Projection Tests

  • Thorsten Dickhaus


We deal with the empirical likelihood (EL) method for testing vector means. A Wilks-type phenomenon for the empirical likelihood ratio test statistic is proved, by means of which a corresponding test can be calibrated. We embed the EL method into a broad class of tests relying on projections of the empirical measure into the space of distributions defined by the null hypothesis. Specific examples like the exponential tilting method are discussed.


  1. Bishop YM, Fienberg SE, Holland PW (2007) Discrete multivariate analysis. Theory and practice. With the collaboration of Richard J. Light and Frederick Mosteller. Reprint of the 1975 edition. Springer, Berlin.
  2. Chandra TK (2012) The Borel-Cantelli lemma. Springer, New York, NY.
  3. Cressie N, Read TR (1984) Multinomial goodness-of-fit tests. J R Stat Soc Ser B 46:440–464Google Scholar
  4. Csiszar I (1975) I-divergence geometry of probability distributions and minimization problems. Ann Probab 3:146–158.
  5. Dickhaus T (2015) Self-concordant profile empirical likelihood ratio tests for the population correlation coefficient: a simulation study. In: Stochastic models, statistics and their applications. Collected papers based on the presentations at the 12th workshop, Wrocław, Poland, February 2015, Springer, Cham, pp 253–260.
  6. Efron B (1981) Nonparametric standard errors and confidence intervals. Can J Stat 9:139–172.
  7. Li X, Chen J, Wu Y, Tu D (2011) Constructing nonparametric likelihood confidence regions with high order precisions. Stat Sin 21(4):1767–1783.
  8. Owen AB (2001) Empirical likelihood. Chapman & Hall/CRC, Boca Raton, FLGoogle Scholar
  9. Schennach SM (2007) Point estimation with exponentially tilted empirical likelihood. Ann Stat 35(2):634–672. MathSciNetCrossRefzbMATHGoogle Scholar
  10. Wilks S (1938) The large-sample distribution of the likelihood ratio for testing composite hypotheses. Ann Math Stat 9:60–62. CrossRefzbMATHGoogle Scholar

Copyright information

© Springer International Publishing AG, part of Springer Nature 2018

Authors and Affiliations

  • Thorsten Dickhaus
    • 1
  1. 1.Institute for StatisticsUniversity of BremenBremenGermany

Personalised recommendations