The open-access scientific journal PLOS Biology has launched a section dedicated to research on research. The new section will analyze how research is designed, executed, communicated and evaluated. With this initiative, PLOS aims to regain public confidence in the scientific method after too many cases of irreproducible research.
According to several meta-research articles, US$200 billion are spent every year on academic works that can’t be reproduced by other researchers. An estimated 85% of worldwide investment in biomedical research is wasted. As a consequence, tax-payers willingness to fund research could decline, and scientists might follow wrong leads, basing their research in inaccurate or false studies. The current high number of retractions makes things even worse. The purpose of the data-driven meta-research section is to improve research standards and avoid the threat of bad practices to the progress of science.
Poor reproducibility and lack of transparency
The section’s two first articles have analyzed reporting in biological research papers between 2000 and 2014, and have found poor reproducibility and transparency. The study directed by J. Ioannidis confirms that most papers didn’t share data or protocols, presented derivative work as original, didn’t inform about funding and didn’t acknowledge conflicts of interest. In the other paper, C. Holman and colleagues used computer modelling to analyze the effects of removing outliers in animal preclinical research, and concluded that the selection bias increases false positives.
The new Meta-Research Section in PLOS Biology underlines the importance that scientific journals always had in preserving good research practices, and represents a call to all journals to regain that role.
Source: PLOS Biology