Pay it Forward and Free your Data! Fear in the Way of Data Sharing in Management Research
反思管理研究者对数据共享的恐惧，包括被抢先、数据滥用等担忧，并呼吁通过制度变革、技术工具和集体行动来克服这些障碍，推动开放科学。
Imagine this scenario: You have just published a multi-year, data-rich paper. It was hard, time-consuming, and consequential work. You pushed the boundaries in a real add-on contribution to knowledge. Two days after the paper is published, you Tweet and LinkedIn post your substantive findings, because that is what we do today. You then receive a spontaneous email: ‘I loved your paper. In fact, I am working on something similar. Can you share your data with me or point me to where your data are stored, and maybe show me some of your study processes and analysis plan or coding?’ What would you do? That's the dilemma we faced when receiving this query. When we asked our colleagues about their reaction to this request, most said they know sharing data is helpful, but were just not confident enough in the act of sharing itself. And this got us wondering; does our outlook on data ownership undermine fundamental principles of high-quality and open science? The journey of scientific discovery is built on the bedrock of data sharing, a principle that has propelled fields from astronomy to medicine towards ground-breaking advancements (Anagnostou et al., 2015). Yet, within the corridors of management research, the notion of sharing our data fills many of us with a sense of dread. This fear isn't unfounded; it stems from a combination of concerns over data misuse, ethical breaches, and the potential for our work to be scooped before we've had the chance to stake a claim in the academic community. Our own journey through the academic landscape has been punctuated by moments that underscore this fear. One of the authors recalls the tension that followed a brown bag session where, for transparency, data were shared, only to face immediate backlash from a co-author worried about the research being captured by those in attendance. This incident reflects a broader sentiment that pervades our field: a guarded approach to research, where the desire for collaboration is often overshadowed by a fear of losing control over valuable data and our contributions. Despite the clear benefits of data sharing, exemplified by the development and dissemination of COVID-19 vaccines through open data practices (Duan et al., 2022), the management research community remains trapped in a paradox: We acknowledge the importance of making our data findable, meaning accessible and transparent (Kowalczyk and Shankar, 2011), yet we hesitate, disincentivized and limited by fear and uncertainty. This reluctance hinders not just the reproducibility and integrity of our work but also its potential impact. This essay is a personal reflection on the barriers that keep us all from sharing our data more freely.[1] It is a call to confront the fears that hold us back, to recognize the value of data openness, not just for the sake of compliance, but for the advancement of knowledge itself. As we navigate these complex issues, we must challenge ourselves and our institutions to foster a culture where sharing is not seen as a risk but as a responsibility – a fundamental shift that could redefine the future of management research. In our opening case, we highlighted how our journey into data sharing began after a major project that required years of data collection and analysis, involving significant time, effort, and ownership rights. Although we took pride in our published work, the idea of sharing our data initially seemed to diminish its value, triggering apprehension linked to four fears deeply embedded in academic life. Recognizing and confronting these fears is essential to fostering better data sharing practices. The email request for our data and methodologies raised concerns about the exposure and vulnerability of our research. These concerns stem from need for self-verification, which is deeply connected to professional identities, status, and the validation of our academic work. This fear, grounded in attribution theory, acts as a protective mechanism for our self-image, with researchers often blaming external factors, like insufficient resources or lack of support, for not sharing (Houtkoop et al., 2018). Data sharing risks increased scrutiny and potential criticism, threatening our expertise and self-worth. Despite journals advocating for data accessibility, researchers’ willingness to share, especially qualitative data, remains low due to associated concerns of exposure risks. This fear highlights the reluctance of researchers to share their ‘investment’ in painstakingly collected data without guaranteed protections. When the researcher (unknown to us) approached us to share our data, another immediate concern was the competitive landscape of academia. Why should we share data that took years to gather, given the risk to ongoing data ownership and use, potentially allowing someone else to benefit from our work? Given data use potential, this mindset is deeply ingrained in many of us from early in our academic training. Reflecting on our formative years as junior faculty, the narrative was clear: ‘hold your data close and trust sparingly’. In the scheme of a publish or perish culture, this fear of being ‘scooped’ by competitor peers – of having research ideas or results published by someone else – is a common anxiety, especially given the high stakes of academic progression. A single instance of someone else publishing your data could jeopardize career aspirations, as we have seen with a junior faculty colleague whose publicly accessible doctoral dissertation data were scooped. Scooping prompts a concern of losing control over intellectual property, or a reluctance to expose work to scrutiny, reflecting a judgmental stance by others, further diminishing the overall motivation to engage in open science practices. This fear has profound implications on identity and self-worth in the academic sphere. When approached, our hesitance to share our published data wasn't merely rooted in self-preservation but also in the altruistic aspect of protecting the confidentiality and security of our participants. Given the legal landscape surrounding data privacy, sharing data without robust safeguards can have far-reaching implications (Nahai, 2019). Our responsibility as researchers extends to ensuring the safety and privacy of our participants. We had to grapple with the dilemma of how much risk is acceptable when weighing scientific progress against the confidentiality of non-sensitive