A new paper sheds light on how current publication practices may distort science. The article, by researchers in the US, uncovers some of the flaws in the current system and wonders whether the digital age will cause the system of publication to evolve yet further and exorcise 'the winner's curse' {1}.

By taking the process of scientific publication back to basics, the authors are able to predict many negative effects from the current model of journal-based scientific publishing. Herding, the economic term used to describe 'follow-the-leader' behaviour, is proposed to narrow and focus the field of research on certain areas that are determined by the papers with the highest profiles. 'The winner's curse', the theory that the winner in an auction will, on average, have paid more that the item is worth, is likened to the proposed exaggeration of results of papers in the most desirable journals. The authors, possibly controversially, suggest that trials with negative results could be published with the same frequency as those with positive results since they should be judged on the quality of the experiment rather than simply positive findings.

The digital age has already extended the reaches of personal networks and allowed personal opinion to be heard across continents. Journal clubs and isolated researchers can highlight the best articles of their reading to others using services like F1000 and, given the apparent limitations of conventional publishing models, surely this can only be a good thing.

Reference: {1} Young et al., PLoS Med 2008, 5:e201 [PMID:18844432]