This article was originally solicited by and published in Nature.
Almost a decade ago, at Arnold Ventures — a US $2-billion philanthropic organization in Houston, Texas — we realized that using evidence to direct our giving required having more confidence in the evidence itself. As vice-president of research, I found myself deep in efforts to improve science, dispersing more than $60 million in grants to make sure researchers can build on others’ results. I was part of the discussions that led to widely adopted guidelines promoting transparency and openness, the clinical-trial repository Vivli and the launch of the Center for Open Science, a non-profit organization in Charlottesville, Virginia.
I’ve seen much positive change since then. But sometimes I worry that we might end up with the worst of all worlds: the pretence of reproducibility without the reality.