Fig. 4From: On the Q statistic with constant weights in meta-analysis of binary outcomesRelative error in the level of the test for heterogeneity of log-relative-risk, vs upper tail area, for four approximations to the null distribution of \(Q_F\) and the chi-square approximation to the null distribution of \(Q_{IV}\), when \(p_{iC} = .1\). The rows correspond to the combinations of \(\rho = 0\) and \(\rho = 1.5\) with \(n = 40\) and \(n = 100\)Back to article page