leave1out.rma.mh: Leave-One-Out Diagnostics for rma.mh and rma.peto Objects

Description

The functions leave1out.rma.mh and leave1out.rma.peto repeatedly fit the specified model, leaving out one observation (i.e., 2x2 table) at a time.

Usage

## S3 method for class 'rma.mh':
leave1out(x, digits=x$digits, transf=FALSE, \dots)
## S3 method for class 'rma.peto':
leave1out(x, digits=x$digits, transf=FALSE, \dots)

Arguments

an object of class "rma.mh" or "rma.peto".

digits

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

transf

logical indicating whether odds ratios or relative risks (and the corresponding confidence interval bounds) should be given in the transformed (meaning: raw) units or in terms of log units (the default).

...

other arguments.

Value

An object of class "list.rma". The object is a list containing the following components:
estimateestimated coefficients of the model.
sestandard errors of the coefficients. NA if transf=TRUE.
zvaltest statistics of the coefficients.
pvalp-values for the test statistics.
ci.lblower bounds of the confidence intervals for the coefficients.
ci.ubupper bounds of the confidence intervals for the coefficients.
Qtest statistics for the tests of heterogeneity.
Qpp-values for the tests of heterogeneity.
The "list.rma" object is formated and printed with print.list.rma.

References

Viechtbauer, W. (2010). Conducting meta-analyses in R with the metafor package. Journal of Statistical Software, 36(3), 1--48. http://www.jstatsoft.org/v36/i03/. Viechtbauer, W. & Cheung, M. W.-L. (in press). Outlier and influence diagnostics for meta-analysis. Research Synthesis Methods.

Examples

Run this code

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

### meta-analysis of the (log) relative risks using the Mantel-Haenszel method
res <- rma.mh(ai=tpos, bi=tneg, ci=cpos, di=cneg, data=dat.bcg, measure="RR")

leave1out(res)
leave1out(res, transf=TRUE)

### meta-analysis of the (log) odds ratios using Peto's method
res <- rma.mh(ai=tpos, bi=tneg, ci=cpos, di=cneg, data=dat.bcg)

leave1out(res)
leave1out(res, transf=TRUE)