A Kenward-Roger approximation and parametric bootstrap methods for tests in linear mixed models: The R Package pbkrtest

By U. Halekoh, Søren Højsgaard. 2014

1,337 citations recorded by the source corpus.

When testing for reduction of the mean value structure in linear mixed models, it is common to use an asymptotic ?2 test. Such tests can, however, be very poor for small and moderate sample sizes. The pbkrtest package implements two alternatives to such approximate ?2 tests: The package implements (1) a Kenward-Roger approximation for performing F tests for reduction of the mean structure and (2) parametric bootstrap methods for achieving the same goal. The implementation is focused on linear mixed models with independent residual errors. In addition to describing the methods and aspects of their implementation, the paper also contains several examples and a comparison of the various methods.

Subfields: Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/587eebe369fbd4dee6e2f1037590ff68c109af40

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Statistics and Probability

Citations in the corpus

Temperature‐ and size‐scaling of phytoplankton population growth rates: Reconciling the Eppley curve and the metabolic theory of ecology (2017, methodology): “Estimates of p-values were obtained using a parametric bootstrapping approach, PBmodcomp, in the pbkrtest package (version 0.4-4; Halekoh and Højsgaard 2014).”
Linear mixed-effects modeling approach to FMRI group analysis (2013, neutral): “However, it is not always realistic to believe that the HDR shape does not vary significantly across subjects, brain regions, tasks/conditions, and trials (Aguirre et al., 1998; Handwerker et al., 2004).”
An Introduction to Mixed Models for Experimental Psychology (2019, positive): “We generally recommend the Kenward-Roger approximation (Halekoh & Højsgaard, 2014; Ken-ward & Roger, 1997), which is based on a modiﬁed F test and also estimates the denominator degrees of freedom, as it is known to provide the best control of Type I errors with the limited sample sizes that are…”
lmerTest Package: Tests in Linear Mixed Effects Models (2017, methodology): “In Halekoh and Højsgaard (2014) an example where LRT leads to misleading results and where Kenward-Roger’s method is accurate is given.”
Consequences of eye color, positioning, and head movement for eye-tracking data quality in infant research (2015, positive): “…session. p -Values for statistical signiﬁcance of each predictor were acquired by comparing the full model with all predictors and random intercepts for operator and participant session, to the model without the predictor of interest using parametric bootstrapping (Halekoh & Højsgaard, 2014).”
Total replacement of fish meal with black soldier fly (Hermetia illucens) larvae meal does not compromise the gut health of Atlantic salmon (Salmo salar) (2020, methodology): “The p value of diet effect was obtained via parametric bootstrap comparisons using the pbkrtest package (Halekoh and Højsgaard, 2014).”
Evaluating significance in linear mixed-effects models in R (2016, methodology): “This function makes use of functions from the pbkrtest package (Halekoh & Højsgaard, 2014) to implement the Kenward-Roger approximation.”
Least-Squares Means: The R Package lsmeans (2016, methodology): “These are obtained from the pbkrtest package (Halekoh and Højsgaard 2014), and they use the Kenward-Rogers method.”
SIMR: an R package for power analysis of generalized linear mixed models by simulation (2016, methodology): “A parametric bootstrap test (e.g. Halekoh & Højsgaard 2014) might be preferred for the ﬁnal analysis.”
Multilevel Modeling and Ordinary Least Squares Regression: How Comparable Are They? (2018, methodology): “We used the pbkrtest (Halekoh & Højsgaard, 2014) and the lmerTest (Kuznetsova, Brockhoff, & Christensen, 2016) packages in R to apply the KR correction.”