[R-meta] Use of Hedges 1992 publication bias correction method
Dear Chisio There are a number of alternative methods mentioned in the CRAN Task View on MetaAnalysis https://cran.r-project.org/view=MetaAnalysis. Look for the section headed "Unobserved studies". Michael (maintainer of the task view)
On 27/09/2018 05:38, Chishio Furukawa wrote:
Dear all, I am writing to ask if anyone knows a statistical package that provides Hedges (1992) method to correct publication bias. It assumes that publication probability is a step function of p-value, and provides estimates using the maximum likelihood with assumptions that the underlying distributions are normal. It is commonly used in economics, but I could not find packages that automate this approach. Please let me know if anyone knows about this! Best regards, Chishio