Problem am wanting to do a leave1out sensitivity analysis of an rma.mv (three-level) meta-analysis, where "1" is a cluster / sample (rather than a single effect size).
I understand that some functions, such as leave1out, do not work with rma.mv model objects (e.g., Metafor package: bias and sensitivity diagnostics).
Attempted solutions
I do not know how to write a "for-loop" (which Wolfgang suggested in post above), so cannot try that.
I have searched the web, but found no guidance on 'for-loops', nor how to do sensitivity analyses for samples / clusters in a three-level meta analysis.
