assessing Regression Models , 1991 .

[33]  Hans-Georg Müller,et al.  Goodness-of-fit diagnostics for regression models , 1992 .

[34]  Jeffrey D. Hart,et al.  Kernel Regression When the Boundary Region is Large, with an Application to Testing the Adequacy of Polynomial Models , 1992 .

[35]  Adonis Yatchew,et al.  Nonparametric Regression Tests Based on Least Squares , 1992, Econometric Theory.

[36]  James G. MacKinnon,et al.  Model Specification Tests and Artificial Regressions , 1992 .

[37]  Jianqing Fan.  Design-adaptive Nonparametric Regression , 1992 .

[38]  Jeffrey M. Wooldridge,et al.  A Test for Functional Form Against Nonparametric Alternatives , 1992, Econometric Theory.

[39]  R. L. Eubank,et al.  Testing Goodness-of-Fit in Regression Via Order Selection Criteria , 1992 .

[40]  Vincent N. LaRiccia,et al.  Testing goodness of fit via nonparametric function estimation techniques , 1993 .

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

[42]  Gerhard Weihrather.  Testing a linear regression model against nonparametric alternatives , 1993 .

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

[44]  P. Gozalo,et al.  A Consistent Model Specification Test for Nonparametric Estimation of Regression Function Models , 1993, Econometric Theory.

[45]  F. Brodeau.  Test for the choice of approximative models in nonlinear regression when the variance is unknown , 1993 .

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

[47]  R. L. Eubank,et al.  Detecting Heteroscedasticity in Nonparametric Regression , 1993 .

[48]  Juei-Chao Chen.  Testing for no effect in nonparametric regression via spline smoothing techniques , 1994 .

[49]  Jianqing Fan,et al.  Local polynomial modelling and its applications , 1994 .

[50]  A cubic smoothing spline based lack of fit test for nonlinear regression models , 1995 .

[51]  Hans-Georg Müller,et al.  On a Semiparametric Variance Function Model and a Test for Heteroscedasticity , 1995 .

[52]  D G Altman,et al.  Statistics notes: Absence of evidence is not evidence of absence , 1995 .

[53]  J. Diebolt.  A nonparametric test for the regression function : asymptotic theory , 1995 .

[54]  W. Manteiga Gonzalez,et al.  Testing linear regression models using non-parametric regression estimators when errors are non-independent , 1995 .

[55]  Bodhini R. Jayasuriya,et al.  Testing for Polynomial Regression Using Nonparametric Regression Techniques , 1996 .

[56]  Jianqing Fan.  Test of Significance Based on Wavelet Thresholding and Neyman's Truncation , 1996 .

[57]  J. Hart,et al.  Smoothing-based lack-of-fit tests: variations on a theme , 1996 .

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

[59]  W. Stute,et al.  NN goodness-of-fit tests for linear models , 1996 .

[60]  Winfried Stute,et al.  Nonparametric model checks for regression , 1997 .

[61]  L. Brown,et al.  An unbiased test for the bioequivalence problem , 1997 .

[62]  Calibrated nonparametric confidence sets , 1997 .

[63]  Adrian Bowman,et al.  Testing for constant variance in a linear model , 1997 .

[64]  Jeffrey D. Hart,et al.  Nonparametric Smoothing and Lack-Of-Fit Tests , 1997 .

[65]  A Simple Goodness-of-fit Test for Linear Models Under a Random Design Assumption , 1998 .

[66]  Winfried Stute,et al.  Bootstrap Approximations in Model Checks for Regression , 1998 .

[67]  Holger Dette,et al.  Estimating the variance in nonparametric regression—what is a reasonable choice? , 1998 .

[68]  C. Czado,et al.  Nonparametric validation of similar distributions and assessment of goodness of fit , 1998 .

[69]  C. Robert,et al.  Model choice in generalised linear models: A Bayesian approach via Kullback-Leibler projections , 1998 .

[70]  Holger Dette,et al.  Validation of linear regression models , 1998 .

[71]  Holger Dette,et al.  Testing heteroscedasticity in nonparametric regression , 1998 .

[72]  Graham B. McBride,et al.  Applications: Equivalence Tests Can Enhance Environmental Science and Management , 1999 .

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

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

[75]  Holger Dette,et al.  Testing linearity of regression models with dependent errors by kernel based methods , 2000 .

[76]  Holger Dette,et al.  Testing model assumptions in multivariate linear regression models , 2000 .

[77]  Jianqing Fan,et al.  Goodness-of-Fit Tests for Parametric Regression Models , 2001 .

[78]  Joel L. Horowitz,et al.  An Adaptive, Rate-Optimal Test of a Parametric Mean-Regression Model Against a Nonparametric Alternative , 2001 .

[79]  Jianqing Fan,et al.  Generalized likelihood ratio statistics and Wilks phenomenon , 2001 .

[80]  On a problem in pharmaceutical statistics and the iteration of a peculiar nonlinear operator in the upper complex halfplane , 2001 .

[81]  Holger Dette,et al.  A consistent test for heteroscedasticity in nonparametric regression based on the kernel method , 2002 .