How to Collaborate with Industry Using Open Science

Authors: Samuel Robson, Brooklyn Corbett, Duncan McCarthy, Gianni Ribeiro, Rachel Searston, Jason Tangen, Matthew Thompson

---

After several years working on fundamental cancer research in the 1990s, immunologist James Allison and his colleagues published research with potential for groundbreaking impact on cancer therapy. Yet when he approached companies for collaboration, many turned him down, and once he did secure an industry partnership the project ended unsuccessfully. Many clinical trials and more than 15 years later, Allison’s Nobel Prize winning research was finally translated into an effective drug available to cancer sufferers. Because the fruits of collaboration are often highly consequential for people and society, it’s important to get it right.

A lot can be gained from collaboration between universities and industry. It fosters research that has real impact, delivers products to end-users more quickly, provides students the opportunity to gain experience in industry, and generates momentum for even grander projects to be undertaken in the future. Despite these benefits, industry collaboration is not without its challenges, as Allison’s experience shows. Without close (and open) collaboration with end-users, there can be too few resources available, legal and administrative processes can be difficult to navigate and different stakeholders can have misaligned goals.

Our research team of several academics, Ph.D. students, and research assistants spanning across four universities, recently received funding to work with six of Australia’s policing agencies. The project aims to improve the training of fingerprint examiners in the police force by applying what we know from psychology and the science of expertise.

A major consideration of ours was how to coordinate a large scale project with so many stakeholders — researchers, industry partners, funding bodies, participants, and the broader public. In the past, we have used the Open Science Framework (OSF) only as a tool to preregister our experiments. It turns out that it is extremely useful in addressing many challenges we faced in this industry collaboration too.

Using the Open Science Framework with Industry
Lack of effective communication is a fundamental barrier that impedes progress, but we have found the OSF, with all its links, tools and customizability, to be a fantastic platform on which to communicate effectively with stakeholders. Providing our partners with links to projects or even adding them as contributors has allowed us to share our work from the ground up at the inception of an idea. Those who are interested in a particular project can receive updates about it and see how it progresses in real time. 

Tailoring a preregistration to make it easier to navigate is therefore critical. We often detail our wiki extensively so that any reader can glimpse a snapshot of the entire study. We include a brief description of the rationale for the project, the participants we need and why, along with any exclusion criteria, the experimental design, materials, procedure, and the planned analyses — the nuts and bolts of any preregistration — but also a few additions to ensure transparency and communication between our partners.

First, we provide videos of our experimental instructions because it allows the reader to take the perspective of a participant in the experiment in only a few short minutes. We also outline our predictions about magnitude and effect size to ensure that they are clear and falsifiable. Precise predictions like these can also clarify to readers what we expect to find and why. Last, we upload pre-written data analysis code (for example, an R Markdown script), run simulated participants through it who respond randomly, and then present figures based on these simulated data. This set-up provides both an end-to-end test of the system and illustrates to readers exactly what kind of data they can expect from the experiment.

Genuine Collaboration
Stakeholders in any collaboration can be contributors on OSF preregistrations like these. In turn, they will feel like collaborators rather than end-users who have the final product of the grant thrust upon them without much input or insight as to what goes on behind the scenes. Access to the preregistration can also foster effective coordination between different stakeholders (who needs what from whom) and ensures that everyone’s goals are met.

If partners are up-to-date with the project as OSF contributors, these regular updates can even spark scientific inquiries. A back and forth between researchers and industry partners over the design choices in the experiments, or involvement in the writing process, can open doors for co-authorship on the resulting scientific papers. It also smooths the way for future collaborative projects and funding opportunities.

Roadblocks to Publication
Many industries, as well as universities, have complicated processes to navigate in order to get things done. It often takes considerable time and effort to sign off on any aspect of a project that is released to the public, including information on the OSF. The problem is magnified further when dealing with sensitive or confidential information, which is all too common in many industries, including police agencies. Dense bureaucracy and vested interests to manage how information is presented to the public means that we need to consider what information is being presented on preregistrations and any resulting publications.

A useful fix is to embargo preregistrations until the very end of the project. An embargo means that only contributors have access to the preregistration until it is lifted, giving plenty of time for collaborators to approve any publicly available information. Once the project ends, the preregistrations can be made public. Embargoes give industry partners peace of mind and significantly reduce administrative and legal barriers because approval is needed only once.

Getting the Most out of Resources 
Resources are always limited when doing research and it is no different when working with industry. Industry collaborators, like the police, contribute significant resources, including funds, on-the-ground coordination, and participants, all of which are crucial in psychological research.

Testing people on a barrage of psychometric tests — like we do with fingerprint examiners — makes them trepidatious about participating in our experiments. In part, this is because it is often unclear to participants what we as researchers are doing at their workplace. The purpose of our visit is often privy only to our senior partners. But, we want the practitioners who are volunteering their time to feel at ease and genuinely part of the research process. So, to combat uneasiness and better involve practitioners in the research process, we have ambassadors from each of the agencies who we work closely with as contributors on projects; these ambassadors are well-respected in the workplace and enthusiastic about the research. They receive alerts about what is happening with the project and can then inform their colleagues about the general nature of participation, which eases their anxiety.

If more researchers adopt open science tools like these when working with industry or end-users, it can substantially improve the quality of collaboration. Customising the OSF to make it more navigable, and keeping partners and ambassadors in on the action, can build trust and communication, and lays the groundwork for longer term and more impactful research opportunities in the future.

---