[R-meta] HKSJ correction
Viechtbauer, Wolfgang (SP)
wo||g@ng@v|echtb@uer @end|ng |rom m@@@tr|chtun|ver@|ty@n|
Fri May 29 17:05:11 CEST 2020
Use it. Always. Essentially every simulation study that has examined this method has shown that it performs better than the alternatives (except maybe for permutation tests, but those methods essentially perform equally well and the former is much quicker). See, for example:
Langan, D., Higgins, J. P. T., Jackson, D., Bowden, J., Veroniki, A. A., Kontopantelis, E., Viechtbauer, W., & Simmonds, M. (2019). A comparison of heterogeneity variance estimators in simulated random-effects meta-analyses. Research Synthesis Methods, 10(1), 83-98.
One exception: When meta-analyzing dichotomous outcomes with rare events. But NOT using the method isn't the solution. One should switch to a different modeling approach then (e.g., logistic mixed-effects models).
>From: R-sig-meta-analysis [mailto:r-sig-meta-analysis-bounces using r-project.org]
>On Behalf Of Ioana Cristea
>Sent: Friday, 29 May, 2020 15:20
>To: r-sig-meta-analysis using r-project.org
>Subject: [R-meta] HKSJ correction
>I am doing a meta-analysis of continuous data (SMD), number of studies (k)
>around 75, heterogeneity high (Tau-squared around 0.15). Our interest is in
>fact in a key subgroup analysis (k 50 and 14 for each subgroup)
>I am using the REML meta-analysis model, is there any scope for also using
>Hartung-Knapp-Sidik-Jonkman variance correction? Could you point me to some
>references that discuss this?
>Ioana-Alina Cristea, Ph.D.
>Department of Brain and Behavioral Sciences
>University of Pavia, Piazza Botta 11, 27100 Pavia, Italy
>METRICS (Meta-research Innovation Center at Stanford)
>Stanford University, California, USA
>Associate editor Systematic Reviews
More information about the R-sig-meta-analysis