data. As the experience of a colleague with qualitative data further highlighted for us, even when data is shared in aggregate form, challenges from readers on finding accuracy and potential legal implications from industry respondents creates a tumultuous sharing environment. The academic landscape isn't just shaped by peer interactions but is also heavily influenced by those higher up in the system: journal editors, publishers, university administrators, funding bodies, media, and senior academics. Given the concentration of power, with their impact on academic progression and reputation, these gatekeepers introduce the element of vertical risk. Sharing data doesn't only open potential scrutiny from peers but also from these authoritative figures. A minor oversight in shared data, if spotlighted by an influential academic or editor, could lead to public retractions or criticisms, jeopardizing our standing. Moreover, when reinterpreted by gatekeepers, data could yield conclusions that overshadow or even challenge original findings. Additionally, there's an institutional layer of concern with universities prioritizing research impact and reputation. Without formal agreement, openly sharing data might be seen as relinquishing control over valuable (technically university-owned) research assets. If data are used in ways misaligned with our institution's goals, could we face repercussions or reduced opportunities? Reflecting status differences in academia, these vertical power structures compound the hesitations around data sharing, adding another dimension to fear. Through our data sharing journey, we have identified two key insights. First, overcoming the fear of sharing begins with self-awareness. Recognizing our initial reluctance as a defence mechanism was crucial, leading us to reconsider our stance on data sharing. We understood that, though legitimate, our fears could be mitigated with a proper mindset and tools. Second, we acknowledged the necessity of transitioning from individual ownership to collective stewardship. This shift entails not only changing practices but also adopting a mindset of collaboration over apprehension. Failure to confront these fears risks reinforcing the notion that sharing is too hard, ignoring the potential for collective management of these concerns. Our experience illustrates the importance of these insights. Initially focused on problems, we shared data in a way that alleviated our fears. Upon reflection, we identified six actionable steps to address data sharing apprehensions, highlighting the value of a comprehensive strategy to tackle these fears. Our initial venture into data sharing resulted in a mentorship program within one of our departments. After our project, we initiated a scheme pairing experienced researchers with junior faculty to guide them on ethical data sharing, anonymization of sensitive information, and intellectual property management. This mentorship, transcending technical advice, was pivotal in fostering confidence and trust in sharing all types of data, alleviating fears of exposure and criticism. The emotional support and practical guidance from mentors proved crucial in helping researchers (and us) navigate the apprehensions of open sharing. Reflecting on this, we recognized the benefits of a comfortable, phased approach to sharing. Starting with sharing data among a small, trusted group allowed us to fine-tune our approach in a safe setting, gradually extending our reach to a broader academic audience and then the public. Implementing structured feedback mechanisms also helped address concerns of data misinterpretation and facilitated constructive peer review sessions. These sessions, emphasizing respect and constructive criticism, promoted a culture of open dialogue and collaboration, essential for shared learning and research advancement. This experience underscores the importance of systemic and collective efforts (Bouckenooghe et al., 2023) to foster an environment of openness and collaboration. The lack of standardized data sharing protocols often pushes researchers towards makeshift solutions, making sharing daunting. This issue is particularly evident in journal policies, where data sharing norms vary significantly. Our experience with submitting supplementary data from a multi-year study to journals underscores this challenge. We encountered resistance from journals that deemed our paper ‘too long’ due to the additional data (especially its qualitative data), despite our intention to promote transparency. The optional nature of data sharing in several management journals inadvertently reinforces the prevailing culture of data hoarding, stifling potential progress. Without mandatory data sharing policies, the implicit message is that data sharing is a secondary concern, failing to address common fears of criticism or being scooped. Transitioning to mandatory data sharing and management plans by journals and funding bodies could cultivate a more open research culture. Adopting editorial policies like those of Science (https://www.science.org/content/page/science-journals-editorial-policies#data-and-code-deposition), serve as a model for boosting research visibility and impact, necessitating a strong commitment from all involved parties. Furthermore, the academic community often struggles with the specifics of responsible data sharing, a challenge exacerbated in the realm of qualitative data, due to a lack of guidance on suitable software, licensing, and storage. Bridging this gap with comprehensive data management training could normalize data sharing, transforming it from a punitive measure to a standard practice. Such training would provide researchers with the skills needed for secure data sharing, including sensitive qualitative data, thereby reducing risks and promoting openness. While training individual researchers on responsible data sharing is crucial, it is important to acknowledge that broader, collective solutions are necessary to address systemic data sharing challenges. Institutional reform is crucial for enhancing data sharing among researchers by mitigating perceived barriers. This open science action includes providing data management resources, establishing supportive policies, and incentivizing sharing. Such reform can increase a sense of agency among researchers. They need not be radical but should integrate into existing frameworks through collaborative efforts, promoting a culture where data sharing is normalized and valued without prejudice. Initiatives could involve embedding data sharing