Equity, Diversity, and Inclusion in Institutional Research Data Management Strategies: A Contributive Justice Approach

: Research data management (RDM) is a field of emerging concern for academic librarians. As funder agencies increasingly mandate institutions and researchers to ethically and responsibly manage their research data, academic librarians are frequently tasked with creating institutional strategies and services to support researchers. This article explores how a racialized librarian at a medium-sized, teaching-focused Canadian university created an institutional research data management strategy through a process informed by critical librarianship research and contributive justice (Gomberg, 2016; Honma & Chu, 2018). It examines the lack of equity, diversity, and inclusion (EDI) principles in both funder directives and RDM research literature and proposes an approach to do institutional RDM work in an EDI-centered way. To cite this article:


INTRODUCTION
In 2021, three federal research funding agencies in Canada published the unified Tri-Agency Research Data Management Policy, which acted as an impetus for some small-and medium-sized, teaching-focused Canadian higher-education institutions to develop and implement research data management (RDM) services for their researchers.Among its requirements, the policy calls for institutions to publish an institutional research data management strategy in order to remain eligible for Tri-Agency funding opportunities (Canada, 2021a).While the policy outlines some points that institutions should include in their strategies, institutions were otherwise left with flexibility to craft their strategy for their local context and needs.Importantly, the data sovereignty of First Nations, Métis, and Inuit communities, collectives and organizations was emphasized as a distinct item in the policy (Canada, 2021a); yet the policy lacked a more expansive application of equity, diversity, and inclusion (EDI) principles despite the agencies releasing the Tri-Agency EDI Action Plan 2018-2025 in the same year.The research literature on RDM is similarly quiet on the application of EDI principles in institutional RDM strategies and services.Scholars explore how the RDM space is negotiated between different groups at institutions (Andrikopoulou et al., 2022;Castle, 2019;Cox et al., 2016;Cox et al, 2017;Donner, 2023;Faniel & Connaway, 2018;Howie & Kara, 2022;Pinfield et al., 2014;Verbaan & Cox, 2014), but do not address how systemic inequities impact these collaborations.
The terms equity, diversity, and inclusion can be used to identify and parse systemic inequities.Equity, diversity, and inclusion are interconnected concepts that are sometimes treated as one; however, they are distinct.Diversity is about individuals and the variety of unique characteristics and experiences they each possess (Canadian Centre for Diversity and Inclusion [CCDI], 2023).Diversity may be expressed through race, ethnicity, gender, sexual orientation, religion, economic status, and physical abilities, but is not limited to those characteristics (CCDI, 2023).Equity is where people are "treated according to their diverse needs in a way that enables all people to participate, perform, and engage to the same extent" (CCDI, 2023, p. 10).Inclusion is a "mindful and equitable effort" to "creat[e] a culture that embraces, respects, accepts, and values diversity" (CCDI, 2023, p. 11).While scholars such as Sara Ahmed (2012) have problematized EDI's overuse by institutions, these concepts may still be useful in discussing the gaps that exist at institutions.
As a racialized librarian charged with creating my institution's strategy, I felt the notable absence of EDI principles in the RDM literature reflected the unspoken whiteness dominant in our libraries and institutions, as identified in critical librarianship research (Chiu et al., 2021;Ettarh, 2018;Honma & Chu, 2018;Jordan, 2023;Roh, 2018).In this article, concepts developed in critical librarianship are used to address the gaps in the RDM literature; namely how vocational awe (Ettarh, 2018) and invisible labour (Jordan, 2023) shape institutional RDM work.Drawing on the contributive justice model (Gomberg, 2016, Honma & Chu, 2018), this article offers a reframe of intra-institutional collaborations.I will discuss and reflect upon my process to create a strategy based on relationships with the people who would be affected by it and how that shifted the strategy creation process from an administrator-led one to a colleague-led one, establishing a foundation for doing research data management work at the institutional level that is EDI-and personcentered.This article contributes a narrative to the RDM literature that reflects my experiences as a racialized academic librarian creating an institutional RDM strategy while navigating a predominantly white institution.

CANADIAN CONTEXT: Tri-Agency Research Data Management Policy
In Canada, three federal agencies provide funding to "support research, research training and innovation in Canadian postsecondary institutions": the Canadian Institutes of Health Research (CIHR), the Natural Sciences and Engineering Research Council of Canada (NSERC) and the Social Sciences and Humanities Research Council of Canada (SSHRC) (Canada, 2023a, para. 1).Referred to together as the Tri-Agency, CIHR, NSERC, and SSHRC released the Tri-Agency Statement of Principles on Digital Data Management in 2016, followed by the more detailed Tri-Agency Research Data Management Policy in 2021.
