1 of 47

“Open wide: Open Research Training across the disciplines”

2 of 47

Abstract

What I promised:

The UK Reproducibility Network (UKRN) is a national peer-led consortium that aims to ensure the UK retains its place as a centre for world-leading research. The Open Research Programme is a five-year programme supported by Research England, UKRN institutional members and project partners which aims to raise the standard of open research practices across the UK research ecosystem. I will talk about the first of three programme work packages: Training. https://www.ukrn.org/open-research-programme/

What I’ll also speak to:

Responses to open research from different disciplines

3 of 47

UKRN & ORP

4 of 47

“..in this field of behavioural science, there are a lot of charlatans...the whole field is riddled with duff studies and memes that people believe are true but are not true.”

5 of 47

UKRN

“a national peer-led consortium that aims to ensure the UK retains its place as a centre for world-leading research”

Image: UKRN institutional membership 2022,Adam Kenny

6 of 47

7 of 47

5 year, £8.5m Project

3 workstreams:

  1. Delivering training
  2. Developing a framework for evaluation
  3. Sharing effective practice

Adam is Open Research Coordinator and Administrator (ORCA) for University of Sheffield. Every institutional partner should have an ORCA!

Adam Partridge,

Open Research Training Lead,

Research Services

8 of 47

Open Research Programme scope

“to accelerate the uptake of high-quality open research practices”

  • Research co-production
  • Conducting open research consistent with relevant legal, ethical, and regulatory constraints
  • Transparent Qualitative Data Analysis
  • Defining the data, code, or other evidence on which your research findings will be based and how this will be managed and shared before the start of data collection and analysis
  • Pre-registration of research protocols
  • Using open source software created by others
  • Creating your own open source software/analysis code to share with others
  • Version control of research products
  • Your own data analysis is computationally reproducible
  • Sharing your own data, code, or other evidence according to the FAIR principles
  • Guidelines for recognising the specific substantive contribution of everyone involved in research projects
  • Declaring conflicts of interest
  • Publishing pre-prints
  • Ensuring accepted publications (and other final outputs of research) are Open Access

9 of 47

Year 1 report : Jim Grange & Hannah De’Bell

10 of 47

Findings

- There is a large diversity in the amount of open research training resources available between different UKRN member institutions.

- Support for different aspects of open research practice is also variable, with the “planning” and “analysing” stages of the research lifecycle better supported than the “conducting” or “disseminating” stages.

- The institutions surveyed perceived the majority of open research practice areas they were asked about as valuable, and, where they were available, training resources on open research were perceived to be high quality.

- Only 10.5% of total resources are open-access, raising the costs for each individual institution to create training resources.

- Training provision is usually delivered “in house” by research support or library services.

11 of 47

Components

Schema (Sarah Callaghan)

Endorsement (Joe Cornelli, David Foxcroft)

Accreditation (Steve Boneham)

Delivery (Adam Partridge)

12 of 47

Schema (Sarah Callaghan)

13 of 47

Endorsement (Joe Cornelli & David Foxcroft)

Evaluation, feedback and a commitment to act on it

14 of 47

Accreditation (Steve Boneham)

Badges, “agreed expectations”, communities of practice

15 of 47

Delivery (Adam Partridge)

External partners: Including… The Software Sustainability Institute (UK), VIRT2UE (Netherlands) and Griffith University (Australia), TIER, COS

And supporting “home grown” training and use of open resources from other institutions.

Data management

16 of 47

Dependencies / vision

The year is 2026, the Open Research Programme is coming to the end of its initial funding. The programme represents a gold standard for training in open research practices, with a model which is internationally admired and emulated. Leaders at UKRN member institutions use the UKRN schema to identify training areas the programme can address and are confident of the standards that the programme holds for its training. External training providers compete to be able to say they are official partners. Within institutions, local contacts - ORCAs - know they can rely on the programme to facilitate training for staff and value the network/community of practice which the UKRN programme provides for them and staff responsible for training in their institution. Staff involved in training within institutions are proud of the accreditation offered by the programme and are able to use it for career progression. The programme fosters locally excellent practice in open research training, with groups who offer the most comprehensive training on particular topics joining the programme to train trainers at other UKRN institutions. The programme records hundreds of trainers who have taken part in the programme and they have reported, in turn, thousands of researchers who have been trained in open research practices. The UKRN and UKRN partners are able to share best-practice training resources across all areas of open research (both via public facing platforms like the UKRN’s living website and informally via community of practice forums). The programme supports the creation of new, openly licensed materials, in areas where there is a gap. Across the UK, standards for open research are elevated, driving greater consistency in research practice, wider adoption of better standards and greater engagement with developing and defining best practice, which continues to evolve and be reflected in the trainings provided by the programme.

17 of 47

Aims

Support institutional, academic led training on open research

Generate materials

Raise engagement with open research across UK

Work to develop best/better practice

Reach underrepresented disciplines in open research

18 of 47

Criticisms of Open Research

19 of 47

TUoS Open Research Statement

“The University of Sheffield recognises that Open Research is a vital part of our commitment to enabling and promoting research excellence across our community.”

sheffield.ac.uk/openresearch

20 of 47

Image: University of Sheffield

21 of 47

Extra effort

personally costly?

unnecessary?

distracting?

Image: Sisyphus, 1548 by Titian via titian.org

22 of 47

Data positivism

Also, confidentiality risks

Image: AI generated, via stablediffusionweb.com.

