Cook's distance (Cook, 1977) is often used to assess the influence of single observations or sets of observations in the linear regression model, and it is available, at least for single observations, in many statistical packages. After calculating Cook's distance for a set of observations, one needs some guidelines to flag influential sets deserving special attention. Cook (1977) suggested guidelines based on a confidence ellipsoid for the regression parameter β. In this paper, we suggest reference values for Cook's distance via Monte Carlo simulation. Also, a relative influence measure using Cook's distance is defined and its reference values are given.
No takes yet. Share an insight, caveat, or question.
Kim et al. (1996) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: