Abstract
Reiser (2001) proposes a method of forming confidence interval for a Mahalanobis distance that yields intervals which have exactly the nominal coverage, but sometimes the interval is (0,0). We consider the case where Mahalanobis distance quantifies the difference between an individual and a population mean, and suggest a modification that avoids implausible intervals.
Original language | English |
---|---|
Pages (from-to) | 131-137 |
Number of pages | 7 |
Journal | Statistics and Probability Letters |
Volume | 127 |
Early online date | 12 Apr 2017 |
DOIs | |
Publication status | Published - Aug 2017 |
Bibliographical note
The work reported here was supported by a project grant from the Medical Research Council, Grant Number MR/J013838/1.Keywords
- Credible interval
- Noncentral F
- Objective prior