Values and Attitudes

Values and Attitudes of Open Science and Open Source

We asked interviewees and survey respondents about the core values of their software projects in order to understand what primary motivations exist for individual contributors and teams. While the responses varied, most participants spoke about impact, improving science, equity, accessibility, openness, or quality/stability of the software. To dig into the practical, symbolic, and ideological importance of these projects being Open Source projects, we also asked our interviewees what the effects of their projects being OS is: Why is OS important for their work?

Notably, no interviewees mentioned usability or design as central values of their projects.

While many discussed the utility of openness and the ability for users to contribute to scientific progress, these aspects of usability were not explicitly acknowledged as such. And, from our other sections on Perceptions of Design and Usability, we can infer that project leaders did not mean to imply an understanding or affinity to usability and design in these responses per se.

Contributors to SROSS are motivated by ethical and practical considerations.

OSS is an ethical choice

“If you get to put profit aside entirely and it’s all about sharing knowledge, it’s really hard to argue that things shouldn’t be open.”

“More often than not doing things in the open is the right thing to do.”

Many participants explained their affinity to Open Source and open practices was based on an ethical argument: not only is practicing open source “the right thing to do” but it actively works against forces that interviewees disavowed, such as being profit-motivated. One respondent referred to OS as “a movement” with idealist purposes, highlighting that they contribute to OS projects to work towards a better and more ethical future.

OS is practical - despite recurring complications around maintenance and financing, the use and ideology behind OS is practical in essence.

“For an analysis tool, it makes perfect sense for it to all be open, because you’re literally going to be coding it”

Beyond ethics, many participants referred to OS as an obviously practical way of working. In simple terms, it “makes the software better.” Some emphasized that, more than anything, it’s free, which is an inherent advantage. Two people mentioned that it enables researchers in the US to get government funding, because the US government requires the projects they support to be built with open-sourced code. As researchers, some respondents highlighted the practicality of OS in supporting research. OS projects enable research values such as open data, open access to information, and reuse of research infrastructure (like formats and metadata). Two respondents pointed out that other researchers, scientists, and academics are drawn to being part of “free knowledge movements” and that the free and open environments that OS projects cultivate are a factor in attracting new project contributors.

OS’s practical value is also rooted in its adaptability. Multiple participants highlighted that OS tools give researchers and programmers the freedom to adapt the tools to their own needs – no matter how niche they are. One participant noted that this allows people to research “questions that would probably be otherwise unanswerable.” Another said she is able to “go from theory to practice really quickly.” Not only is that useful for expanding opportunities for research, but it also helps the tool and software itself improve over time. Participants said that the freedom to change things and address issues is central to OS’s power.

Open Source improves and furthers scientific work and progress.

“One of the core tenets of science is that it needs to be reproducible. That’s how we establish facts. With models that are developed based on computer code, it’s 100% essential that that code is shared openly so that other people can run it for themselves and prove that they are getting the same results.”

“Openness leads to better science”

We heard repeatedly that OS is in ideological and methodological alignment with the scientific method: its practices are based upon replication, validation, challenges, and dialogue. Many respondents said that when doing code-based analysis, OS is critical for peers to be able to analyze and replicate findings. Beyond simple replication, OS enables science to progress and improve. As one interviewee said, “the important part is really to be able to build upon each other’s work, and test each other’s work and critique each other’s work.” Others emphasized that openness signals stronger science as it invites collaboration and challenges. Throughout a number of responses was a sense of altruism and dedication to the common cause of science, rather than a competitive spirit between researchers.

Openness actively helps to interrupt some of the more difficult dynamics and barriers within academia.

Following the spirit of optimism for the “good of science” noted by participants, we also heard that OS culture runs up against “the culture in some of the fields where people want to keep their code to themselves, so that they’re the only one that can do it.” Inherently, OS invites participation and forecloses the option to be exclusive and proprietary to get ahead.

Participants also emphasized that working in the open enables them to do “simple things” that are, for many reasons both cultural and institutional, difficult to achieve within academic circumstances. For example, storage of data and content for reuse and access over time is often tied up in ownership and rights issues within institutions. One respondent said that through open source practices:

“We are simply talking about storing content, and augmenting the chances that you can find it back. So the idea that if you publish an article, you should be able to find it associated with data that allows you to reuse them.”

The same respondent went on to explain that commonly, researchers run into frustrations, like your data being in “an institutional repository there in your old computer that is broken.” Open licenses allow researchers to access their data even if they don’t work at the institution any longer.

Academic competition and privacy concerns often interrupt the OS research ideal.

“Openness is a continuum and has trade-offs associated with it.”

While nearly all participants espoused the virtues of Open Source, many identified limitations that interrupt “perfect” openness, including academic pressures and contextual needs for privacy. In certain situations, all code, content, and data cannot be made open. Participants referred to working with data related to sensitive subjects (like biomedical research) that for good reason cannot be made openly accessible. One person explained a compromise: their project requires a verification process to ensure that only authorized researchers can access sensitive data. Others work for government entities that keep some information under lock and key for security reasons.

On the other hand, some urged that the pressures on people working in academic contexts can lead to competition and secrecy at the expense of openness. One participant shared:

“We may think that the kind of culture of open science is idealistic and better than commerce, but in reality, professors are extremely competitive, labs are extremely competitive, and everyone is trying to get their little piece of the pie.”

A lot of this scarcity mindset is framed by the need to publish, limited resources, and fears that sharing one’s tools, data, or findings will undermine the value of a potential publication with respect to the author’s career growth

Institutionalizing Open Science and Open Source at large within academia and labs requires intentional planning and cultural buy-in.

While Open Source as a practice can interrupt some of the difficulties researchers experience within academia, participants were clear that the process of codifying and popularizing Open Science practices is still an uphill battle. Many urged that this process itself needs to be carefully planned and strategically managed. Universities are not “prepped” to work openly, because “Open source infrastructure and academic infrastructure aren’t linked or in conversation.” Part of this broken linkage exists in logistical aspects, like data infrastructure, which one participant cited as critical to successful OS and open science. The research pipeline, from ideation to dissemination and storage, also needs to be retrofitted for openness:

“The element of thoughtfulness and design that goes into actually developing research is kind of lacking outside of: ‘Hey, this is an interesting problem, how do we solve it?’ There’s a whole aspect of that which is: How is it applied? How do people engage with it? How do you publish it? How do you make it open? All those things aren’t really thought about.”

This need for intentionality stems from researchers doing their work in very distinct and often stressful contexts with competing incentives and specific requirements. On the most basic level, researchers are often most concerned with sustaining their work in the short term, including securing grants, publishing to achieve tenure, job stability, etc – “why should they care about openness when they need to survive?” one respondent said. Another brought up a resonant fear: “people don’t want to harm their CV or next grant application by sharing” too much of their work. Openness isn’t currently measurable in the “incentives economy” of academia. It is, to some, overshadowed by publishing, tenure, grants and “newness.”

While cultural overhaul of academic incentives is certainly one way to address one of the challenges participants identified, there is also an opportunity for developing cultures of openness in a way that can give researchers the currency they require within their institutions. Where a participant identified a “growing nervousness in OSS generally about metrics and measuring” perhaps OSPOs and others at the forefront of Open strategy can seek to develop methods of measuring openness.

Our recommendation/s:

• Choose to store your design, usability and research data with institutions that are more likely to reuse it: Work toward openness by choosing to store your research data with institutions that are more likely to reuse it, or even partnering and collaborating with people who are more likely to reuse it.

• Measuring openness: OSPOs and others working on open strategies should develop methods of measuring openness.