TITLE

Goodness of Fit in Regression Analysis - R² and G² Reconsidered

AUTHOR(S)
Hagquist, Curt; Stenbeck, Magnus
PUB. DATE
August 1998
SOURCE
Quality & Quantity;Aug98, Vol. 32 Issue 3, p229
SOURCE TYPE
Academic Journal
DOC. TYPE
Article
ABSTRACT
There has been considerable debate on how important goodness of fit is as a tool in regression analysis, especially with regard to the controversy on R² in linear regression. This article reviews some of the arguments of this debate and its relationship to other goodness of fit measures. It attempts to clarify the distinction between goodness of fit measures and other model evaluation tools as well as the distinction between model test statistics and descriptive measures used to make decisions on the agreement between models and data. It also argues that the utility of goodness of fit measures depends on whether the analysis focuses on explaining the outcome (model orientation) or explaining the effect(s) of some regressor(s) on the outcome (factor orientation). In some situations a decisive goodness of fit test statistic exists and is a central tool in the analysis. In other situations, where the goodness of fit measure is not a test statistic but a descripitive measure, it can be used as a heuristic device along with other evidence whenever appropriate. The availability of goodness of fit test statistics depends on whether the variability in the observations is restricted, as in table analysis, or whether it is unrestricted, as in OLS and logistic regression on individual data. Hence, G² is a decisive tool for measuring goodness of fit, whereas R² and SEE are heuristic tools.
ACCESSION #
10191640

 

Related Articles

  • The Anderson-Darline Test for Goodness of Fit.  // CORS Journal;Mar68, Vol. 6 Issue 1, p65 

    This article focuses on the Anderson-Darling test for goodness of fit. The test for goodness of fit between an observed and a theoretical cumulative distribution function is illustrated with an example, and its advantages in relation to other tests are indicated.

  • THE KOLMOGOROV GOODNESS-OF-FIT TEST OF INDEPENDENCE BASED ON COPULAS. Güven, Bilgehan // Communications Series A1 Mathematics & Statistics;2003, Vol. 52 Issue 2, p21 

    We present a method which reduces the Kolmogorov goodness-of-test of independence to the Kolmogorov-Smimov one sample test. The null distribution of test statistic is the same as the Kolmogorov-Smimov test statistic in this test of independence.

  • Finding local departures from a parametric model using nonparametric regression. Opsomer, J. D.; Francisco-Fernández, M. // Statistical Papers;Jan2010, Vol. 51 Issue 1, p69 

    Goodness-of-fit evaluation of a parametric regression model is often done through hypothesis testing, where the fit of the model of interest is compared statistically to that obtained under a broader class of models. Nonparametric regression models are frequently used as the latter type of...

  • Goodness-of-Fit Tests for the Power Function Distribution Based on the Puri-Rubin Characterization and Their Efficiences. Volkova, K.; Nikitin, Ya. // Journal of Mathematical Sciences;May2014, Vol. 199 Issue 2, p130 

    We construct integral and supermum type goodness-of-fit tests for the family of power distribution functions. Test statistics are functionals of U-empirical processes and are based on the classical characterization of the power function distribution family belonging to Puri and Rubin. We...

  • Goodness-of-Fit Tests Using Dual Versions of Characterizations via Moments of Record Values. Morris, K. W.; Szynal, D. // Journal of Mathematical Sciences;Aug2004, Vol. 122 Issue 4, p3384 

    Focuses on goodness-of-fit tests using dual versions of characterizations via moments of record values. Sequence of independent and identically distributed random variables with cumulative distribution function and probability distribution function; Discusses of related theorems, corollaries and...

  • Goodness-of-Fit Tests Using Dual Versions of Characterizations via Moments of Order Statistics. Morris, K. W.; Szynal, D. // Journal of Mathematical Sciences;Aug2004, Vol. 122 Issue 4, p3365 

    Focuses on goodness-of-fit tests using dual versions of characterizations via moments of order statistics. Use of dual versions of characterizations to construct good-of-fit tests; Discusses of related theorems, corollaries and equation.

  • Two criteria of goodness-of-fit for Cauchy distributions based on characterizations. Litvinova, V. // Journal of Mathematical Sciences;May2005, Vol. 127 Issue 1, p1752 

    In this paper, we construct two criteria of goodness-of-fit for Cauchy distributions. These critiria are based on characterizations. The first characterization is due to Arnold, and the second one goes back to Ramachandran. The Bahadur efficiency for these criteria is found. Bibliography: 10 titles.

  • Tests Resulting from Characterizations Using Record Values. Morris, K. W.; Szynal, D. // Journal of Mathematical Sciences;Dec2005, Vol. 131 Issue 3, p5646 

    Focuses on the construction of goodness-of-fit tests resulting from characterizations via moments of the kth record values using U-statistics. Implication of the characterization; Estimation of the expectations by the U-statistics; Mean vector and variance matrix of the null distribution.

  • Characterizations of Continuous Distributions and Associated Goodness of Fit Tests. Morris, K. W.; Szynal, D. // Journal of Mathematical Sciences;Dec2005, Vol. 131 Issue 3, p5630 

    Discusses the characterizations of continuous distributions and associated goodness-of-fit tests. Probability design function; Test procedures; Dual test statistics arising from the characterization; Analysis of the behavior of the statistics.

Share

Read the Article

Courtesy of THE LIBRARY OF VIRGINIA

Sorry, but this item is not currently available from your library.

Try another library?
Sign out of this library

Other Topics