metapb: Detecting and Quantifying Publication Bias/Small-Study...

In altmeta: Alternative Meta-Analysis Methods

Detecting and Quantifying Publication Bias/Small-Study Effects

Description

Performs the regression test and calculates skewness for detecting and quantifying publication bias/small-study effects.

Usage

metapb(y, s2, data, model = "RE", n.resam = 1000)

Arguments

y	a numeric vector specifying the observed effect sizes in the collected studies; they are assumed to be normally distributed.
s2	a numeric vector specifying the within-study variances.
data	an optional data frame containing the meta-analysis dataset. If data is specified, the previous arguments, y and s2, should be specified as their corresponding column names in data.
model	a characher string specifying the fixed-effect ("FE") or random-effects ("RE", the default) model. If not specified, this function uses the Q statistic to test for heterogeneity: if the p-value is smaller than 0.05, model is set to "RE"; otherwise, model = "FE".
n.resam	a positive integer specifying the number of resampling iterations.

Details

This function derives the measures of publication bias introduced in Lin and Chu (2018).

Value

This function returns a list containing measures of publication bias, their 95% confidence intervals, and p-values. Specifically, the components include:

n	the number of studies in the meta-analysis.
p.Q	the p-value of the Q-test for heterogeneity.
I2	the I^2 statistic for quantifying heterogeneity.
tau2	the DerSimonian–Laird estimate of the between-study variance.
model	the model setting ("FE" or "RE").
std.dev	the standardized deviates of the studies.
reg.int	the estimate of the regression intercept for quantifying publication bias.
reg.int.ci	the 95% CI of the regression intercept.
reg.int.ci.resam	the 95% CI of the regression intercept based on the resampling method.
reg.pval	the p-value of the regression intercept.
reg.pval	the p-value of the regression intercept based on the resampling method.
skewness	the estimate of the skewness for quantifying publication bias.
skewness.ci	the 95% CI of the skewness.
skewness.ci.resam	the 95% CI of the skewness based on the resampling method.
skewness.pval	the p-value of the skewness.
skewness.pval.resam	the p-value of the skewness based on the resampling method.
combined.pval	the p-value of the combined test that incorporates the regression intercept and the skewness.
combined.pval.resam	the p-value of the combined test that incorporates the regression intercept and the skewness based on the resampling method.

References

Egger M, Davey Smith G, Schneider M, Minder C (1997). "Bias in meta-analysis detected by a simple, graphical test." BMJ, 315(7109), 629–634. <\Sexpr[results=rd]{tools:::Rd_expr_doi("10.1136/bmj.315.7109.629")}>

Lin L, Chu H (2018). "Quantifying publication bias in meta-analysis." Biometrics, 74(3), 785–794. <\Sexpr[results=rd]{tools:::Rd_expr_doi("10.1111/biom.12817")}>

Examples

data("dat.slf")
set.seed(1234)
metapb(y, s2, dat.slf)

data("dat.ha")
set.seed(1234)
metapb(y, s2, dat.ha)

data("dat.lcj")
set.seed(1234)
metapb(y, s2, dat.lcj)

altmeta documentation built on Sept. 11, 2024, 9 p.m.