Early career researchers in open science: vanguard or cannon fodder?

Moves towards a more reproducible, transparent, open way of 'doing' science have been creeping into the mainstream over several years. At the vanguard of this culture shift has been a generation of relatively junior front-line scientists, many sharing a shiny vision of robustness and transparency. Perhaps the most prominent example of a bottom-up movement headed by early career researchers (ECRs) is ReproducibiliTea, an tea-drinking journal club led and organised predominantly by PhD students which centres around discussion and debate of all things reproducible. At the last count, it has homes in 92 institutions and represents 25 different countries.Therefore, it is perhaps unsurprising that the first ReproducibiliTea virtual meeting, hosted by Surrey Reproducibility Society, attracted an online crowd of 300+ attendees and a suitably energised Twitter following. 

Anne Scheel, a PhD student at Eindhoven University, led the proceedings with a lecture on ECR contributions to the quest for reproducible science. Scheel started with a justification for why ECRs are well placed in informing, advocating for, and promoting the values of open science. Scientists early in their career trajectory are, as Scheel notes, generally 'progressive and idealistic', not least because there is no career forged through 'closed science' to defend or protect. There are also some practical benefits to promoting open science principles to junior researchers. In a research lab, PhD students and post-docs are usually the ones who are 'closest to the data' and, therefore, have the most influence over the reproducibility and robustness of the way that data are handled. As Scheel puts it, 'we need the people who interact with the data to adopt new practices'. 

That is not to say, however, that there is nothing to lose in engaging with these practices as a PhD student. It is important to consider how 'transparency makes you vulnerable'. This may likely be the most immovable barrier to working openly as an early career researcher attempting to publish for the first time. Openness, be it of data, analysis code, or research materials, may be progressive, robust, and useful for the future of the scientific field. But it is also hard. 'Open science takes a lot of time and extra resources,' Scheel said. 'That's easy to overlook if you haven't done it yourself.' 

New practices can also be incredibly exposing. For example, when senior researchers 'go digging around' in the open data, the capacity for criticism can 'place a huge burden on ECRs', Scheel warned. 'People who do open science really well – we will find more errors in their work, compared with those who work in a less transparent way.' Whilst we should remain sensitive to these issues, our capacity to catch errors when they exist 'is a central pillar of science'. 

There are some important power dynamics at play here too, which feed into these discussions and must also be acknowledged if academia is to progress. Scheel mentioned ECRs who had raised issues with the data of senior researchers then receiving 'a tonne of emails' in response, which she likened to a 'denial of service' attack overwhelming the capacity of the ECR. These attacks can, if science is not careful, serve to deter and silence the progressive, new voices of academia. 

Scheel ended her presentation by stressing that open science can be a slow, frequently frustrating process that brings with it issues of power, credibility, and support in the early-career stage. ECRs are a unique group, in that they may be largely invested in the practices of transparent and robust working, but also at risk of reputational scrutiny at a time of continuous learning and skill refinement. We must, therefore, respond to these pressures with the same principles of openness and collaboration that the very movement of open science is based on. 

In line with the theme of transparency of experience as well as 'the science' itself, following Scheel's talk, we heard from a panel of ECR and established academics, each with their own story of reproducible science. 

This discussion quickly got into the meaty, tangible 'how to's' of open science in the early career stage. Given all of the very present barriers highlighted by Scheel, what should open science look like in the early career stage? How can ECRs engage meaningfully and openly with this process? Jess Butler, a health data researcher at the University of Aberdeen, noted that her adoption of open science principles made her way of research feel 'not radical, just rearranged'. This sparked a discussion of how moves towards open science should be communicated beyond the silos of researchers who readily promote and adopt these practices. 

Dr Amy Orben (University of Cambridge) suggested that the reputation of the open science movement itself is crucial. It is important, when faced with the task of shaking up the fundamentals of research culture and integrity, to 'make things feel official', Orben said. This can, as Butler echoed, add a sense of 'legitimacy to the cause'. Grassroots, ECR-led, person-centred movements are most effective when they have institutional and organisational support. Without this support, long-lasting change becomes more difficult. 

Following this recommendation, all panel members emphasised the continuing need to change mindsets of these more emergent research practices. As Professor Emily Farran (University of Surrey) noted, this can be especially challenging for PhD students who work with supervisors who are not readily engaged with open science practices. Against a backdrop of job precarity in the early stage of an academic career and the enduring pressures to publish, this can be a significant barrier. To work past this, Farran recommended that ECRs consider framing the benefits of open science towards the distinct motivations of those who may 'govern' elements of the research process, for example, supervisors, PIs, or lab managers. Ben Bleasdale (Wellcome Trust) supported this, noting that until open science is valued by those in positions or seniority for the careers of 'those they want to see succeed, things won't change.'

The concept of wide, system-level, comprehensive, change largely follows on from some of the fundamental ideas in Scheel's opening presentation. For example, as a useful analogy to science as a whole, Scheel emphasised that teaching about open science principles and skills at the undergraduate level, including coding, data sharing, and sophisticated statistics, should not be considered an 'extracurricular add-on' that accompanies core research training. Rather, these principles should be embedded into a 'bigger shift of our workflows'; these new workflows, with accompanying priorities, should be set up to accommodate data sharing, robust practices, and a 'slower' more considered way of doing science. 

Overall, both Scheel's talk and the panel discussants recognised the significant barriers that ECRs face in contributing to long-term systemic change within research culture and practices. Precarity, which can be related to employment, finance, and researcher identity, was cited frequently as a fundamental barrier that all ECRs face, in some way. This cannot be overlooked in the conversation surrounding 'best practice' in research. As Butler summarised 'we must not be ignorant to politics'. This was epitomised by one delegate during the Q&A segment of the panel discussion, who asked 'is reproducible and responsible science only for those who are financially privileged?'. 

This question neatly summed up the distinct challenges faced by early-career researchers and made the crucial differentiation between ECRs and more established academics in permanent employment. If 'gold standard' research, with transparency, reproducibility, and robustness at its core, relies heavily on local resources, security of position, access to support, and capacity to work in a 'slowed down system', does this not preclude those without access to these privileges? The webinar ended with an agreement that accessibility, including a critical appraisal of how privilege feeds into this, should be a priority for the future of open science progression. 

- Madeleine Pownall is a Graduate Teaching Assistant and PhD student at the University of Leeds.