de_mean() computes group- and de-meaned versions of a
variable that can be used in regression analysis to model the between-
and within-subject effect.
de_mean(x, ..., grp, append = TRUE, suffix.dm = "_dm",
suffix.gm = "_gm")A data frame.
Names of variables that should be group- and de-meaned.
Quoted or unquoted name of the variable that indicates the group- or cluster-ID.
Logical, if TRUE (the default) and x is a data frame,
x including the new variables as additional columns is returned;
if FALSE, only the new variables are returned.
String value, will be appended to the names of the
group-meaned and de-meaned variables of x. By default, de-meaned
variables will be suffixed with "_dm" and grouped-meaned variables
with "_gm".
For append = TRUE, x including the group-/de-meaned
variables as new columns is returned; if append = FALSE, only the
group-/de-meaned variables will be returned.
de_mean() is intended to create group- and de-meaned variables
for complex random-effect-within-between models (see Bell et al. 2018),
where group-effects (random effects) and fixed effects correlate (see
Bafumi and Gelman 2006)). This violation of the so called
Gauss-Markov-assumption can happen, for instance, when analysing panel
data. To control for correlating predictors and group effects, it is
recommended to include the group-meaned and de-meaned version of
time-varying covariates in the model. By this, one can fit
complex multilevel models for panel data, including time-varying,
time-invariant predictors and random effects. This approach is superior to
simple fixed-effects models, which lack information of variation in the
group-effects or between-subject effects.
A description of how to translate the
formulas described in Bell et al. 2018 into R using lmer()
from lme4 or glmmTMB() from glmmTMB can be found here:
for lmer()
and for glmmTMB().
Bafumi J, Gelman A. 2006. Fitting Multilevel Models When Predictors and Group Effects Correlate. In. Philadelphia, PA: Annual meeting of the American Political Science Association.
Bell A, Fairbrother M, Jones K. 2018. Fixed and Random Effects Models: Making an Informed Choice. Quality & Quantity. 10.1007/s11135-018-0802-x
# NOT RUN {
data(efc)
efc$ID <- sample(1:4, nrow(efc), replace = TRUE) # fake-ID
de_mean(efc, c12hour, barthtot, grp = ID, append = FALSE)
# }
Run the code above in your browser using DataLab