Behrens-Fisher Problem from a Model Selection Point of View

  • Jeon, Jong-Woo (Department of Computer Science and Statistics, Seoul National University) ;
  • Lee, Kee-Won (Department of Statistics, Hallym University)
  • Published : 1991.12.01

Abstract

Behrens-Fisher problem is viewed from a model selection approach. Normal distribution is regarded as an approximating model, A criterion, called TIC, is derived and is compared with selection criteria such as AIC and a bootstrap estimator. Stochastic approximation is used since no closed form expression is available for the bootstrap estimator.

Keywords