Figure 1. Proportion of reviewers identifying each error for those who did and did not recommend rejection of each paper

A: Reviewers rejecting Paper 1 (n=335); B: Reviewers not rejecting Paper 1 (n=156); C: Reviewers rejecting Paper 2 (n=346); D: Reviewers not rejecting Paper 2 (n=71); E: Reviewers rejecting Paper 3 (n=325); F: Reviewers not rejecting Paper 3 (n=74)

Errors: 1, Biased randomization procedure; 2, Inconsistent denominator; 3, Unjustified conclusions; 4, Poor response rate; 5, Poor justification; 6, Discrepancy between abstract & results; 7, ITT would be appropriate; 8, No sample size calculation; 9, Unknown reliability & validity; 10, No explanation of drop outs; 11, No ethics approval; 12, Hawthorne effect; 13, Word reversal; 14, Inconsistency between text & tables