DOI QR코드

DOI QR Code

Influential Points in GLMs via Backwards Stepping

  • Jeong, Kwang-Mo (Department of Statistics, Research Institute of Computer, Information and Communication, Pusan National University) ;
  • Oh, Hae-Young (Department of Statistics, Pusan National University)
  • 발행 : 2002.04.01

초록

When assessing goodness-of-fit of a model, a small subset of deviating observations can give rise to a significant lack of fit. It is therefore important to identify such observations and to assess their effects on various aspects of analysis. A Cook's distance measure is usually used to detect influential observation. But it sometimes is not fully effective in identifying truly influential set of observations because there may exist masking or swamping effects. In this paper we confine our attention to influential subset In GLMs such as logistic regression models and loglinear models. We modify a backwards stepping algorithm, which was originally suggested for detecting outlying cells in contingency tables, to detect influential observations in GLMs. The algorithm consists of two steps, the identification step and the testing step. In identification step we Identify influential observations based on influencial measures such as Cook's distances. On the other hand in testing step we test the subset of identified observations to be significant or not Finally we explain the proposed method through two types of dataset related to logistic regression model and loglinear model, respectively.

키워드

참고문헌

  1. Categorical Data Analysis Agresti, A.
  2. Biometika v.68 Two Graphical Displays for Qutlying and Inflential Observations in Regression Atkinson, A. C. https://doi.org/10.1093/biomet/68.1.13
  3. Refression Diagnostics: Identifying Influential Data Sources of Collinearity Belsley, D. A.;Kuh, E.;Welsch, R. E.
  4. Statistical Models in S Chambers, J. M.;Hastie, T. J.
  5. Technometrics v.19 Detection of Influential Observations in Linear Regression Cook, R. D. https://doi.org/10.2307/1268249
  6. Technometrics v.22 Characterizations of an Empirical Influence Function for Detecting Influential Cases in Regression Cook, R. D.;Weisberg, S. https://doi.org/10.2307/1268187
  7. Residuals and Influence in Regression Cook, R. D.;Weisberg, S.
  8. Biometrika v.34 The Estimation from Individual Records of the Relationship Between Dose and Quantal Respose Finney, D. J. https://doi.org/10.1093/biomet/34.3-4.320
  9. American Statistician v.32 The Hat Matrix in Regression and ANOVA Hoaglin, D. C.;Welsch, R. E. https://doi.org/10.2307/2683469
  10. Generalized Linear Models McCullagh, P.;Nelder, J. A.
  11. The Annals of Statistics v.9 Logistic Regression Diagnostics Pregibon, D. https://doi.org/10.1214/aos/1176345513
  12. Technometrics v.17 On the Detection of Many Outliers Rosner, B. https://doi.org/10.2307/1268354
  13. Technometrics v.10 Detecting Outlying Cells in Two-Way Contingency Tables via Backwards Stepping Simonoff, J. S.