Science is based on experimental testing and data. Yet ironically the publishing of science was until recently based on faith not study. The BMJ Group has a vision of becoming the world's first evidence based publisher. We have been publishing around evidence based medicine for some years and hope that the same thinking can transform publishing.

We carry out research into all aspects of the BMJ Group's operation with the objective of developing, promoting, and incorporating evidence-based publishing into our own practices. Our research projects usually originate from within the BMJ Group, but our work has a wider focus: nearly all of the issues we research have relevance for journal editors, authors, peer reviewers and publishers working across biomedical science. We also conduct collaborative research projects with researchers in international institutions. As a publisher of numerous journals we have a rich dataset of submitted articles and privileged contact with authors and reviewers worldwide.

Here you will find useful information on research undertaken in the BMJ Group into all aspects of publishing science including peer review, conflicts of interests, editorial decision making, open access publishing, authorship/contributorship, presentation of research, research ethics, publication bias, and education. Projects undertaken include randomised controlled trials of open peer review and the effectiveness of peer review training, author surveys on the use of statistical expertise in medical research, preferences for presentational style of scientific research, perceptions of electronic publishing, attitudes towards open access publishing, and reader surveys to understand the effect of competing interest statements on research credibility. If you have any ideas for collaborative research in this area, please contact us.

Sara Schroter, senior researcher, BMJ