In our research integrity and transparency work to date, we have focused on doing audits of the reproducibility and quality of the scientific literature; on producing meta-research to highlight problems and identify best practices; on improving the standards at journals, professional societies, and funders; and on developing the tools to make it easier for researchers to preregister their work and share data.
An exciting new area of change is universities themselves. Universities have tremendous influence over academic behavior and research standards – they train up-and-coming scholars, hire new professors out of PhD programs or post-docs, and promote existing professors to tenure.
To some, universities have seemed difficult to influence directly. For example, hiring and tenure committees conduct their work in a black box, as far as outsiders are concerned, and their practices can differ department by department within the same university.
Even so, universities do not exist in a vacuum. The same scholars who edit or publish in journals, interact with funders, and lead professional societies also work at universities and exert influence there. Thus, as journals, societies, and funders have engaged with the reproducibility movement, we are starting to see early signs that university policies are moving in the right direction as well.
- The University of Oregon Department of Psychology released a job ad last year stating, “Our Department embraces the values of open and reproducible science, and candidates are encouraged to address (in their statements and/or cover letter) how they have pursued and/or plan to pursue these goals in their work.” (There are several other examples from European universities’ psychology departments.)
- The SMU Department of Psychology unanimously adopted a “Policy on Research and Open Science” last year. That policy provides that “the department encourages all faculty members, graduate students, and undergraduates, whenever possible, to engage in what is known as open science practices and employ an online platform such as the Open Science Framework (osf.io) or ClinicalTrials.gov.” The policy then specifically recommends that “all investigators preregister their studies.”
- The University of Southern California’s Academic Senate adopted reproducibility principles that endorse transparency, replication, sharing of data and methods, and even the exact software tool that Arnold Ventures helped develop (“Researchers should in particular consider use of the Open Science Framework as the vehicle for recording and sharing research design, hypotheses, and research data”).
- The University of California — one of the largest research organizations in the U.S. — announced that they would no longer renew a deal with the world’s largest publisher, Elsevier. That announcement followed on the University of California’s 2018 manifesto entitled, “Declaration of Rights and Principles to Transform Scholarly Communication,” which announced UC’s expectations for negotiations with publishers, including “no copyright transfers,” “no restrictions on preprints,” and “no barriers to data availability.”
Perhaps the most thorough and potentially transformative policy was proposed in March 2019 by MIT’s Open Access task force, which has been mandated to improve MIT’s commitment to open science. The task force has recommended that MIT’s official policy on open science include the following:
- A campus-wide open access policy for books and articles written by MIT faculty and staff, and for existing and new grant programs
- A campus-wide Open Data Fund to provide funding to new data repositories (where one doesn’t already exist) and to create discipline-wide databases
- Policies for publishing code openly so as to “reduce the potential negative impact of … software patents”
- An Open Access Infrastructure Fund that would support MIT usage of outside projects like arXiv and the Open Science Framework
- Department-level ways to reward open science practices “in annual reviews and in tenure and promotion packets”
- Greater “recognition and credit” for researchers who engage in open science practices, including new “metrics for assessing impact of non-traditional research outputs, such as data, software, and educational materials”
- Consistent advocacy with the federal government in support of open access policies
These recommendations have yet to be adopted, but MIT is primed to move in the right direction.
Finally, we are currently funding a National Academies Roundtable on Aligning Incentives for Open Science. The goal is to “convene critical stakeholders from universities, funding agencies, societies, foundations, and industry to discuss the effectiveness of current incentives for adopting Open Science practices, current barriers and disincentives of all types, and ways to move forward to align incentives that support common missions and values.”
Members of the roundtable include the president of the Association of American Universities, as well as multiple university presidents (including Johns Hopkins, Trinity, Arizona State, and the University of Houston). With any luck, the National Academies Roundtable and the universities mentioned above can serve as a model to the university system nationwide.
In short, we hope that these efforts at changing scientific culture will eventually affect universities more broadly. In the end, best practices in research will spread further if they are reinforced by all the rewards that a university can offer.