Intended to support Canadian research excellence, the policy has three main pillars (Canada, 2021a).The first pillar requires funding-eligible institutions to create and publicly share an institutional research data management strategy by March 1, 2023 (Canada, 2021a).The use of 'institutional strategy' in this first pillar suggests that institutions should approach it as an action plan rather than a governance policy: "An institutional RDM strategy describes how the institution will provide its researchers with an environment that enables and supports RDM practices" (Canada, 2023b, 4.a.).The other two pillars of this policy are more focused on the researcher and are not yet fully implemented (Canada, 2023b).They include the requirement for researchers to submit data management plans with their funding applications and the requirement for researchers to share their data in open repositories, as appropriate (Canada, 2021a).In order to maintain eligibility for Tri-Agency funding, institutions and researchers must fulfill these requirements as they are incrementally implemented (Canada, 2021a).While larger Canadian research institutions and other institutions actively engaged in research activities had established robust RDM supports before the policy's release (Cheung et al., 2022), the Tri-Agency policy provided an impetus, as well as resources and guidance, for the development of institutional research data management strategies at small-and medium-sized Canadian institutions that are in the process of transitioning to support research activities.As noted in the literature, research funders' mandates were an effective stimulus for the development of research data management strategies and services in other countries (Cox et al., 2016;Howie and Kara, 2022;Koltay, 2016).At small and mediumsized Canadian institutions that are only starting to establish research support services, the Tri-Agency policy encouraged the development of institutional RDM strategies and services along the timeframe set by the policy.

MY APPROACH
"think of strategy not as the absence or bracketing of thought…but as the unfolding of thought" (Ahmed, 2012, p.5) Given the task of creating my institution's strategy, I felt strongly that equity, diversity, and inclusion principles should form an essential part of both the final strategy and, more importantly, the strategy creation process.My positionality plays a role in this perspective.I am a cis-gender, racialized librarian working at a predominantly white institution.I identify as Chinese Canadian with roots in Hong Kong.In my professional career I have only worked in predominantly white institutions with white administrators, and at times, have been the sole librarian of colour.My reflections are inspired by Sara Ahmed's (2012) approach in her work On being included: Racism and diversity in institutional life, in which reflections of her experiences prompt her to consider her relationship to institutional life and policy.As much as this article is about creating a research data management strategy, it is also about navigating institutional life as a racialized person.
Through her interviews with diversity practitioners, Ahmed (2012) noted the frequent criticism that "too much research is premised on findings that institutions want found: from toolboxes to good practice" (p.10) rather than reflecting the messiness of how this work is actually done or reflecting the problematic power structures that hold up our institutions.The goal of this article is not to provide a toolbox for accomplishing EDI in RDM; rather, I contribute my experiences in how I incorporated EDI principles in RDM work.I use racial and gender equity as an entry point to problematize RDM discourse; however, this is not to minimize the other ways people experience marginalization in academic institutions.Writing from my perspective as a racialized librarian, this article contributes to filling a gap identified in the library and information studies (LIS) literature by Bright (2018).Bright (2018) notes the lack of attention on women of colour or people with intersectional identities in LIS research.In particular, there is no research on women of colour or people with intersectional identities who hold specific job roles; the little research available is primarily focused on collection development and book reviews, LIS as a whole, and LIS leadership (Bright, 2018).Bright (2018) posits that this lack of research leads to further marginalization of these LIS workers.In this article, I offer reflections on how the concepts of diversity, equity, and inclusion can and should shape the creation of institutional research data management strategies, which includes making visible my experiences of invisible labour as a racialized librarian.This work is important to me as it is informed by my belief that EDI work should not be confined to isolated units within an institution, that justice should be an important factor in all institutional work.Importantly, this work needs to be situated in my institutional context: where I am, where I stand in my institution's structure and spaces.
I work at a mid-sized public post-secondary institution in British Columbia, Canada.While it began as a college in the 1960s, the institution was designated a special purpose, teaching university in 2008 by the provincial government (University Act, 2008).Distinct from researchfocused universities, a special purpose, teaching university is a post-secondary institution mandated to serve a specific geographical area and provide educational programs (University Act, 2008).There is no external mandate for the university to support research.Faculty at my institution are hired to teach, not to do research; however, faculty and staff have and continue to pursue research regardless of mandate and lack of pay.The university has only started to provide institutional support for research through the creation of a research office and a research ethics board in the past decade.This institutional context is important as the university's identity is still strongly grounded in its roots as a college and in its positioning of research as ancillary to teaching, with research as "integral to building and maintaining our culture of learning and innovation" (Capilano University, n.d., para.1).
