| metainf.meta | R Documentation |
Performs an influence analysis. Pooled estimates are calculated omitting one study at a time.
## S3 method for class 'meta'
metainf(
x,
pooled,
sortvar,
prediction,
overall = x$overall,
text.pooled,
no = 1,
cid = NULL,
cid.below.null = NULL,
cid.above.null = NULL,
small.values = "desirable",
...
)
metainf(x, ...)
## Default S3 method:
metainf(x, ...)
x |
An object of class |
pooled |
A character string indicating whether a common effect
or random effects model is used for pooling. Either missing (see
Details), |
sortvar |
An optional vector used to sort the individual
studies (must be of same length as |
prediction |
A logical indicating whether to report prediction intervals. |
overall |
A logical indicating whether overall results should be reported. |
text.pooled |
A character string used in printouts and forest plots to label the pooled effect estimate. |
no |
A numeric specifying which meta-analysis results to consider. |
cid |
A numeric value or vector specifying clinically important
differences (CID) / decision thresholds used to calculate expected
proportions of clinically important benefit or harm
(see |
cid.below.null |
A single numeric defining the decision threshold below
the null effect to distinguish clinically important from not important
effects (see |
cid.above.null |
A single numeric defining the decision threshold above
the null effect to distinguish clinically important from not important
effects (see |
small.values |
A character string specifying whether small
treatment effects indicate a beneficial ( |
... |
Additional arguments (ignored). |
Performs a influence analysis; pooled estimates are calculated
omitting one study at a time. Studies are sorted according to
sortvar.
If argument pooled is missing, a random effects model is assumed
(pooled = "random") whenever prediction intervals are requested,
irrespective of x$random. If argument prediction is missing,
the setting x$prediction is used; an explicit value of
prediction overrides this setting.
If prediction intervals are not requested, a random effects model is
assumed when x$common = FALSE and x$random = TRUE.
Otherwise, a common effect model is assumed. An explicit choice of
pooled takes precedence over these defaults.
An object of class "metainf" with dedicated print and forest
functions.
The following list elements provide results from meta-analyses, each
excluding one study at a time (see meta-object for more
information on these list elements):
| studlab, TE, seTE, df.random, lower, upper, statistic, pval, |
| lower.predict, upper.predict, df.predict, w (sum of weights), |
| tau2, se.tau2, lower.tau2, upper.tau2, tau, lower.tau, upper.tau, |
| I2, lower.I2, upper.I2, Rb, n.harmonic.mean, t.harmonic.mean, |
| k, k.study, k.all, k.TE, k.MH. |
The following list elements contain results of the original meta-analysis:
| TE.pooled, seTE.pooled, df.random.pooled, |
| lower.pooled, upper.pooled, statistic.pooled, pval.pooled, |
| lower.predict.pooled, upper.predict.pooled, |
| df.predict.pooled, w.pooled, |
| tau2.pooled, se.tau2.pooled, lower.tau2.pooled, upper.tau2.pooled, |
| tau.pooled, lower.tau.pooled, upper.tau.pooled, |
| I2.pooled, lower.I2.pooled, upper.I2.pooled, Rb.pooled, |
| n.harmonic.mean.pooled, t.harmonic.mean.pooled, |
| k.pooled, k.study.pooled, k.all.pooled, k.TE.pooled, k.MH.pooled. |
Guido Schwarzer guido.schwarzer@uniklinik-freiburg.de
Cooper H & Hedges LV (1994): The Handbook of Research Synthesis. Newbury Park, CA: Russell Sage Foundation
forest.metainf, print.metainf,
cidprop
data(Fleiss1993bin)
ma1 <- metabin(d.asp, n.asp, d.plac, n.plac,
data = Fleiss1993bin, studlab = study, sm = "RR", method = "I")
ma1
metainf(ma1)
metainf(ma1, pooled = "random")
forest(metainf(ma1))
## Not run:
forest(metainf(ma1, pooled = "random"))
# Create a PDF file with a forest plot containing UTF-8 characters
# for tau^2, tau, and I^2
forest(metainf(ma1, pooled = "random"),
label.tau2 = "\u03c4\u00b2", label.tau = "\u03c4", label.I2 = "I\u00b2",
file = "forest_with_unicode.pdf", device = "cairo_pdf")
metainf(ma1, sortvar = study)
metainf(ma1, sortvar = 7:1)
ma2 <- update(ma1, title = "Fleiss1993bin meta-analysis", backtransf = FALSE)
metainf(ma2)
data(Fleiss1993cont)
ma3 <- metacont(n.psyc, mean.psyc, sd.psyc, n.cont, mean.cont, sd.cont,
data = Fleiss1993cont, sm = "SMD")
metainf(ma3)
## End(Not run)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.