Erratum

The Editorial by Curtis et al. (2015) contained an error in the following paragraph from page 3468:

‘Thus there is clearly a risk of type 2 error (failure to detect a real effect) if one attempts to correct for false positives by increasing the height of the “P bar”, and this will be without obtaining any extra benefit of avoiding false negatives (type 1 error).’

‘Type 1 error’ has been put in the wrong part of the sentence. It should read:

‘Thus there is clearly a risk of type 2 error (failure to detect a real effect) if one attempts to correct for false positives (type 1 error) by increasing the height of the “P bar”, and this will be without obtaining any extra benefit of avoiding false negatives.’

We thank Catherine Gliddon and John C. Ashton, Department of Pharmacology & Toxicology, School of Medical Sciences, University of Otago, New Zealand for identifying this error.