While other larger, research-focused Canadian institutions had done years of work establishing RDM strategies and services ahead of the Tri-Agency policy (Cheung et al., 2022), my mid-sized, teaching-focused institution did not have the capacity to address research data management before it was required by funder mandate.This is not uncommon: Perrier et al. (2018) noted in a meta-ethnographic analysis of academic libraries that "[r]eactionary responses were the norm, as sufficient time was not provided to be proactive" (p.177-8).Patterton et al. (2018) focused on resource-constrained institutions' implementation of RDM services and "found that it was not so much the researcher behaviour, but rather the approach to implementing RDM that needed to be adapted to make provision for a different set of circumstances and priorities" (p.15).Pinfield et al. (2014) note that whether an institution is research-led and teaching-led plays a role in how research data management services are offered, and others acknowledge that the size and nature of the institution determines the scope of these services (Cox et al., 2017;Donner, 2023;Patterton et al., 2018).In addition to its teaching focus, another key part of my institution's character is the collegial management model, which supports decision-making at the functional level (Agreement, 2019).Unlike institutions with a vertical organizational structure, this flatter organizational structure creates a much different relationship between faculty and administrators.In specified functional areas, faculty groups make recommendations rather than having direction handed down from administration (Agreement, 2019).To create the strategy, I received release time from some of my librarian work with funding from the research office and worked with the Associate Vice President, Creative Activity, Research and Scholarship and Graduate Studies; however, this collegial structure provided me with flexibility in how I approached the task of creating the strategy and offered me the freedom to explore how I could make EDI central to strategy development.
As I began the work of creating my institution's research data management strategy, I looked to Tri-Agency documents and the academic literature for guidance on how EDI principles can be enacted in RDM work.

EDI AND THE TRI-AGENCY
In 2021, the year that the federal research agencies released the Tri-Agency Research Data Management Policy, the agencies also published the Tri-Agency EDI Action Plan for 2018-2025.Problematically, the action plan does not provide individual definitions of equity, diversity and inclusion, noting that "the phrasing of equity, diversity and inclusion (EDI) [is used] as the overarching objectives informing the initiatives" (Canada, 2021b, endnote 2), and outlines some examples of underrepresented groups.The action plan directs the agencies to "address systemic barriers that limit the full participation of all talented individuals" and that they "must create a culture where embedding equity, diversity and inclusion (EDI) considerations into all aspects of research is second nature" (Canada, 2021b, para. 1).The use of 'second nature' to describe how these principles should be incorporated implies complacent inattention: when something is so ingrained it is done without attention or focus.It renders intentional actions and systemic processes invisible, and in doing so, upholds the status quo.In this environment, it would be impossible to name, much less address, systemic barriers.
Contrary to this 'second nature' approach, the Tri-Agency EDI Action Plan has been adopted by some agencies with language describing specificity and intentionality: for example, by NSERC in its grant evaluations which requires applicants to "describe specific, concrete practices that you will put into place to ensure that EDI is intentionally and proactively supporting in your project training plan" (NSERC, 2022, para.2).Further, this procedure points to NSERC's (2017) Guide for applicants: Considering equity, diversity and inclusion in your application, which provides separate and distinct definitions for each term.While there is criticism that the terms equity, diversity, and inclusion are overused and used to obscure problems (Ahmed, 2012), it is still meaningful that they are used at all.It signifies that they can be perceived; or perhaps, the use of these terms "can be understood as a form of practical knowledge of the difficulty in getting through" systemic barriers (Ahmed, 2012, p. 175).
However, in the Tri-Agency Research Data Management Policy, the words equity and inclusion are not used at all, and the only mention of diversity is in relation to methodologies and research: "The agencies acknowledge the diversity of models of scientific and scholarly inquiry that advance knowledge within and across the disciplines represented by agency mandates" (Canada, 2021a, para. 4).With the omission of equity and inclusion and a definition of diversity limited to research models, the Tri-Agency Research Data Management Policy signals that EDI as a whole is not relevant to RDM.The Tri-Agency Research Data Management Policy only goes so far as to advise a "distinctions-based approach" for research with or by First Nations, Métis, or Inuit communities and organizations (Canada, 2021a, para. 5).While it is important that the distinct data rights and sovereignty of these groups be recognized and respected as an aspect of Indigenous self-determination, the Tri-Agency falls short of fulfilling its EDI Action Plan goal in Tri-Agency Research Data Management Policy with the exclusion of other marginalized groups.It does not encourage institutions or researchers to consider how people from marginalized communities or with intersectional identities may face systemic barriers in managing their research data.Equity, diversity, and inclusion is not a box that can be checked off with the inclusion of some communities and the exclusion of others.In the Tri-Agency Research Data Management Policy, the absence of the words equity and inclusion, and use of diversity only in the context of research practices, signifies to researchers and institutions that EDI principles have a specific and limited application in RDM.

