Detection of a random alteration in a multivariate observation when knowing probable direction

Barry P. Katz, Morton B. Brown

Research output: Contribution to journalArticle

1 Scopus citations


An observation from a multivariate distribution may be subject to perturbation in a known subset of the variables. A likelihood radio statistic is developed to test whether or not there has been an addition of a random quantity in a prespecified direction to an observation from a multivariate normal distribution. When the variance of this addition is unknown, a secondarily Bayes approach is used to eliminate this variance which acts as a nuisance parameter. The testing procedure is based on a distribution-free tolerance interval.

Original languageEnglish (US)
Pages (from-to)145-155
Number of pages11
JournalComputational Statistics and Data Analysis
Issue number2
StatePublished - Mar 1988



  • Outlier detection
  • Quality control
  • Random shift in location
  • Secondarily Bayes approach

ASJC Scopus subject areas

  • Computational Theory and Mathematics
  • Statistics, Probability and Uncertainty
  • Electrical and Electronic Engineering
  • Computational Mathematics
  • Numerical Analysis
  • Statistics and Probability

Cite this