Using a Behavioural Framework to Understand and Modify Researchers’ Data Management Practices

Ms Kylie Poulton1, Ms Sue Hickson2, Ms Maria Connor3

1Griffith University, Nathan, Australia,

2Griffith University, Gold Coast, Australia,

3Griffith University, Gold Coast, Australia

 

INTRODUCTION

The release of the Productivity Commission’s Inquiry Report into Data Availability and Use [1], may herald a new era of data eminence in Australia.  While the major funding bodies, the ARC and NHMRC, have policies that strongly encourage researchers to deposit their research data in an appropriate repository, the Commission’s recommendations could see a strengthening and tightening of this policy.

With stronger mandates in place researchers may be compelled to, at the very least, formally plan how they will manage their data and may possibly be required to make their data accessible to others.  Librarians and data managers are in a unique position to enable researchers’ to transition to new data management requirements.

UNDERSTANDING DATA MANAGEMENT BEHAVIOUR

This presentation will report on the second phase of a unique research project that investigated if librarians and data managers can modify researchers’ data management practices by understanding their practices through a conceptual framework, the A-COM-B model.

In the first phase of the project the research team investigated the data management behaviours of members of a high-profile, interdisciplinary, social science research group with the objective of encouraging them to use the institutions data management solutions.  The project analysed data management behaviours using the A-COM-B framework.  Wolski and Richardson [2] developed the A-COM-B framework, based on the COM-B model developed by Michie et al. [3].  As illustrated in Figure 1, in the A-COM-B model, A = Attitude; C = Capability; O = Opportunity; M = Motivation; and B = Behaviour.  The project aimed to find out whether understanding data management behaviours through the A-COM-B lens could be used to plan intervention strategies.

 

Figure1: A-COM-B framework for understanding behaviour.

The initial findings of the research project were reported in Hickson et al. [4]  The authors concluded, “By using this framework, practitioners can design intervention strategies that are aligned to individual need, and that lead researchers to using safe and secure institutional solutions and services.” [4]

 

MODIFYING DATA MANAGEMENT PRACTICES

In this presentation the authors will discuss their response to the initial research findings and report on Phase 2 of the research project, specifically:

  1. The results of a follow up survey to discover whether researchers’ have modified their data management behaviour in response to interaction with librarians.
  2. The development of best practice guidelines tailored for the research centre based on the initial research findings.
  3. The development of a toolkit based on the initial research findings and framed around the A-COM-B model that can be used by librarians and data managers to inform their engagement with researchers around data management.
  4. The development of a training program for librarians and data managers focusing on researcher engagement using the A-COM-B framework.

This presentation will be of interest to librarians, data managers, project managers and researchers and would be suited for “The Connected Researcher” stream.

REFERENCES

  1. Productivity Commission, Australian Government, Data availability and use: Productivity Commission Inquiry Report Overview and Recommendations, Productivity Commission, Editor. 2017.
  2. Wolski, M. and J. Richardson. Improving data management practices of researchers by using a behavioural framework. in THETA 2015 Create, Connect, Consume. 2015. Gold Coast, Queensland, Australia: CAUDIT.
  3. Michie, S., M.M. van Stralen, and R. West, The behaviour change wheel: A new method for characterising and designing behaviour change interventions. Implementation Science, 2011. 6(42): p. 1-12.
  4. Hickson, S., et al., Modifying researchers’ data management practices: A behavioural framework for library practitioners. IFLA Journal, 2016. 42(4): p. 253-265.

 


Biographies

Kylie Poulton began her library career as a researcher in investment banking before joining Information Services at Griffith University.  During her career at Griffith, she has held roles in Liaison Librarianship, Project Management and Change Management.  She has also led Griffith’s Higher Education Research Data Collection team and worked as a project officer in the University’s Research Office.  She is currently Griffith University’s Business Librarian. Her research interests include data management and text and data mining.

