000198639 001__ 198639
000198639 005__ 20181203023504.0
000198639 0247_ $$2doi$$a10.1080/01621459.2013.839451
000198639 022__ $$a0162-1459
000198639 02470 $$2ISI$$a000333787300024
000198639 037__ $$aARTICLE
000198639 245__ $$aAccurate Directional Inference for Vector Parameters in Linear Exponential Families
000198639 260__ $$bAmerican Statistical Association$$c2014$$aAlexandria
000198639 269__ $$a2014
000198639 300__ $$a13
000198639 336__ $$aJournal Articles
000198639 520__ $$aWe consider inference on a vector-valued parameter of interest in a linear exponential family, in the presence of a finite-dimensional nuisance parameter. Based on higher-order asymptotic theory for likelihood, we propose a directional test whose p-value is computed using one-dimensional integration. The work simplifies and develops earlier research on directional tests for continuous models and on higher-order inference for discrete models, and the examples include contingency tables and logistic regression. Examples and simulations illustrate the high accuracy of the method, which we compare with the usual likelihood ratio test and with an adjusted version due to Skovgaard. In high-dimensional settings, such as covariance selection, the approach works essentially perfectly, whereas its competitors can fail catastrophically.
000198639 6531_ $$aContingency table
000198639 6531_ $$aCovariance selection
000198639 6531_ $$aExponential family model
000198639 6531_ $$aHigher-order asymptotics
000198639 6531_ $$aLikelihood ratio test
000198639 6531_ $$aLogistic regression
000198639 700__ $$0240476$$g111184$$uEcole Polytech Fed Lausanne, EPFL FSB MATHAA STAT, CH-1015 Lausanne, Switzerland$$aDavison, A. C.
000198639 700__ $$uUniv Toronto, Dept Stat, Toronto, ON M5S 3G3, Canada$$aFraser, D. A. S.
000198639 700__ $$uUniv Toronto, Dept Stat, Toronto, ON M5S 3G3, Canada$$aReid, N.
000198639 700__ $$aSartori, N.
000198639 773__ $$j109$$tJournal Of The American Statistical Association$$k505$$q302-314
000198639 909C0 $$xU10124$$0252136$$pSTAT
000198639 909CO $$pSB$$particle$$ooai:infoscience.tind.io:198639
000198639 917Z8 $$x111184
000198639 937__ $$aEPFL-ARTICLE-198639
000198639 973__ $$rREVIEWED$$sPUBLISHED$$aEPFL
000198639 980__ $$aARTICLE