Abstract
This paper describes how to compute robust confidence intervals for differences of the effects using the likelihood ratio testF M in the two-way analysis of variance. The probability for the α-error and the average length of the confidence intervals withF m and the quadratic formQ M are investigated and compared with the classical confidence intervals fort-distributed and lognormal errors. We also give a warning of building confidence intervals withF M andQ M in the presence of heterogeneous scale parameters, because these tests which do not regard heteroscedasticity are then much too liberal.
Similar content being viewed by others
References
Hampel, F.R., Rousseeuw, P.J., Ronchetti, E.M., Stahel, W.A. (1986):Robust Statistics. John Wiley & Sons, Inc., New York
Huber, P.J. (1964):Robust Estimation of a Location Parameter. Ann. Math. Statist.35,73–101
Huber, P.J. (1973):Robust Regression: Asymptotics, Conjectures and Monte Carlo. The Annals of Statistics1,799–821
Huber P.J. (1977):Robust Statistical Procedures. Regional Conference Series in Applied Mathematics27, Society for Industrial and Applied Mathematics, Philadelphia, Pennsylvania
Huber, P.J. (1981):Robust Statistics. John Wiley & Sons, Inc., New York
Jeffreys, H. (1961):Theory of Probability. 3rd Edition, Clarendon Press, Oxford
Schrader, R.M., Hettmansperger, T.P. (1980):Robust Anslysis of Variance Based Upon a Likelihood Ratio Criterion. Biometrika67, 93–101
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Bachmaier, M., Precht, M. Robust confidence intervals for contrasts based upon a likelihood ratio test. Stat Papers 36, 215–236 (1995). https://doi.org/10.1007/BF02926035
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/BF02926035