Sue Hickson is a Library Services Manager (Business) at Griffith University and a librarian by profession. Sue is responsible for leading a team of Librarians, Learning Advisors and Digital Capability Advisers who support academics, researchers and students in the areas of Research, Learning and Teaching. Previously Sue was a Faculty Librarian supporting Griffith’s Health Group.

Maria Connor is the Business Librarian at Griffith University Gold Coast campus. She holds an MLIS from Victoria University, Wellington, New Zealand.

Research tools: Why Institutions Should Care

Mr Malcolm Wolski1, Ms Louise Howard1

1Griffith University, Brisbane/Gold Coast, Australia

 

SUMMARY

Tools are critical to undertake successful research. Researchers are increasingly utilising software tools as an integral part of the research lifecycle to manage, process, analyse and integrate data from multiple sources.

This increasing use of tools has implications for not only researchers but also the institutions where the researchers are located. Support for research tools by institutions has high-level impacts, ranging from financial, strategic, and compliance through to capacity, capability, and connectivity.

This presentation outlines the ramifications for institutions, particularly universities, in supporting the increasingly complex tools which are used in the research lifecycle. It also explores the use of a trust framework to address this challenge.

 

ABSTRACT

The eResearch environment in Australia has changed significantly over the last decade. The first investment wave saw substantial resources channelled into developing research infrastructure, e.g. servers, storage, and high performance computing (HPC). However, there has been growing recognition that “Research outputs, whether data, software, methods or publications, are critical inputs to future research and underpin innovation” (O’Brien, 2016).

In more recent times much attention has been given to supporting researchers with their data management requirements. Less attention has been given to the tools that researchers use that are critical to successful research. And yet support for research tools by institutions has high-level implications, ranging from financial, strategic, and compliance through to capacity, capability, and connectivity. The current wave of cybersecurity incidents has also highlighted institutional exposure to the development and use of such tools.

Kramer and Bosman have extensively researched the use of tools by researchers. Of the more than twenty thousand responses to their 2016 online survey, researchers accounted for seventy-two percent and librarians for just seven percent. The average number of tools reported per person was twenty-two. The current list of tools exceeds six hundred.

How, therefore, do researchers determine which tools they should use? When they do find them, how can they assess the risks and determine the credentials for use of the tools?

This presentation will discuss reasons why institutions should take an interest in tools their researchers are using and also explore the application of a trust framework to assess tools. This framework can be used as a guide by support and outreach staff to provide advice to the research community.


Biographies

Louise Howard is is an information professional with over 20 years’ experience in information management and governance in the public sector, including local and state government and higher education. She is currently part of the senior leadership team providing library, information and IT services at Griffith University to staff, students and researchers, with particular responsibility for a comprehensive range of  information technology infrastructure services. Her current focus is on continuing to evolve Griffith’s IT services and practices to ensure agility and flexibility in service design and delivery.

Malcolm Wolski is the Director, eResearch Services, at Griffith University. In his role, he is responsible for the development, management and delivery of eResearch services to support the University’s research community, which includes the associated information management systems, infrastructure provision, data management services, and media production. Malcolm is a part of the senior leadership team providing library, information and IT services at the University, and he works closely with these groups to provide service desk, infrastructure and outreach services to the research community.

Leveraging projects for institution-wide benefit – Expect the best, plan for the worst, and prepare to be surprised

Dr Andrew Williams1, Ms Eva Fisch1Simon Huggard1

1La Trobe University, Bundoora, Australia

 

It was only comparatively recently that the value of researchers’ datasets and the need for those datasets to be organized and discoverable has been widely recognized. At La Trobe University, we are addressing these challenges through key projects and partnerships with academics, ICT, and the Australian National Data Service (ANDS) to develop a new technical infrastructure and to generate cultural change. The key components of our scheme to date include:

The University Digital Research Strategy, which philosophically underpins these plans through its four main goals:

  • Enable researchers to actively engage with information and computing technologies
  • Develop Digital Research expertise among researchers within the University
  • Proactively enhance institutional research capabilities
  • Maximise the use and enable proper management and curation of research data

Sustainable and successful ANDS (Australian National Data Service) High Value Collections Program projects:

