The Likelihood Displacement: A Unifying Principle for Influence Measures.
Technical summary rept.,
WISCONSIN UNIV-MADISON MATHEMATICS RESEARCH CENTER
Pagination or Media Count:
The young field of statistical diagnostics has produced an array of competing statistics for measuring the influence of individual cases. The identification of influential cases seem generally accepted as an important part of linear regression analysis. Although there are many diagnostic methods available for this, two specific diagnostic statistics--d sub i as proprosed by Cook, and DFFITS sub i as proposed by Belsley, Kuh and Welsch 1980--appear to be used most frequently since they are available in many widely distributed regression packages. For further progress and a deeper understanding of available methodology, larger perspectives seem necessary.
- Statistics and Probability