importance in university recruitment, paralleling mandatory training in ergonomics or safety, and fostering inter-institutional collaboration to streamline processes like metadata storage. Leveraging models from other fields, academic bodies can adopt standardized, accessible data sharing systems, akin to two-factor authentication, to notify researchers about the use and impact of their data, thereby encouraging a culture of mutual benefit. Furthermore, funding agencies and institutions could mandate data sharing for grant or tenure considerations, prioritizing its role in research impact. The key challenge lies in initiating this shift, determining who will lead the way in institutionalizing data sharing as a fundamental research practice. Recently, while preparing a manuscript, we faced a common dilemma: a journal required data availability for submission. Intended to enhance research quality, this requirement sparked concerns about moving beyond our comfort zone. However, the journal's precise instruction and guidelines led to a positive encounter, as an exercise in providing rigour and transparency to sharing. This transparency can help bolster the credibility of our findings. The experience taught us that data sharing need not be an all-or-nothing approach but can be more nuanced and tailored to recognize the breadth of sharing challenges. Institutions and funding bodies play a vital role in this landscape, offering training and resources on data sharing benefits and incentivizing researchers for their past sharing efforts. The adoption of open science methods for all types of data, like the transparency and openness promotion (TOP) guidelines (Miguel et al., 2014), provides options such as pre-registration of data collection and open access publishing. Additionally, negotiating fee waivers with publishers for data sharing can cultivate a culture of transparency. Public data repositories with open access plans, supported by platforms like Figshare, Zenodo, Dataverse, and Data Dryad, are crucial for this shift. Journals can encourage conditional data sharing and citation, ensuring security and longevity through such stable repositories and controlled access via DOIs. Sharing data without explicit consent, however, requires a balance of ethical concerns and potential IRB processes. Ethics committees must navigate these complexities, possibly redrawing conduct codes or creating broader data access types. In searching for data sharing solutions, we also considered the potential of blockchain technology for managing sensitive data. By leveraging encryption within a decentralized network, blockchain ensures participant anonymity and addresses privacy concerns, thereby fostering trust in the research ecosystem. This technology offers incentives for data sharing such as tokens or compensation for open access publication fees, while aligning with initiatives like the Center for Open Science's open data badges (https://www.cos.io/initiatives/badges), nurturing a sharing culture. Blockchain reframes errors in data reporting as part of the scientific process, diminishing the fear of judgement and criticism, while innovations like smart contracts guarantee data access solely to authorized researchers, mitigating worries about data misuse (Macrinici et al., 2018). Simultaneously, blockchain helps bolster sharing sensitive qualitative data, by prioritizing the implementation of sophisticated deidentification techniques. Utilizing advanced methods like natural language processing (NLP) and AI-based tools, including named-entity recognition and pattern matching, enables researchers to effectively anonymize sensitive information, safeguarding participant privacy without compromising data integrity. Institutions championing these advanced deidentification methods take a proactive stance in privacy protection, enhancing the willingness among researchers and participants to share data, and thereby enriching the collective research landscape. The integration of blockchain technology and advanced deidentification techniques, can enable more knowledge transparency. To effectively mitigate the vertical risks in academia associated with data sharing, gatekeepers need to spearhead initiatives that foster a supportive and collaborative research environment. Creating a protective framework for data sharing is a critical first step, where policies safeguard researchers by ensuring that critiques or reanalyses of shared data involve the original contributors, maintaining context and integrity. This approach should be complemented by promoting a culture of constructive engagement, where the academic community prioritizes collaborative improvement and knowledge advancement over punitive scrutiny. Such a culture relies on open dialogue and constructive feedback on data, helping to alleviate fears of reputational damage or data misinterpretation. By hosting forums, workshops, and discussions that bring together researchers, policymakers, and the public, gatekeepers can cultivate a more nuanced understanding of the complexities and benefits of data sharing. This collective effort not only demystifies the process but also highlights the communal benefits of open science, ensuring that the pursuit of knowledge is a shared endeavour, marked by transparency, respect, and mutual support. In drafting this essay, one of us faced a common hurdle in academia: seeking specific data (correlation coefficients) from authors and encountering non-responsiveness or refusal. This experience underscores a wider reluctance in management research to share data, fuelled by valid concerns and perpetuating a cycle of withholding information. But this reluctance stalls the development of data sharing guidelines and policies, limiting the field's growth potential. Open science holds universal benefits, yet fears about data sharing persist, necessitating innovative approaches and a shift in mindset. Drawing from successful models in fields like science or medicine, we can overcome these barriers for both qualitative and quantitative data. Advancing management research requires a collective effort to promote data availability, engage in dialogue on research practices, and sharing training. By confronting fears around data sharing, we can foster a more collaborative future in management research, embracing open science to unlock our data's full potential. Open access publishing facilitated by University of New South Wales, as part of the Wiley - University of New South Wales agreement via the Council of Australian University Librarians.