Dr Agustina Martínez-García, Head of Open Research Systems, Digital Initiatives
Cambridge University Library (CUL) is piloting participation in the Thoth Archiving Network, which allows small presses to use a simple deposit option to archive their publications in multiple repository locations, creating the opportunity to safeguard against the complete loss of their open books catalogue, should they cease to operate.
Participation in the pilot has allowed us to explore the implementation of suitable infrastructure, built on interoperable, open, and widely adopted platforms to support discovery, access, and long-term availability of open scholarly works.
Work done so far
We are pleased to share that the Cambridge repository platform participating in the Thoth network is now live at https://thoth-arch.lib.cam.ac.uk/home, and now includes a full back catalogue of two open monograph publishers. This repository is based on the open-source DSpace software.
Through the implementation phase, we have worked very closely with the Thoth technical team to support the implementation and testing of standard and automated deposit mechanisms into DSpace-based repositories. This work has allowed us to further our knowledge and expertise on scholarly and research platforms by using well adopted repository platforms (DSpace) in a new area: open access books and monographs. It has also provided us with the opportunity to test the implementation of additional infrastructure to support discovery, access, and dissemination of such open access content, and potentially experiment with other types of scholarly work.
What’s next
Now that the repository platform is live, we would like to gather insights about volume of content, required storage and staff resources (both infrastructure and user support). This will help us estimating associated costs for provision of such a service as well as preservation costs for the longer term, during the 3-year pilot.
In terms of long-term preservation, we will explore several preservation options, including preserving the content in-house as part of the Libraries’ wider Digital Preservation Programme. The types of material hosted in this platform can provide an exemplary use case of scholarly content that is “preservation ready”, uses open and standard file formats (i.e., PDF and epub) and is accompanied by rich, high quality descriptive metadata.
See this post by the Open Book Futures Team for more details about the pilot:
Dr Agustina Martínez-García, Head of Open Research Systems, Digital Initiatives
We are pleased to announce that our Diamond Open Access Journals at Cambridge platform has launched in May and can be accessed at https://diamond-oa.lib.cam.ac.uk/home. This service will be available initially as part of a one-year pilot project undertaken by the Open Research Systems (ORS) and Office of Scholarly Communication (OSC) teams within Cambridge University Library (CUL).
Project overview
The main aim of the Diamond project is to support Cambridge’s research community in the context of a changing open research and scholarly publishing environment. To meet increasing demand to share research findings we are scoping, assessing, and implementing future services and systems that meet those needs, while contributing to a growing wider open research community and ecosystem. The pilot is being launched off the back of a project to understand the community-led publishing landscape at Cambridge (findings to be shared soon). Researchers in the Office of Scholarly Communication uncovered a vibrant ecosystem of DIY publishing projects at Cambridge that the library is exploring how to support through technical and resource-based approaches.
As part of the project, we are engaging with Cambridge researchers and exploring whether open and community-developed platforms meet their needs around institutional publishing and can be used as the basis for service development in this area. We are using the DSpace repository platform to support this pilot. DSpace is a widely adopted, open-source repository platform, and it is currently the solution underpinning Apollo, Cambridge’s Institutional Repository. In its newest version, it offers advanced functionality and features that can potentially make it a suitable platform for journal publishing, an area we are keen to explore with this pilot.
Where we are at
Main activities of the project are focusing on:
Exploring the implementation of suitable infrastructure, built on interoperable, open, and widely adopted platforms.
Gathering use cases of community-led open access journals at Cambridge, focusing on discipline, journal type, frequency of publication, production standards.
Gathering insights and inform future service development in this area by a) assessing the suitability of the DSpace open-source repository platform as a journal publishing platform; and b) estimating the associated costs and resourcing requirements, both in terms of service management and infrastructure (long-term access, storage, and preservation costs).
The following four Cambridge student-led journals have agreed initially to participate in the pilot, and we are also exploring opening participation to additional journals in the upcoming months.