Prompt: “a malevolent tower overlooks a barren wasteland”

23 of 47

Exclusive and performative

Alex McLean ‘Performing openness in academic publishing’ https://slab.org/2022/12/08/performing-openness/

Image: AI generated, via stablediffusionweb.com.

Prompt: “smug scientists sit in preening superiority on thrones of diamond”

24 of 47

Privilege entrenching

Image: AI generated, via stablediffusionweb.com.

Prompt: “slaves building the pyramids in impressionist style”

25 of 47

Smith, A. C., Merz, L., Borden, J. B., Gulick, C. K., Kshirsagar, A. R., & Bruna, E. M. (2022). Assessing the effect of article processing charges on the geographic diversity of authors using Elsevier’s “Mirror Journal” system. Quantitative Science Studies, 2(4), 1123-1143.

APC (“Gold”)

APC (hybrid)

free (hybrid)

Mirror journals with OA/APC and closed/no APC routes shows more authors from low and middle income countries published by non-OA route

"Our results for Elsevier’s Mirror-Parent system are consistent with the hypothesis that APCs are a barrier to OA publication for scientists from the Global South."

26 of 47

Olejniczak, A. J., & Wilson, M. J. (2020). Who’s writing open access (OA) articles? Characteristics of OA authors at Ph. D.-granting institutions in the United States. Quantitative science studies, 1(4), 1429-1450.

Across 182,320 US researchers, publishing APC-funded OA articles was associated with being a man, at a prestigious institution, being in STEM, having more funding, and at a more senior career stage.

"Participation in APC OA publishing appears to be skewed toward scholars with greater access to resources and job security."

Image CC BY-SA 3.0, Wikimedia commons

27 of 47

More on critiques of open research

Ross-Hellauer, T. (2022). Open science, done wrong, will compound inequities. Nature, 603(7901), 363-363.

Lancaster, A. (2016). Open science and its discontents. Ronin Institute.

Naomi Truan (2021). Dealing With Being Exposed: Setting Boundaries While Being Open

Mirowski, P. (2018). The future (s) of open science. Social studies of science, 48(2), 171-203.

Callard F. (2022). Replication and reproduction: Crises in psychology and academic labour. Review of General Psychology, 26, 199–211.

Uygun-Tunç, D., Tunç, M. N., & Eper, B. (2022). Is Open Science Neoliberal?. Perspectives on Psychological Science, 20(X). https://doi.org/10.1177/17456916221114835

Levin, N., & Leonelli, S. (2017). How does one “open” science? Questions of value in biological research. Science, Technology, & Human Values, 42(2), 280-305.

Peterson D., Panofsky A. (2021). Arguments against efficiency in science. Social Science Information, 60(3), 350–355. https://doi.org/10.1177/05390184211021383

28 of 47

Synthesis

29 of 47

Process quality > output quality

Reproducibility …?

More repeatable

More reusable, extendable

More translatable

Better documented

Better archived

More secure

30 of 47

Principles vs practices, an example

PRACTICE: preregistration

publicly archived trial design, outcome measures and statistical models

EXCEPTIONS

exploratory research, feasibility/pilot studies,

PRINCIPLE: avoid (self) deception in results narration

formal record of what you intended when you started

alertness to researcher bias

31 of 47

Principles vs practices, example II

PRACTICE: open data

public, permanent, archive of study data

EXCEPTIONS:

commercially or clinically sensitive data, DOUS (data of unusual size), lack of consent

PRINCIPLE: reproducibility, audit, reuse

share post-processed data at earlier possible point

share metadata for discovery, synthesis

32 of 47

Onwards

33 of 47

“Training researchers in research integrity and the need to ensure reproducibility is inconsistent and often absent. We recommend mandating the provision of such training at undergraduate, postgraduate and early career researcher stages.”

https://publications.parliament.uk/pa/cm5803/cmselect/cmsctech/101/summary.html

34 of 47

EVENT: 2023-05-18

35 of 47

Teaching Reproducible Research and Open Science’, 20-22 June 2023, Sheffield.

Dr Aneta Piekut

Organisers: University of Sheffield (Sheffield Methods Institute, Open Research Working Group and University Library) and Project TIER.

This three day event consists of a one-day symposium with conference style presentations, a workshop on teaching transparent methods of empirical research, and individual or small group in-person consultations with Project TIER Directors.

https://www.sheffield.ac.uk/smi/events/teaching-reproducible-research-and-open-science-conference

36 of 47

END

37 of 47

END

38 of 47

Reserve slides follow

39 of 47

40 of 47

41 of 47

Tour of Depts

42 of 47

Image: Stable Diffusion (via huggingface.co),

prompt “ancient foundations building towards a sky of stars”

43 of 47

Open Research Tour (of Departments…and Flagships?)

1. Meet with DRIs

- use open research aspects of REF environment statements / 5 year strategy plans to structure discussion on enhancing open research

2. Provide researchers with updates on TUoS activities and opportunities

3. Hear from researchers about barriers to research improvement in their work

44 of 47

Collectively we already know what to do

but this knowledge is

- under-shared

- under-evidenced

- not packaged for dissemination

- not clearly distinguished from suboptimal practices

investigating the factors that contribute to robust research

disseminate

best practice,

support training

evaluate

implement

45 of 47

Research Improvement Strategy

This slide stolen from Malcolm MacLeod

46 of 47

A workable strategy

Add one new practice per project

47 of 47

“Made in Sheffield”