find_second_split function

A function to find the optimal combination of first two splits, and the corresponding Q-between given the first split

find_second_split(xk, first.splits, y, vi, minbucket, minsplit)

Arguments

• xk: moderator vector
• first.splits: possible first splits
• y: effect size
• vi: sampling variance
• minbucket: the minimal number of studies in child nodes
• minsplit: the minimal number of studies in parent node

Returns

a list including all possible combinations of the triplet