THE LITERATURE ON RDM INSTITUTIONAL STRATEGIES:
WHO AND HOW Examining the RDM literature retrieved in LISTA, two relevant groupings emerged that illuminate the RDM strategy development landscape: articles focused on the stakeholders or participants in RDM strategy development, or the who (Andrikopoulou et al., 2022;Cox et al., 2016;Cox et al., 2017;Faniel & Connaway, 2018;Howie & Kara, 2022;Patterton et al., 2018;Perrier et al., 2018;Pinfield et al., 2014;Verbaan & Cox, 2014), and those on the processes or steps involved in creating and enacting strategies, the how (Castle, 2019;Cox et al., 2016;Cox et al, 2017;Donner, 2023).This first grouping of articles examine the who of RDM, delving into which institutional units are involved in RDM strategy and service development (Andrikopoulou et al., 2022;Cox et al., 2016;Cox et al., 2017;Faniel & Connaway, 2018;Howie & Kara, 2022;Patterton et al., 2018;Perrier et al., 2018;Pinfield et al., 2014;Verbaan & Cox, 2014).Since RDM does not sit within an established portfolio, many institutional units can stake a claim on this space (Perrier et al., 2018).Collaborations between units within the institution are essential to the development of RDM services (Cox et al., 2017).The library, IT services, researchers and research support office were frequently listed as collaborators (Andrikopoulou et al., 2022;Cox et al., 2017;Verbaan & Cox, 2014), but Pinfield et al. (2014) also identified records management and legal counsel as potential partners and Patterton et al. (2018) included research ethics committees.These occupational subcultures all hold different views on how to respond to RDM needs, which can lead to conflict (Verbaan & Cox, 2014).Howie and Kara (2022) identified that the library's involvement in RDM has been primarily in the "advisory space" rather than "technical and infrastructure space" (p.11).Cox et al. (2017) noted that while IT services is most frequently mentioned as the library's partner, "issues of prioritization and coordination" in the IT-library relationship impede "[t]he development of coordinated, cohesive and integrated services" (p.2188).In addition to these differing occupational views, the ambiguity around roles and leadership has been further identified as an issue in developing RDM services (Andrikopoulou et al., 2022;Perrier et al., 2018); with Andrikopoulou et al. (2022) expanding that this uncertainty can lead to competition rather than collaboration between units.With all of these players in the RDM game, Cox et al. (2016) noted that organizational levels and roles also play a part in how an individual perceives and responds to RDM, contrasting the view of senior leadership with liaison librarians and library staff offering front-line support, and suggest that this disconnect may be challenging for the implementation of RDM services.Pinfield et al. (2014) also identified complex university governance structures and power dynamics between managers as key challenges for implementing RDM programs.These studies illustrate the institutional labyrinth that needs to be navigated in order to make space for RDM at institutions.
The second grouping of articles focus more specifically on the how, the processes by which institutional RDM strategies are created and enacted (Castle, 2019;Cox et al., 2016;Cox et al., 2017;Donner, 2023).The term top-down was frequently used to describe the strategy creation processes (Castle, 2019;Cox et al., 2016;Cox et al, 2017;Donner, 2023).Top-down, as characterized in this literature, is a policy-driven, administrator-led approach to creating research data management services (Castle, 2019;Cox et al., 2016).Cox et al. (2016) conducted a series of semi-structured interviews with UK librarians and revealed a common approach to creating institutional strategy: an upper-level administrator responsible for research creates and chairs a working group or task force with members from research support and IT services, and this working group reports to an institutional research committee.This highly structured, bureaucratic approach was noted by the interviewees as unsuccessful in gaining significant engagement with the academic community (Cox et al., 2016).In order to gain more researcher engagement, Castle (2019) proposed augmenting a top-down, centralized approach with bottom-up strategies such as holding direct meetings with researchers, interviewing and surveying researchers, and hosting open informational meetings with funders for researchers.In these studies, regardless of whether an institution takes a top-down or bottom-up approach to strategy creation, the library or librarian plays a key role in this process.