In a recent ANDS-funded project, HubZero has been implemented to facilitate publication of and interaction with x-ray photoelectron spectroscopy (XPS) research data by a surface science research community. The imperative of ANDS funding has also given La Trobe the opportunity to explore the complementary potential of two genomics browsers. The Degust browser was developed by the Victorian Life Sciences Computation Centre and facilitates discovery of known genes using known gene names. The eFP browser was originally developed at the University of Toronto and is suited to analysing expression of known genes. The resulting publication and visualisation platform using these tools is now offered as part of La Trobe genomics platforms suite of services.  In this presentation we will discuss how the partnership in these projects worked to:

  • implement a custom metadata schema and custom data visualisation widgets
  • engage with industry – influencing the direction of instrument development through data publication
  • engage with the research discipline through publishing a large reference dataset
  • amplify the researcher’s voice in a technical development project
  • strengthen the role of the data librarian in a project enabling researcher engagement with the community
  • leverage boutique projects for institution-wide benefit.

The University ERDMS (Enterprise research data management system) project:

The Library and ICT are working together in a three year project to implement an Enterprise Research Data Management System, developing and building the University’s capacity to collect, curate and disseminate research data underpinning research outputs.

The open platforms used will enable the University to better manage research data outputs, in line with its obligations to under the Australian Code for the Responsible Conduct of Research, in line with our research data management policy, as well as anticipated future requirements of funding agencies to require researchers to provide a research data management plan and to publish research datasets in the public domain. Software being considered, trialed, or put in place include:

  • research data management plans ( ReDBox)
  • CloudSTOR
  • Hubzero
  • OMEKA
  • eMU
  • Lab Archives
  • eFP and Degust browsers

The system will also enable researchers to improve the management and tracking of their own research, collaborate more easily with internal and external research partners, and provide a safe, backed up and fully managed environment in which to manage their data.  A concomitant cultural change in the University is also required and will be supported.

These projects together are testing, piloting and implementing a suite of systems, software, services and tools to streamline and simplify researchers’ data management needs.

 


Biographies

Simon Huggard is Deputy Director, Research & Collections at La Trobe University Library. He has worked at a number of other institutions including Monash University and the State Library of Victoria. He is responsible for two major portfolios in the Library: Research and Collections. The Research team provide services to researchers to help use electronic resources , databases and print collections, as well as providing research impact reports, advice on open access, publication through the institutional repository and assistance with annual publication collection and research data management. The collections team manage La Trobe University library’s search and management systems, reading list software, acquisition and management of all of our resources and collections and interlibrary loans. I am interested in how others provide maximum benefit for students, staff and all of our users in accessing resources and managing research data in an effective way.

The Research Data Lifecycle: One institution, three experiences

Professor Debra J. Searles1,2, Jill Penridge2, Helen Morgan3

1Centre for Theoretical and Computational Molecular Science, Australian Institute for Bioengineering And Nanotechnology, The University of Queensland, Brisbane, Australia,

2The Australian Institute for Bioengineering and Nanotechnology, The University of Queensland, Brisbane, Australia,

3Learning and Research Services, The University of Queensland , Brisbane, Australia

 

INTRODUCTION

“It is not sufficient simply to focus on exposing, collecting, storing, and sharing data in the raw. It is what you do with it (and when) that counts” JISC report – The future of data-driven decision making

People often talk about the research data lifecycle from the perspective of the data that flows through it, in a clean and linear way… planning, collecting, analysing, publishing… is the journey the data goes on. Neglected are the stories from the people involved in making this work. These stories are often compelling and full of challenges and controversy. To get data to flow neatly as it is curated through the data lifecycle requires hard work and persistence on the part of many different stakeholders. In this paper, we hear from three of the university stakeholders who are very much at the forefront of research and who are, each in their own way, driving towards best-practice research data management.

Modern research faces existential dilemmas, key among them the growing lack of confidence in both the outcomes and the need for fundamental research as a societally valuable activity deserving of resources. The causes of this increasing uncertainty are being simultaneously driven by scandals around reproducibility, the commodification of research and the politics of science denialism. These pressures come at the same time as a significant generational shift in the research and research support workforces.

