作者：Robert, CP; Aitkin, M; Cox, DR; Stephens, M; Polymenis, A; Gilks, WR; Nobile, A; Hodgson, M; OHagan, A; Longford, NT; Dawid, AP; Atkinson, AC; Bernardo, JM; Besag, J; Brooks, SP; Byers, S; Raftery, A; Celeux, G; Cheng, RCH; Liu, WB; Chien, YH; George, EI; Cressie, N; Huang, HC; Gruet, MA; Heath, SC; Jennison, C; Lawson, AB; Clark, A; McLachlan, G; Peel, D; Mengersen, K; George, A; Philippe, A; Roeder, K; Wasserman, L; Schlattmann, P; Bohning, D; Titterington, DM; Tong, H; West, M
作者单位：University of Newcastle; University of Oxford; University of Glasgow; MRC Biostatistics Unit; University of Bristol; University of Nottingham; De Montfort University; University of London; University College London; University of London; London School Economics & Political Science; University of Valencia; University of Washington; University of Washington Seattle; University of Kent; University of Texas System; University of Texas Austin; Iowa State University; Universite Paris Saclay; INRAE; University of Bath; University of Queensland; University of Abertay Dundee; Queensland University of Technology (QUT); Universite de Rouen Normandie; Carnegie Mellon University; Free University of Berlin; Duke University
作者：Fung, WK; Kwan, CW
作者单位：University of Hong Kong
摘要：Object functions other than the likelihood displacement, such as a parameter estimate or a test statistic, can also be used in local influence analysis. The normal curvatures of these object functions have been studied in situations where the slopes were non-zero. In these situations, we show that the normal curvature is not scale invariant and thus ambiguous conclusions will be drawn. Comments on the application of the general normal curvature formula are presented.