THE LIBRARY AS A 'NEUTRAL BODY' FOR RDM
In both the who and the how RDM literature, the library's role and participation in the development and implementation of RDM services is undisputed (Andrikopoulou et al., 2022;Castle, 2019;Cheung et al., 2022;Cox et al., 2016;Cox et al., 2017;Donner, 2023;Faniel & Connaway, 2018;Howie & Kara, 2022;Patterton et al., 2018;Perrier, 2018;Pinfield et al., 2014;Tayler & Jaffrey, 2021;Verbaan & Cox, 2014), even though it may involve staking a new jurisdiction (Verbaan and Cox, 2014) or have a negative impact on the library's identity (Andrikopoulou et al., 2022).Perrier et al. (2018), exploring the role of academic libraries in supporting RDM, note that researchers, librarians and library staff saw the library "in a positive light and … [as] a neutral body that could provide central services" (p.178).In Perrier et al.'s (2018) finding, I discerned an application of critical librarianship research (Chiu et al., 2021;Ettarh, 2018;Honma & Chu, 2018;Jordan, 2023;Roh, 2018) to research data management, specifically in providing a theoretical frame for interrogating this neutrality as justification for the library's role in RDM support and the shaping of institutional RDM work.
The use of neutrality as a library virtue is not new (Chiu et al., 2021), and offers one possible avenue for exploring why EDI is not addressed in the literature on institutional RDM strategies and services.One key theoretical lens for viewing library neutrality is vocational awe (Ettarh, 2018).Vocational awe "refers to the set of ideas, values, and assumptions librarians have about themselves and the profession that result in beliefs that libraries as institutions are inherently good and sacred, and therefore beyond critique" (Ettarh, 2018, para. 3).Thinking back to Cheung et al.'s (2022) consideration for equitable RDM infrastructure access for graduate students, libraries and librarians clearly can identify inequities when it comes to our users, yet under the blindfold of vocational awe, we do not see ourselves nor our libraries with the same critical eye.Vocational awe espouses the myth of library neutrality (Chiu et al., 2021).If a space is neutral, it is beyond critique -it carries no values that can be interrogated or dismantled.In the RDM literature, vocational awe is expressed in two ways: in its depiction of the library as institution and in the interactions between the librarian and the larger institution.
To expand on the first instance, the library institution is sometimes described as the active body in the RDM literature: in this framing, "the library" and "the researchers" are posed as participants (Andrikopoulou et al., 2022;Castle, 2019;Donner, 2023).Studies noted that it is important that "the library" be recognized and accepted as a leader and partner in RDM which can be achieved by "the library" proactively promoting these services (Andrikopoulou et al., 2022;Donner, 2023).While "the researchers" are presented as a group of individuals sharing an occupational role, the institution of "the library" subsumes the individual roles of different workers.This characterization accomplishes two things: first, it removes the possibility of critique because of the belief that libraries are inherently good and neutral (Ettarh, 2018); second, it erases the labour of library workers.The library is imbued with agency and labour; it is not the RDM librarian or library staff that is promoting and offering the services, but the library itself.
Vocational awe also plays out on an interpersonal level for librarians: in the instances where librarians are named as actors in the process of creating institutional strategies, librarians are described to navigate institutional structures and power dynamics and instigate cultural shifts (Andrikopoulou at al., 2022;Cox et al., 2016;Cox et al., 2012;Pinfield et al., 2014;Verbaan & Cox, 2014).Due to this need, the RDM librarian should not only possess technical skills but also "advocacy and liaison skills, training skills" (Pinfield et al., 2014, p.20), "effectiveness in the workplace" (Andrikopoulou at al., 2022, p. 358), "people and collaboration skills" (Cox et al., 2016, p. 13) and "strategic understanding and influencing skills" (Cox et al., 2012, Table 1).Succinctly put by Cox et al. (2016), "...it is not simply about applying skills to a problem but also persuading user groups that they are relevant" (p.11).Cox et al.'s (2016) interviews with academic librarians in the UK revealed an emphasis on niceness in developing and maintaining intra-institutional relationships: "Although the funder mandates opened up the possibility of stressing the stick over the carrot, interviewees seemed inclined to avoid punitive approaches, recognising the complexity of the challenge and the need to maintain goodwill" (p.14).In the literature's characterization of RDM work (Andrikopoulou at al., 2022;Cox et al., 2012;Cox et al., 2016;Pinfield et al., 2014;Verbaan & Cox, 2014), the RDM librarian not only needs to possess technical and political skills, but also must be perceived as nice -all in service of RDM education.This 'play nice' mentality further demonstrates how vocational awe is embodied in RDM work as librarians navigate institutional structures and relationships, informed by the belief that librarianship is a calling for the greater good (Ettarh, 2018).Ettarh (2018) expands on how vocational awe leads to librarian overwork and burnout in its positioning of librarians as allsacrificing.Faniel and Connaway's (2018) interviews of academic librarians in the US revealed that "difficulty meeting existing job responsibilities along with new RDM demands was mentioned most frequently" (p.107).It sets up a challenging if not impossible expectation for the librarian to be all capable, and this can lead to librarian burnout and overwork.
