Graphical Methods for Influence Diagnostics

  • Dae Heung Jang (Department of Applied Mathematics, Pukyong National University, Pusan, 608-737, Korea)
  • Published : 1997.08.01

Abstract

Unusual observations can greatly influence the results of least wquares estimation. I propose graphical methods which can detect the influential observations.

Keywords

References

  1. An Introduction to Regression Graphics Cook, R.D.;Weisberg, S.
  2. Classical and Modern Regression with Applications Myers, R.H.
  3. Methods and Applicationsof Linear models Hocking, R.R.
  4. Communications in Statistics-Theory and Methods v.12 The regression dilemma Hocking, R.R.;Pendleton, O.J.
  5. Technometrics v.31 Regression Diagnosis with dynamic graphics Cook, R.D.;Weisberg, S.
  6. Residuals and Influence in Regression Cook, R.D.;Weisberg, S.
  7. Diagnostics : Identifying Influential Data and Sources of Collinearity Belsley, D.A.;Kuh, E.;Welsch, R.E.
  8. Biplots Gower, J.C.;Hand, D.J.
  9. Biometrika v.58 The biplot-graphic display of matrices woth application to principal component analysis Gabriel, K.R.