Abstract Meta‐analyses are subject to bias for many of reasons, including publication bias. Asymmetry in a funnel plot of study size against treatment effect is often used to identify such bias. We compare the performance of three simple methods of testing for bias: the rank correlation method; a si...
No comments yet
Be the first to share your thoughts!