Meta-analysis challenges a textbook example of status signalling and demonstrates publication bias