THE RESEARCHER PERSPECTIVE

For most researchers, the challenges associated with data are growing as the capabilities to generate and store it grow, the pressure to publish high research output rapidly increases and formal requirements associated with data storage, reproducibility and integrity are enforced. Certainly some research is not possible without fast storage and access to huge volumes of data, and for researchers in high performance computing this is vital. However, availability of resources also makes it easier to store data that is not required and ultimately more difficult to locate appropriate data when required.

Many new resources are becoming available which assist with data management, storage and availability, however as with any new technology, uptake can be slow due to the need to change processes and culture within a research group and the history of technologies, software systems and storage systems becoming outdated. Data carefully filed and stored on Zip drives and other outdated technologies become almost useless. Challenges associated with developing data storage processes within a research group that are practical, versatile, enduring and suitable need to be overcome, and this will be discussed.

As a researcher in high performance computing, digital data storage has always been a key issue.  Today, this is becoming important to the broader research community that brings with it associated opportunities that will also be highlighted.

THE TRANSLATIONAL RESEARCH FOCUSSED INSTITUTE PERSPECTIVE

Data is the lifeblood of the basic research from which scientific discoveries are generated. In translational research, the reliability of data is equally or more important as the justification for the commercial value ascribed by commercial sponsors in exchange for their financial and reputational support of that research.

At an operational level, the data-related challenges to be addressed in the current research environment are varied. The existing cohort of senior academics should be encouraged to recognize the practicalities of data as a resource that needs to be sustainably managed and preserved to ensure business continuity risks are addressed. Incentives should be introduced to motivate recognition of the veracity and providence of data as the underpinning source of value, and therefore resources, to industry.

In particular, translational research also requires pragmatic navigation of pre-“big data” university and government policies that are influenced by entrenched institutional views of how research was undertaken in the past. One final area of potential focus is the need for increased IT literacy among the senior executive of universities, who may not have recent experience with IT in a research context, to improve institutional decision-making.

Solutions will be proposed to address these challenges but ultimately the necessary changes must be supported by effective cultural support at all levels of modern universities to have a positive impact.

THE WHOLE-OF-INSTITUTION PERSPECTIVE

Institutions face challenges in this space that are driven by a rapidly moving national and global landscape. Funders and Journals are increasing requirements to provide research data on publication or completion of project work.
This requires consistency with local institutional policy, and continued development of governance and systems to meet varying needs across the institution.  Increasing ease of compliance for researchers without policing or administrative burden is requiring the development of new, novel workflows. These have to be designed and implemented in consultation with both researchers and operations managers.

Institutions have much to be gained by building reputations as trusted providers of data. The current swing towards innovation and emphasis on collaborating with industry to solve real-world problems means now more than ever data management is a priority activity. The ability to validate, verify and reproduce research outputs, and to have a clear audit trail from published results to the underlying raw data is a clear advantage for any organization.

CONCLUSION – GOAL of RESEARCH INTEGRITY, BEST-PRACTICE AT ALL LEVELS, TRUSTED DATA and RESEARCH OUTPUTS

Good data management throughout the data lifecycle requires significant engagement with the processes, systems and people at all levels. How this looks can be quite different and present different challenges to everyone involved. Ultimately good curation of data happens when there is a strong culture and shared understanding by researchers of the value created.

The ways in which these three perspectives intersect will be discussed and common points of friction highlighted. The practical implications will be considered through the lens of researchers as the effectiveness of their activities is the primary source of research success.

About the conference

eResearch Australasia provides opportunities for delegates to engage, connect, and share their ideas and exemplars concerning new information centric research capabilities, and how information and communication technologies help researchers to collaborate, collect, manage, share, process, analyse, store, find, understand and re-use information.

Conference Managers

Please contact the team at Conference Design with any questions regarding the conference.

© 2017 - 2018 Conference Design Pty Ltd