THE FREE SHARING OF IDEAS
THE FREE SHARING OF IDEAS: REPOSITORIES OF STERILIZED DATA OR THE EVOLUTION OF THE ORAL HISTORY OF A DISCIPLINE
By Marianne Evola
During presentations on responsible research, I frequently discuss how scientific progress critically relies on the free sharing of ideas. Generally, I raise the topic when speaking on how plagiarism damages trust within the scientific community and inhibits our willingness to freely share our ideas. However, recently during a presentation to Chemical Engineering graduate students, a student in the audience raised his hand to challenge me on how "free" it actually is to share our ideas when research publications charge individuals and institutions exorbitant rates for access to published works. Of course, I explained that as authors, we are not paid for the work that we submit for publication, thus as individuals, we are freely sharing our ideas. As for the student's critique of our academic world, his criticism of the expense associated with journal subscriptions for individuals and libraries was not the first that I had heard. Rather, the controversy has been addressed for quite a few years and I have read heated discussions online regarding the issue. In fact, it is not uncommon to read about academics boycotting various publishers due to excessive rates.
As often happens when a good question or critique captures my attention, shortly after the student challenged me, I began to attend to alternative options for sharing ideas that were sent to me by colleagues. As such, I thought that I would put them together because the future of the "free sharing of ideas" may be a bit different than how it was constructed in the past. And hopefully, it may be a bit richer than I'd previously imagined.
Shortly after the student challenged me, I received an email regarding federal regulations on making published literature freely accessible to the public. Obviously, the student's concerns regarding the high cost of free ideas are shared by our federal research sponsors. Basically, the federal government has been working on legislation to make publically funded research more accessible to the public. The public access policy states that authors of federally funded research are now required to make peer-reviewed publications in scholarly journals freely available to the public within 12-months of initial publication (NSF 15-52). The email described the NSF Public Access Repository (NSF-PAR) and mechanisms to enable investigators to meet the requirements. Although the email was interesting and timely with regard to the student's critique of our publishing system, full details would be rather dry reading so I encourage interested parties to follow the link provided above for further information.
However, I encountered a more intriguing presentation about a week or so after I received the email on the NSF requirements. I attended a library presentation by Brian Quinn on Data Repositories. The library has been taking a lead on assisting the TTU community prepare their data for Archiving and thus data management and repositories are a frequently addressed topic in responsible research. However, Brian's presentation got me thinking about data archives in a more insightful and useful way. During the presentation, Brian presented the option of archiving negative data or unpublished manuscripts. His suggestion made me realize that as the networks of shared data grow and are more widely utilized, data archiving will provide a means for us to also share negative or inconclusive results with our colleagues. In fact, arguably repositories of negative data and unpublished manuscripts may someday be a widely utilized measure of productivity while also providing our colleagues with a more comprehensive collection of research results and interpretations.
Historically, and probably currently, a research career is measured on the publication of peer reviewed original research. We submit our work to a journal for publication, it is sent out for the critical review by our peers, and then the critique is addressed and ultimately, the best polished work is published and shared with our colleagues. Our colleagues build on our published work, and with their citation, our careers grow. The peer review system, at it's best, polices the quality of the published work and serves the research world by minimizing poor research from being published.
However, our system has some well recognized flaws. First, there is publication bias for positive results rather than negative or inconclusive results. Even though the scientific community widely accepts that all data is valuable, it is considerably less likely that a study presenting negative research results will be published. Thus most laboratories, have file cabinets and computers loaded with unpublished, thus unshared, negative results. The inability to publish or share negative results wastes resources because colleagues inadvertently repeat work was never published due to the negative results and the publishing bias. This phenomenon is largely recognized as a flaw in our peer review system. However, during Brian's presentation, I realized that unpublished results and even manuscripts could be readily archived and shared with our colleagues. And when you imagine how the network of archived work could ultimately grow, we will not only be able to access the polished and successful work of our colleagues, we will also be able to access their failures. Access to this negative data will save us time and resources that would have been wasted on repeating failed work. However, access to the negative data and unpublished work of our colleagues will also enrich our understanding of our respective disciplines.
The more I thought about this network of unpublished data and manuscripts, the more I realized how rich and valuable it could be. Just before I made the transition from research into administration, I was finally getting comfortable talking to senior scientists in my field about research. By talking, I don't mean presenting. Presenting is like publishing, it is the polished story, the story that we wish was true. Rather, I had developed the comfort of talking with senior colleagues regarding all of the humorous slips and trips that preceded the polished research story. I loved their stories about all of the flaws and unrealized research questions that they encountered and the many research paths that had not been followed. But also, the oral history they share often contains stories of simple curiosity rather than a clear research agenda.
It was during those informal discussions that I fully realized that a great deal more research was conducted as compared to the amount of research that was published in the literature. At that time, our lab was conducting a series of experiments to assess learning and cognition in mice and we were replicating the work of a senior colleague. We invited him to give a seminar for our department and during his visit, I had a great deal of time to talk to him about all of my ideas for various experimental manipulations of his research paradigm. The interesting thing was that during this conversation, as I was proposing all of my "innovative" ideas. He continually responded with: "Yeah, we tried that", then he would describe the problems or insights that the experiment had produced. None of it was published, because they were unfinished concepts or failed projects, but it was all in his research memory. And I realized during these conversations that there is so much more research knowledge in our senior colleagues than could ever be shared in the published literature.
My senior colleague had conducted a lot of experiments because the research was fun and the students were curious. There was so much more information in those casual research conversations than there could ever be in a polished peer-reviewed publication because the publication is missing most of the story. The publication is missing all of the false starts, bad ideas, research errors and of course, all of the data that those slips and trips provided to that research team. There is a large repository of valuable research knowledge that cannot ever be accessed via the published literature but will be lost to those that do not learn to access it via the oral history that only research conversations provide. That information was never sufficiently polished for publication.
When I realized that much of that work could be shared in data archives, I realized how incredibly valuable a growing network of archives could be. Most of the presentations on data repositories and archives have addressed a need for formatting data so that it can be utilized for data mining in future research projects. Definitely, there is a need to be aware of this progression because it is happening rather quickly as I have previously discussed. However, when I thought about the potential for sharing unpublished work, I realized that repositories of career data and/or lab notebooks from a research career could be rich and interesting collections of the true, unpolished history of our respective disciplines. The unpolished story or an unpublished manuscript could contain a great deal of information on the slips and trips of a real research project rather than merely the polished story after all of the experimental details had been mastered for a final publishable project. I don't know how data repositories will evolve, because the system is in it's infancy. Perhaps it will only be large and varied collection of sterile data points that can be numerically accessed for new questions. But more interestingly, perhaps it could also be the evolutionary equivalent of the oral history of a discipline, a rich history of the messiness of science and/or the memoirs of our many mistakes along with the abandoned or unrealized ideas of our academic idols. I guess it all depends on what our colleagues choose to share.
Marianne Evola is senior administrator in the Responsible Research area of the Office of Research & Innovation. She is a monthly contributor to Scholarly Messenger.