PublicationBiasBenchmark (version 0.1.3)

method.EK: Endogenous Kink Method

Description

Implements the endogenous kink (EK) method proposed by Bom and Rachinger for publication bias correction in meta-analysis. This method modifies the PET-PEESE approach by incorporating a non-linear relationship between publication bias and standard errors through a kinked regression specification. The method recognizes that when the true effect is non-zero, there is minimal publication selection when standard errors are very small (since most estimates are significant), but selection increases as standard errors grow. The kink point is endogenously determined using a two-step procedure based on the confidence interval of the initial effect estimate. See bom2019kinked;textualPublicationBiasBenchmark for details.

Usage

# S3 method for EK
method(method_name, data, settings = NULL)

Value

Data frame with EK results

Arguments

method_name

Method name (automatically passed)

data

Data frame with yi (effect sizes) and sei (standard errors)

settings

List of method settings (no settings version are implemented)

References

Examples

# Generate some example data
data <- data.frame(
  yi = c(0.2, 0.3, 0.1, 0.4, 0.25),
  sei = c(0.1, 0.15, 0.08, 0.12, 0.09)
)

# Apply EK method
result <- run_method("EK", data)
print(result)