Comparing tests of autoregressive versus moving average errors in regression models using bahadur's asymptotic relative efficiency

Colin R Mckenzie, Michael McAlecr

Research output: Contribution to journalArticle


The purpose of this paper is to use Bahadur's asymptotic relative efficiency measure to compare the performance of various tests of autoregressive (AR) versus moving average (MA) error processes in regression models. Tests to be examined include non-nested procedures of the models against each other, and classical procedures based upon testing both the AR and MA error processes against the more general autoregressive-moving average model.

Original languageEnglish
Pages (from-to)1349-1371
Number of pages23
JournalCommunications in Statistics - Theory and Methods
Issue number8
Publication statusPublished - 2002 Aug
Externally publishedYes



  • Autoregressive model
  • Bahadur efficiency
  • Inappropriate alteratives
  • Lagrange multiplier test
  • Moving average model
  • Separate (non-nested) tests

ASJC Scopus subject areas

  • Statistics and Probability
  • Safety, Risk, Reliability and Quality

Cite this