Dear Statalist,
I am using the meta forestplot command. Depending on how many subgroup variables one defines the output either gets presented with single study estimates:
Array

or with summary estimates per group:
Array
I would like to control this myself. E.g. when presenting with a single subgroup variable with plenty of categories, I might prefer the second style. How do I do this?

The code I used for the sample graphs in Stata 16.1:
Code:
clear all
set more off
webuse pupiliq
meta set stdmdiff se, random studylabel(studylbl)
meta forest, subgroup(tester)
meta forest, subgroup(tester setting)