SCIENCE (real science, not the summaries in popular books and the media) is needlessly closed to the outside world. Worse, it is closed within itself, with every lab its own silo, and little sharing of data or materials.
Top researchers work frantically behind the scenes, but only occasionally do they share their results, by describing selected experiments in publications. This way of working may have been sensible in the paper past, but as the digital age advances it should not be the way we work now.
Recently one of us (MT) directed a fully open project in which all data and ideas were shared online. People completely unknown to the existing team came to the project and helped.
Some were employed by corporate laboratories but nevertheless provided significant inputs free of charge. This should not be surprising, as it is in our nature to help others solve problems. It is also in our nature to compete, and the open nature of the project yielded competition that helped achieve a common end. On this playing field, the greatest expertise could win out.
In an ongoing project involving open source drug discovery, a scientific dispute was settled by a discussion between an undergraduate and a professor in public view - the very best kind of educational practice.
Elsewhere in the world, a Dutch schoolteacher discovered a mysterious astronomical object by participating in the open GalaxyZoo project and was included as an author on the resulting publication.
More recently a high school student was able to devise an algorithm for detecting pancreatic cancer by reading free online scientific journals.
Our desire to see science thrive outside its walled garden and in full view of the public motivated us to write an open letter to the new CEO of the Australian Research Council (ARC).
That letter is included in full below. If readers agree with the letter, we invite them to add their signature to it by clicking on the link at the bottom of the letter. If a more cooperative, open research culture was actively encouraged by funding agencies we believe science would be more efficient and powerful.
The way the ARC evaluates scientists inadvertently fuels the secretive competition we scientists can get caught up in. We focus on competing for publications in prestigious journals. But the ARC can change its policies to lead us towards openness.
The ARC could make data sharing a condition of the grants it gives us.
An open letter to Professor Aidan Byrne, CEO, the Australian Research Council
Dear Professor Byrne,
We know you have floated plans to broaden access to scientific publications supported by the ARC. We enthusiastically applaud such moves, as we believe that citizens and businesses should be able to access the science that their tax dollars pay for. We also believe the ARC should go beyond opening access to journal articles and take steps to increase the sharing of scientific data.
As science is currently practised, many research publications are more like press releases or executive summaries than detailed records of what was done and found. They are descriptions of the primary outcome the authors wish to reveal, often with insufficient supporting data for readers to validate the conclusions themselves.
As a result, science is facing a reproducibility crisis. Recent efforts to replicate published findings have yielded a low rate of success (Begley & Ellis, 2012; Prinz, F., Schlange, T. & Asadullah, K., 2011). More details of what was actually done and found when reports are published are needed to remedy this. We think these details should usually include the raw data.
Not only would such open science facilitate replication, but it would also result in further discoveries. For example, many more publications have arisen from people re-analysing Hubble Space Telescope data than from the scientists who originally acquired the data.
The cost of conducting the Human Genome Project is acknowledged to be dwarfed by its economic benefit to scientific productivity because the data are publicly available. For a physically more isolated country like Australia to participate in cutting-edge international science, the free sharing of research data is particularly important.
In the future, data sharing will result in discoveries beyond the capability of any one lab, as data can be mined by computer algorithms to seek trends and patterns that reveal hidden truths. In a recent report entitled Science as an Open Enterprise, the Royal Society wrote that: "Governments should recognise the potential of open data … to enhance the excellence of the science base. They should develop policies for opening up scientific data that complement policies for open government data, and support development of the software tools and skilled personnel that are vital to the success of both."
It may be only through open science, with massively collaborative efforts, that urgent problems of the world can be solved. Open science can reduce wasteful duplication of efforts (because it is clear what everyone is doing) while maintaining the highest quality work (because everything is continually peer-reviewed in a competitive environment).
We'd like to close with an invitation. We are two members of a group organising a three-day conference in Auckland in early 2013 examining how open research fits the Australian/New Zealand context. We cordially invite you to attend this meeting, and would be delighted if you were able to deliver a keynote outlining your vision for the future practice of science.
Alex O. Holcombe
These opinions and those of the other signatories do not necessarily reflect the positions of their institutions.
Academics at ANU, Harvard, Yale, and other universities have joined in signing our letter, together with people from outside the university sector.