In naming the phenomenon of vocational awe, Ettarh (2018) allows library and information professionals to expose the problematic underpinnings of our institutional and interpersonal interactions.It is important to name and reveal these dynamics since "[t]he very tendency to 'look over' how everyday and institutional worlds involve restrictions and blockages is how those restrictions and blockages are reproduced" (Ahmed, 2012, p. 181).In the RDM literature, these restrictions and blockages are framed in terms of organizational level and occupational group (Andrikopoulou et al., 2022;Cox et al., 2016;Cox et al., 2017;Pinfield et al., 2014;Verbaan & Cox, 2014), concealing other structural power dynamics with the rhetoric of neutrality.The presence of vocational awe and neutrality in our institutions is not isolated or rare, but pervasive and deeply ingrained; or as Chiu et al. (2021) aptly describe, "not the shark, but the water."(p.49).This is not to say that we should passively accept vocational awe as status quo, but rather that we should be scrutinizing the institutional waters we are in and intentionally using equity, diversity, inclusion, and justice to navigate our course.

SEEING INVISIBLE LABOUR IN RDM
For racialized librarians, exacerbating overwork is the invisible labour of navigating predominantly white institutions.Although I did not find it reflected in the existing RDM literature, I found my concerns echoed in Jordan's (2023) writing about her experiences as a minority open educational resources (OER) creator.Jordan (2023) describes the context that she and other racialized individuals work within: "navigating microaggressions, explaining why we are in a room, dedicating time to advocate for ourselves, and even redirecting and rewriting our own impostor-syndrome-related thoughts, which mirror society's perception of our ability (para.13).While RDM was seen as a natural extension of my scholarly communication portfolio by my institution, like Jordan, I struggled with whether I was the right person to be doing this work, or whether I was simply there to bring a bit of colour into the room; or expressed more concisely by Hswe, as quoted in Roh (2018), "Which is it: competence or color?" (p.435).Jordan (2023) finds that these "well-meaning efforts" for diversity and inclusion "can place faculty of color at risk for burnout and exploitation" with "a staggering invisible workload" (para.18).Although Jordan writes about OER creation, the same well-meaning efforts could well apply to research data management.Recognition of diversity does not lead to equity when the performance of equity, diversity and inclusion is prized over the actual enactment of these principles, and it places what Roh (2018) describes as a "[b]urden to perform not just to prove ourselves, but on behalf of others" (p.436).The work of making research data management strategies and services diverse, equitable and inclusive should not be unduly placed on librarians and faculty from marginalized groups.
The RDM literature on intra-institutional collaborations does not address the cost of these collaborations for people with intersectional identities.It presumes institutions, libraries and librarians are homogenous because it does not address positionality.Chiu et al. (2021) explain how this presumed whiteness/neutrality is perpetuated: "By virtue of their privileged position, librarians of the status quo can afford to remain neutral toward issues that don't affect them personally.The legacy of a homogenous profession has given power to this position.Oppression is seemingly overlooked for its lack of impact, whether direct or indirect, not only by the individuals who make up the profession, but also by the profession itself."(p.56).As with the absence of equity, diversity, and inclusion from the Tri-Agency Research Data Management Policy, the lack of attention on equity, diversity, and inclusion in the research literature presumes that these principles are not important or applicable to the people doing the RDM work at institutions, and therefore, there is no cost to collaborations because these systemic barriers simply do not exist in a homogenous white world.

A PROPOSAL: CONTRIBUTIVE RATHER THAN COLLABORATIVE
Collaboration can be fraught when navigating power structures in a predominantly white institution.In my experience as a racialized person, collaboration in a predominantly white institution can mean white administrators taking my ideas and presenting them as collaboratively generated, or worse, misunderstanding or misrepresenting my words to cause harm to other marginalized individuals.Tokenism can be at play, where the "very fact of your existence can allow others not to turn up" (Ahmed, 2012, p. 5).Models that tout collaboration as an unquestioned good should be interrogated: who is participating and whose voices are being heard?What has each person contributed?Honma and Chu (2018) suggest that the contributive justice model proposed by Paul Gomberg (2016) provides a way forward for equitable participation in libraries: "Contributive justice is the process whereby each of us has the opportunity to contribute.By everyone contributing, then there will be the opportunity to ensure that all of our stories, all that we can offer, will be accessible to others.With all contributing, there is greater possibility of accessibility of a broader range of information" (Honma & Chu, 2018, p. 460).This contributive justice model offered me a way to enact principles of equity, diversity, and inclusion in the creation of an institutional research data management strategy.Contributive justice prizes diversity by including a range of voices and stories.It creates equity in providing flexibility in how people can contribute, and ultimately supports creating a culture of inclusion where peoples' individual contributions are valued.
