Global Validation of Linear Model Assumptions

An easy-to-implement global procedure for testing the four assumptions of the linear model is proposed. The test can be viewed as a Neyman smooth test and relies only on the standardized residual vector. If the global procedure indicates a violation of at least one of the assumptions, then the components of the global test statistic can be used to gain insight into which assumptions have been violated. The procedure can also be used in conjunction with associated deletion statistics to detect unusual observations. Simulation results are presented indicating the sensitivity of the procedure in detecting model violations under a variety of situations, and its performance is compared with three potential competitors, including a procedure based on the Box–Cox power transformation. The procedure is demonstrated by applying it to a new car mileage dataset and a water salinity dataset that has been used earlier to illustrate model diagnostics.

[1]  J. Tukey.  One Degree of Freedom for Non-Additivity , 1949 .

[2]  J. Schmee.  An Introduction to Multivariate Statistical Analysis , 1986 .

[3]  S. Weisberg,et al.  Residuals and Influence in Regression , 1982 .

[4]  T. W. Anderson.  An Introduction to Multivariate Statistical Analysis, 2nd Edition. , 1985 .

[5]  R. Cook.  Detection of influential observation in linear regression , 2000 .

[6]  N. Hjort,et al.  Frequentist Model Average Estimators , 2003 .

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

[8]  A. Hossain,et al.  A comparative study on detection of influential observations in linear regression , 1991 .

[9]  Ross Ihaka,et al.  Gentleman R: R: A language for data analysis and graphics , 1996 .

[10]  John C. W. Rayner,et al.  Smooth test of goodness of fit , 1989 .

[11]  J. Durbin,et al.  Testing for serial correlation in least squares regression. I. , 1950, Biometrika.

[12]  H. Theil,et al.  Testing the Independence of Regression Disturbances , 1961 .

[13]  F. J. Anscombe,et al.  The Examination and Analysis of Residuals , 1963 .

[14]  D. Ruppert,et al.  Trimmed Least Squares Estimation in the Linear Model , 1980 .

[15]  N. Hjort,et al.  The Focused Information Criterion , 2003 .

[16]  Chan‐Fu Chen,et al.  Score Tests for Regression Models , 1983 .

[17]  William H. Swallow,et al.  A review of the development and application of recursive residuals in linear models , 1996 .

[18]  D. Cox,et al.  An Analysis of Transformations , 1964 .

[19]  H. Akaike,et al.  Information Theory and an Extension of the Maximum Likelihood Principle , 1973 .

[20]  Olivier Thas,et al.  Smooth tests of goodness of fit , 1989 .

[21]  G. Schwarz.  Estimating the Dimension of a Model , 1978 .

[22]  D. Pierce.  The Asymptotic Effect of Substituting Estimators for Parameters in Certain Types of Statistics , 1982 .

[23]  D. F. Andrews,et al.  Significance tests based on residuals , 1971 .

[24]  H. Theil.  The Analysis of Disturbances in Regression Analysis , 1965 .

[25]  John C. W. Rayner,et al.  Neyman-type smooth tests for location-scale families , 1986 .

[26]  P. Bickel,et al.  Using Residuals Robustly I: Tests for Heteroscedasticity, Nonlinearity , 1975 .

[27]  Chan‐Fu Chen,et al.  Robustness Aspects of Score Tests for Generalized Linear and Partially Linear Regression Models , 1985 .

[28]  Edsel A. Peña,et al.  Variance Estimation in a Model With Gaussian Submodels , 2005, Journal of the American Statistical Association.

[29]  J. Durbin,et al.  Testing for serial correlation in least squares regression. II. , 1950, Biometrika.

[30]  E. J. Hannan.  TESTING FOR SERIAL CORRELATION IN LEAST SQUARES REGRESSION , 1957 .

[31]  Nils Lid Hjort,et al.  Discussion: Rejoinder to 'The focussed information criterion' and 'Frequentist model averaging' , 2003 .

[32]  V. Barnett,et al.  Applied Linear Statistical Models , 1975 .

[33]  John W. Tukey,et al.  Exploratory Data Analysis. , 1979 .

[34]  R. Cook.  Regression Graphics , 1994 .

[35]  Donald A. Pierce,et al.  Neyman's Smooth Goodness-of-Fit Test When the Hypothesis Is Composite , 1979 .

[36]  F. J. Anscombe,et al.  Examination of Residuals , 1961 .

[37]  J. Doornik,et al.  An Omnibus Test for Univariate and Multivariate Normality , 2008 .