cint.rma.uni: Confidence Intervals for rma.uni Objects

Description

The function calculates a confidence interval for the amount of (residual) heterogeneity in random- and mixed-effects models that were fitted with the rma.uni function. Confidence intervals for the model coefficients can also be obtained.

Usage

## S3 method for class 'rma.uni':
cint(object, fixed=FALSE, random=TRUE, level=object$level, 
     digits=object$digits, control=list(), ...)

Arguments

object

an object of class "rma.uni".

fixed

logical indicating whether confidence intervals for the model coefficients should be returned (default is FALSE).

random

logical indicating whether a confidence interval for the amount of (residual) heterogeneity should be returned (default is TRUE).

level

a numerical value between 0 and 100 specifying the confidence interval level (the default is to take the value from the object).

digits

an integer specifying the number of decimal places to which the results should be rounded (the default is to take the value from the object).

control

a list of control values for the iterative algorithm. Defaults to an empty list, which means that default values are defined inside the function. See Note.

...

other arguments.

Value

Either a single data frame or a list with two data frames (named fixed and random) with the following elements:
estimateestimate of the model coefficient or variance component.
ci.lblower bound of the confidence interval.
ci.ubupper bound of the confidence interval.

Details

The confidence intervals for the model coefficients are simply the usual Wald-type intervals which are also shown when printing the fitted object or when calling coef.rma. The confidence interval for the amount of (residual) heterogeneity is obtained via the Q-profile method as described by Hartung and Knapp (2005) and Viechtbauer (2007). The method provides an exact confidence interval for $\tau^2$ in random- and mixed-effects models. The square root of the interval bounds is also returned for easier interpretation. For random-effects models, a confidence interval for $I^2$ is also provided. Since $I^2$ is just a monotonic transformation of $\tau^2$, the confidence interval for $I^2$ is also exact.

References

Hartung, J.,& Knapp, G. (2005) On confidence intervals for the among-group variance in the one-way random effects model with unequal error variances. Journal of Statistical Planning and Inference, 127, 157--177. Viechtbauer, W. (2007) Confidence intervals for the amount of heterogeneity in meta-analysis. Statistics in Medicine, 26, 37--52.

Examples

Run this code

### load BCG vaccine data
data(dat.bcg)

### meta-analysis of the log risk rates using a random-effects model
res <- rma(ai=tpos, bi=tneg, ci=cpos, di=cneg, 
           data=dat.bcg, measure="RR", method="REML")

### confidence interval for the total amount of heterogeneity
cint(res)

### mixed-effects model with absolute latitude in the model
res <- rma(ai=tpos, bi=tneg, ci=cpos, di=cneg, mods=ablat,
           data=dat.bcg, measure="RR", method="REML")

### confidence interval for the residual amount of heterogeneity
cint(res)

Run the code above in your browser using DataLab

Last chance! 50% off unlimited learning