In the spirit of contributive justice, and aware of the reality that RDM work is messy and often unaligned with best practices, I determined that it would be best to first understand how researchers at my institution manage their research data rather than start with drafting a strategy based on best practice and the research literature and only then soliciting research feedback.RDM librarians frequently use surveys to gather data on researchers' data practices and concerns (Castle, 2019;Cheung et al., 2022;Cox et al., 2016;Patterton et al., 2018); however, surveys have been noted in the literature to have limited response from researchers (Cox et al., 2016).Rather than creating a survey, I began with talking to researchers and listening to their diverse stories, experiences, worries, and feelings.I was aware that these researcher conversations would take time, but this approach heeded Okun's (2023) warning that a sense of urgency to complete the strategy is characteristic of White Supremacy culture, and that when a sense of urgency arises, one should build time for equity.With support from the research office and liaison librarians in identifying researchers, I contacted staff and faculty researchers at my institution for short chats about their research data practices.Some researchers chose to respond via email, but the majority (fifty researchers) agreed to a virtual or in-person meeting.Due to the deadline provided by the Tri-Agency policy, faculty availability during the academic year, and my own workload, most of these meetings took place in the last three weeks of a busy fall semester.Ahead of time, I prepared some questions focused on researchers' RDM practices and current and anticipated needs; however, these were only to be used to prompt conversation and not to fully direct the chat.This semi-structured interview approach allowed me to approach these conversations with RDM in mind; but not restrict our conversations to a specific set of questions or topics.During these chats, these researchers generously and openly spoke about their experiences and needs for research data management support, framed in their specific stories and positionalities.Many of their concerns aligned with researcher themes identified in the research literature, but in having more unstructured conversations rather than sticking to a prescribed script or set of survey questions, we uncovered concerns around EDI, workload, justice, and institutional politics that were not included in my list of initial prompts and questions.When it came time to circulate the policy draft, I received comments from researchers outside of the initial group of fifty researchers I spoke with, unlike the low researcher engagement resulting from the top-down approaches as described in the literature (Cox et al., 2016).I felt this spoke to the importance of building space at the start of the strategy creation process to get to know the researchers and resisting the sense of urgency to draft the strategy.
In addition to researchers, I also sought feedback on the working draft from administrators at the vice president, associate vice president and dean levels and employees in IT services, library, research ethics board, teaching and learning centre, and Indigenous faculty group, a list of contributors reflective of those described in the literature (Andrikopoulou et al., 2022;Cox et al., 2016;Cox et al., 2017;Faniel & Connaway, 2018;Howie & Kara, 2022;Patterton et al., 2018;Perrier et al., 2018;Pinfield et al., 2014;Verbaan & Cox, 2014).The RDM literature poses role ambiguity as a challenge for intra-institutional cooperation (Andrikopoulou et al., 2022;Perrier et al., 2018;Pinfield et al., 2014) but under "contributive justice, where everyone has the ethic of contributing, we all have a role to play, and everyone takes ownership in doing their part.Those who can contribute more will do so, while those may contribute less, but all contribute to the shared labor."(Honma & Chu, 2018, p. 460).All of these people provided contributions to the strategy that were informed not only by their occupational and organizational roles, but also their positionality and experiences at this institution and others.For example, the labour concerns raised by researchers led me to request the involvement of the faculty association, which had not been identified in the literature as a contributor to institutional RDM strategies.The faculty association provided a larger scale view of faculty and their expertise was invaluable in ensuring that union rights and the principle of academic freedom were explicitly stated in the strategy.Through the lens of contributive justice, I was not evaluating the quantity or quality of people's contributions, but simply "respected [them] for contributions that are real and important" (Gomberg, 2016, p. 50).The recognition of contribution as a reciprocal benefit (Gomberg, 2016) is essential in the context of my teaching-focused, rather than research-focused, institution.Since these researchers are doing this additional work on the side of their recognized, paid roles, it is important that people are invited to contribute: while not every individual will care about RDM, an invitation to participate acknowledges the legitimacy of each researcher's work and recognizes that each researcher has the right to contribute to the strategy.It is an equity-creating step that also acknowledges the work done by people to perform or support these research activities, especially those who are not employed in faculty roles or those in "non-instructional" faculty roles such as librarians and counsellors.I felt that this open, iterative methodology reduced tunnel vision by highlighting issues that at times seemed to be out of scope or peripheral but together formed a big picture of intertwining systems and structures that inform how people work together at my institution and how they might manage their research data.
