r2glmm: Computes R Squared for Mixed (Multilevel) Models
The model R squared and semi-partial R squared for the linear and
generalized linear mixed model (LMM and GLMM) are computed with confidence
limits. The R squared measure from Edwards et.al (2008) <doi:10.1002/sim.3429>
is extended to the GLMM using penalized quasi-likelihood (PQL) estimation
(see Jaeger et al. 2016 <doi:10.1080/02664763.2016.1193725>). Three methods
of computation are provided and described as follows. First, The
Kenward-Roger approach. Due to some inconsistency between the 'pbkrtest'
package and the 'glmmPQL' function, the Kenward-Roger approach in the
'r2glmm' package is limited to the LMM. Second, The method introduced
by Nakagawa and Schielzeth (2013) <doi:10.1111/j.2041-210x.2012.00261.x>
and later extended by Johnson (2014) <doi:10.1111/2041-210X.12225>.
The 'r2glmm' package only computes marginal R squared for the LMM and does
not generalize the statistic to the GLMM; however, confidence limits and
semi-partial R squared for fixed effects are useful additions. Lastly, an
approach using standardized generalized variance (SGV) can be used for
covariance model selection. Package installation instructions can be found
in the readme file.
||mgcv, lmerTest, Matrix, pbkrtest, ggplot2, afex, stats, MASS, gridExtra, grid, data.table, dplyr
||lme4, nlme, testthat
||Byron Jaeger [aut, cre]
||Byron Jaeger <byron.jaeger at gmail.com>
Please use the canonical form
to link to this page.