confint.mer {papeR}

Confidence intervals for mixed models

Description

Compute confidence intervals for mixed models from package lme4 (prior to version 1.0). This function is only needed for backward compatibility.

Usage

## S3 method for class 'mer'
confint(object, parm, level = 0.95,
            simulate = c("ifneeded", TRUE, FALSE),
            B = 1000,...)

Arguments

object

Model of class mer.

parm

Parameters to be included in the confidence interval. See confint.default for details.

level

the confidence level.

simulate

If “"ifneeded"” is specified (default), simulated confidence intervals are returned if (and only if) no z-value exists in the corresponding summary and asymptotic confidence intervals will be returned otherwise. If TRUE (or “"TRUE"”) confidence intervals will be estimated using ci from package gmodels which uses mcmcsamp internally. If FALSE (or “"FALSE"”), asymptotic confidence intervals will be returned and an error is given if not possible.

B

number of samples to take in mcmcsamp. Per default 1000 samples are used.

...

Additional arguments. Currently not used.

Value

Matrix with confidence intervals.

Author(s)

Benjamin Hofner, partially based on code from package stats. See source code for documentation.