The contributive model strengthens relationships in that individuals' voices and experiences are acknowledged and valued.Centering people and equity, diversity, and inclusion enriches an institution's research data management strategy because it prioritizes the process over the product.The Tri-Agency policy requires institutions to outline how institutions will support researchers, rather than what they will do (Canada, 2023b), which makes space for strategy creators to focus on the process.Although these reflections have primarily focused on the labour of interpersonal and intra-institutional relationships, the written strategy reflected the spirit of the process The strategy is informed by diversity, equity, inclusion, and justice, and reflects the individuals who engage in research at my institution.It establishes a foundation for doing research data management work at the institutional level that is EDI-and person-centered.

THE FUTURE
"How documents are written affects how they might be taken up.If the document becomes the sole responsibility of an individual within an organization, then that organization can authorize the document, give it a signature, and refuse responsibility for it at the same time" (Ahmed, 2012, p.91).
Writing and publishing the strategy is just the start of the work.It complies with the requirement set by the federal policy, but does not ensure my institution's commitment to taking substantive action to create RDM services rooted in equity and justice.A single document does not instantly dismantle systemic barriers nor does it ensure non-performative accountability to the principles outlined.
However, using a contributive model to create a strategy does give voice to researchers and also amplifies the voices of librarians and other professionals in the research process whose labour and key contributions to research are often sidelined.It values the diversity of the people involved and makes a mindful effort to enable their participation and recognize their contributions.Throughout the process and in the strategy document, faculty and staff researchers' needs, concerns and successes inform the institutional priorities, and their voices make the strategy human.The recognition of strategy creation as a people-centred practice contributes to a culture of inclusion where people feel valued and respected.This strategy creation process reflects Ahmed's (2012) suggestion to "think of strategy not as the absence or bracketing of thought…but as the unfolding of thought" (p.5).Alongside the researchers' stories, the concerns of vocational awe (Ettarh, 2018) and invisible labour (Jordan, 2023), as highlighted in critical librarianship research (Chiu et al., 2021;Ettarh, 2018;Honma & Chu, 2018;Jordan, 2023;Roh, 2018), contributed to the strategy's unfolding.The resulting strategy is a dynamic and imperfect document, not a static best practice that can be uniformly applied across institutions.By treating equity, diversity, inclusion and justice as central to our work, LIS professionals can challenge and shape our practices and the ways we talk about RDM, and consider alternative ways for us to do institutional RDM work are centered on people and their contributions.Because, ultimately, when we tell the stories of research data management, we are telling the stories of people, not data nor organizational units.
(Smith, 2021), 2022)ion in the Tri-Agency Research Data Management Policy, I turned my attention to the academic literature on RDM.I conducted searches on the Library, Information Science & Technology Abstracts (LISTA) database to discover what ways the existing academic literature approached RDM and EDI, and in particular how RDM institutional strategies were created and enacted with these principles in mind.Searches for the keywords research data management, RDM, equity, diversity, inclusion, EDI, justice, and labour/labor retrieved results on institutional RDM strategies and services, but no articles focused primarily on RDM strategy development and the principles of equity, diversity, and inclusion.However, there was one article on RDM service this understanding of diversity does not seem to apply beyond research practices and data to the researchers' identities, those of their research communities and partners, nor the people supporting research data management at an institution.Although the scope of this article is institutional RDM strategy creation and EDI, it is important to acknowledge that there is a vibrant and growing body of literature on Indigenous research data management services at universities(Howie & Kara, 2022), situated in decolonizing methodologies(Smith, 2021), as well as international and national principles created by and with Indigenous peoples, such as the CARE Principles for Indigenous Data Governance (Global Indigenous Data Alliance, 2019) and the First Nations Principles of OCAPⓇ (First Nations Information Governance Centre, 1998) that can inform institutional RDM practices.Aside from the literature on Indigenous research data management and Cheung et al.'s (2022) consideration of equitable access, the LISTA search largely did not uncover academic research that explores equity, diversity, and inclusion for other groups, nor how these principles are applied in the creation of institutional RDM strategies.
(Castle, 2019;Pinfield et al., 2014;of EDI principles, even though it did not use the terms equity, diversity and inclusion: inCheung et al.'s (2022)survey gauging researchers' RDM readiness at Canadian universities, they noted that their survey results "results may reflect differences in access to institutional infrastructure, with fewer graduate students having access to shared drives and/or university or departmental servers" (p.12) and advised institutions to ensure that graduate students have access to institutional infrastructure.To parseCheung et al.'s (2022)finding with an EDI lens, they identified that graduate students have different circumstances than faculty researchers (diversity), and recommended institutions to enable this access to institutional infrastructure (equity), which builds towards a culture of inclusion in institutional RDM support.The other, direct mentions of diversity in the LISTA literature on RDM were descriptions of disciplinary 'diversity' in how research data is managed(Castle, 2019;Pinfield et al., 2014; Tayler  & Jaffrey, 2021).Like the use of 'diversity' in the Tri-Agency Research Data Management Policy,