## Abstract

Hotelling's T

^{2}tests are the classical testing procedure for one- and two-sample location problems under Gaussian assumptions. Although asymptotically valid under non-Gaussian distributions with finite second-order moments, they are highly nonrobust. Robust and rank-based alternatives, resisting violations of the finite second-order moment assumption, are described.Original language | English |
---|---|

Title of host publication | Encyclopedia of Environmetrics, 2nd Edition |

Editors | W. Piegorsch, A. El Shaarawi |

Publisher | Wiley |

Pages | 910-915 |

Number of pages | 3510 |

ISBN (Print) | 9780470973882 |

Publication status | Published - 2012 |

## Fingerprint

Dive into the research topics of 'Hotelling's T^{2}tests (robust versions of)'. Together they form a unique fingerprint.