[R-sig-ME] how to know if random factors are significant?
Fabian Scheipl
Fabian.Scheipl at stat.uni-muenchen.de
Wed Apr 2 09:32:01 CEST 2008
>To obtain a
>p-value, you need to compare with some distribution and a chi-square
>with one df is the default output. Often however a mixture of 0 and 1
>df's are more appropriate, hence a more correct p-value is half the
>one, the software reports.
For linear mixed models with uncorrelated random effects, our package
RLRsim offers a rapid algorithm to determine the exact finite sample
distribution of the restricted likelihood ratio for testing whether
the variance of a random effect is zero.
Using the ChiSquare(1) or a 50:50 mixture of ChiSquare(1) and 0 will
almost always lead to very conservative tests. For a detailed
comparison of various approaches to test for zero variance in linear
mixed models have a look at:
F.Scheipl, S.Greven, H.Küchenhoff (2008): Size and power of tests for
a zero random effect variance or polynomial regression in additive and
linear mixed models.
Computational Statistics & Data Analysis, 52(7):3283-3299
(http://dx.doi.org/10.1016/j.csda.2007.10.022).
More information about the R-sig-mixed-models
mailing list