2337 shaares
J'aime bien la première réponse de Ripley, au sujet des tests sur la surdispersion: "There are, but like formal tests for outliers I would not advise using them, as you may get misleading inferences before they are significant, and they can reject when the inferences from the small model are perfectly adequate".
The moment estimator \phi of over-dispersion (calculé par le rapport Chi-2/ddl) gives you an indication of the likely effects on your inferences: e.g. estimated standard errors are proportional to \sqrt(\phi). having standard errors which need inflating by 40% seems to indicate that the rule you quote is too optimistic (even when its estimator is reliable).
Edit: remarque de Ripley concernant l'article de Lindsey: "And I can add 'it is helpful to know whose authority is being invoked', since (e.g.) some authors are not at all careful." Dommage que la critique ne soit pas plus détaillée.
The moment estimator \phi of over-dispersion (calculé par le rapport Chi-2/ddl) gives you an indication of the likely effects on your inferences: e.g. estimated standard errors are proportional to \sqrt(\phi). having standard errors which need inflating by 40% seems to indicate that the rule you quote is too optimistic (even when its estimator is reliable).
Edit: remarque de Ripley concernant l'article de Lindsey: "And I can add 'it is helpful to know whose authority is being invoked', since (e.g.) some authors are not at all careful." Dommage que la critique ne soit pas plus détaillée.