By default (interval = "none") returns the same numeric vector
as the lme4-style point prediction (preserved byte-for-byte from the
pre-existing predict method). With interval = "confidence" or
"prediction", returns a data frame with columns fit,
lwr, upr, se.
# S3 method for rlmerMod
predict(
object,
newdata = NULL,
re.form = NULL,
ReForm,
REForm,
REform,
terms = NULL,
type = c("link", "response"),
allow.new.levels = FALSE,
na.action = na.pass,
interval = c("none", "confidence", "prediction"),
level = 0.95,
...
)A numeric vector (when interval = "none") or a data
frame with columns fit, lwr, upr, se.
An rlmerMod object.
See the lme4 predict.merMod documentation; these
arguments are forwarded unchanged to the point-prediction code path.
One of "none" (default, returns a numeric
vector for backwards compatibility), "confidence"
(fit-uncertainty interval for \(X\hat{\beta} + Z\hat{b}\)), or
"prediction" (adds residual variance \(\hat{\sigma}^2\)).
Coverage level for the interval; default 0.95.
The fixed-effect contribution to the SE uses the robust cluster
sandwich vcov(object, type = "sandwich"); the random-effects
contribution (when REs are part of the prediction) is computed from
the partial influence function of \(\hat{u}\) (the local-shift
sensitivity). For interval = "prediction" the additional
residual variance \(\hat{\sigma}^2\) is added. Intervals are
fit +/- z * se with \(z = \Phi^{-1}((1 + level)/2)\); for
small numbers of clusters the normal approximation may under-cover,
and a bootstrap CI (via confint(..., method = "boot")) is
preferable.
vcov