In a scientific utopia, communication is open, data are accurate and results are reproducible.That’s the kind of environment The Center for Open Science in Charlottesville is trying to provide for its users. WMRA’s Kara Lofton has the story.
I met Jeffrey Spies at the Center for Open Science’s large, open-floor office late on a Saturday afternoon. The door to the multi-suite office building was locked, but lights were on at COS and a few young developers huddled in the back around a 3-D printer.
The atmosphere was what one would expect for a well-funded tech startup, not a non-profit. But the Center for Open Science is both.
JEFFREY SPIES: Our main mission is to increase transparency, reproducibility and integrity in the sciences. The main goal is to make science more efficient and we believe we can do that via openness.
They achieve this goal by providing free and open services for scientists and researchers through statistical counseling and what they call the Open Science Framework.
SPIES: This is a content and collaboration platform for scientists to use, just to make themselves more efficient. It’s all open source. You can use it just to work with your fellow collaborators.
Erica Baranski is a PhD student in psychology at the University of California at Riverside. Last summer she was a developer intern at the Center for Open Science. She is now using the Open Science Framework for her own research in social and personality psychology.
ERICA BARANSKI: We use the Open Science Framework to facilitate our lab meetings primarily. We also use it to keep track of different versions of data sets and analysis. Before the OSF we would print out all of our analysis and present them to the lab and sort of stash them in a binder that started out really organized but then sort of lost its way quickly. So obviously we lost track of the sorts of things that we were analyzing, the data sets we were using weren’t consistent across different analysis so it was just very, very unorganized.
Another function of the Open Source Framework is the innovative idea of sharing data and methodologies with other scientists in a way that would allow them to use, collaborate or consult on a given research project.
One example is the methods a scientist used to achieve particular results. For example, did the researcher perform an experiment, or a content analysis, or a survey? How can someone later replicate the original study? Spies called methods, materials and original hypotheses “building tools.”
SPIES: This is where innovation occurs. This is where we find out what actually happens in the scientific process. When that’s open we can then examine how things actually happened. It’s very difficult to just take a paper and try to reproduce that. You need a lot of information to do so. It’s difficult then to build on those ideas if you don’t have the full story.
Baranski said her research team frames their research in a very open way and that the data collaborators collect for them is available to view and comment on. For her, this is all part of the integrity of true science.
BARANSKI: Transparency in science has sort of I think in the past been lacking and we are seeing the consequences of that a little bit. So there have been questionable research practices is sort of scratching the surface. There are actually people that are faking their data, which has massive negative consequences for science, but also their collaborators. So transparency in science is a means to limit those incidences, but it’s also in the spirit of science, science is for everyone.
Spies said that in order for science to truly be for everybody, services like those the Center for Open Science offers have to be free. COS has achieved this vision through over 14 million dollars in private funding that they generated in the last two years. The money has come entirely from a variety of generous grants and donations. This allows them to pay competitive tech salaries to their employees without needing to lock in users to their services. It’s like a Google docs or Dropbox that is catered specifically to the needs of researchers.
SPIES: When you monetize the product there has to be a bias somewhere, there has to be something you are selling and so some groups will do that by making it very hard to leave the site by creating lock in. But scientists should not be locked in to the tools they use. If there’s a better tool I should be able to get the data I have on that site and export it to the newer, better tool.