Diagnostic Displays for Assessing Leverage and Influence in Generalized Linear Models
- 1 September 1987
- journal article
- Published by Wiley in Australian Journal of Statistics
- Vol. 29 (3) , 233-243
- https://doi.org/10.1111/j.1467-842x.1987.tb00741.x
Abstract
Summary: A general technique for assessing leverage and influential observations in Generalized Linear Models is described. The procedure takes the form of Half‐Normal plots with envelopes derived from simulation to enhance overall assessment of the model. This procedure of assessment is more informative and provides additional insight compared with procedures based on the largest sample leverage and influence statistics. Application of the method is illustrated with an example in logistic regression.Keywords
This publication has 15 references indexed in Scilit:
- Small Sample Properties of Robust Analyses of Linear Models Based on R-Estimates: A SurveyPublished by Springer Nature ,1991
- Assessing Partial Influence in Generalized Linear ModelsBiometrics, 1988
- Residuals in Generalized Linear ModelsJournal of the American Statistical Association, 1986
- The Conditional Distribution of Goodness-of-Fit Statistics for Discrete DataJournal of the American Statistical Association, 1986
- K-Clustering as a Detection Tool for Influential Subsets in RegressionTechnometrics, 1984
- Efficient Bounded-Influence Regression EstimationJournal of the American Statistical Association, 1982
- Two graphical displays for outlying and influential observations in regressionBiometrika, 1981
- Influential Observations and Outliers in RegressionTechnometrics, 1981
- Characterizations of an Empirical Influence Function for Detecting Influential Cases in RegressionTechnometrics, 1980
- Generalized Linear ModelsJournal of the Royal Statistical Society. Series A (General), 1972