Cambridge Journal of Climate Research (Climate Research Society, first issue now available in the Diamond platform)
Cambridge Journal of Human Behaviour (Anthropology)
Cambridge Journal of Visual Culture (History of Art)
Scroope (Architecture)
What’s next
The next iteration of work for the pilot will focus on assessing the resources and costs involved in transitioning from pilot to service. Ensuring long-term preservation and access comes with several associated costs and it is critical to assess these when evaluating sustainable approaches to service development. Examples of cost elements that we will consider include onboarding (initial implementation) fees, hosting and maintenance fees, volume of content and storage costs, persistent identifier (DOIs and ISSN) minting and publisher databases indexing services costs, etc. We will also explore suitable long-term content preservation options, including approaches such as integrations with existing preservation services such as CLOCKSS (https://clockss.org/), or assessing in-house preservation via the services that are currently being developed as part of CUL’s Digital Preservation Programme.
Dr Mandy Wigdorowitz, Open Research Community Manager, Cambridge University Libraries
The promise and pitfalls of preregistration meeting was held at the Royal Society in March 2024. It was organised to address the utility of preregistration and initiate an interdisciplinary dialogue about its epistemic and pragmatic aims. The goal of the meeting was to explore the limitations associated with preregistration, and to conceive of a practical way to guide future research that can make the most of its implementation.
Preregistration is the practice of publicly declaring a study’s hypotheses, methods, and analyses before conducting a research study. Researchers are encouraged to be as specific as possible when writing preregistration plans, detailing every aspect of the research methodology and analyses, including, for instance, the study design, sample size, procedure for dealing with outliers, blinding and manipulation of conditions, and how multiple analyses will be controlled for. By doing so, researchers commit to a time-stamped study plan which will reduce the potential for flexibility in analysis and interpretation that may lead to biased results. Preregistration is a community-led response to the replication crisis and aims to mitigate questionable research practices (QRPs) that have come to light in recent years, some of which include HARKing (Hypothesising After Results are Known), p-hacking (the inappropriate manipulation of data analysis to enable a favoured result to be presented as statistically significant), and publication bias (the unbalanced publication of statistically significant findings or positive results over null and/or unexpected findings) (Simmons et al., 2011; Stefan & Schönbrodt, 2023).
The meeting brought together scholars and publishers from a range of disciplines and institutions to discuss whether preregistration has indeed lived up to these aims and whether and to what extent it has solved the problems it was envisioned to address.
It became clear that the problems associated with QRPs have not simply disappeared with the uptake and implementation of preregistration. From the perspective of meta-research, the success of preregistration appears to be largely disciplinary and legally dependent, with some disciplines mandating and normalising it (e.g., clinical trial registration in biomedical research), others greatly encouraging and (sometimes) requiring it (e.g., psychological science research), and others having no expectations about its use (e.g., economics research). The effectiveness of preregistration was shown to be linked to these dependencies, but also related to the quality and detail of the preregistration plan itself. Researchers are the arbiters of their research choices and if they choose to write vague or ambiguous preregistration plans, the problems that preregistration are assumed to address will inevitably persist.
Various preregistration templates exist (such as on the Open Science Framework, OSF) and some incentives for preregistration are recognised, such as the preregistration badges awarded by some journals, making it a systematic and straightforward exercise. In practice, however, it is not always the case that sufficient information is provided, and even in cases where preregistered plans are detailed, they are not always followed for various pragmatic or other (not always nefarious) reasons. As such, the research community are cautioned to not assume that preregistration equates to better or more trustworthy research. Rather, the preregistration plan needs to be critically reviewed as a standalone document in conjunction with the published study. This is important because preregistration plans that are usually deposited into repositories (e.g., OSF, National Library of Medicine’s Clinical Trials Registry) are seldom evaluated as entities of their own or against their corresponding research articles. Note that this is unlike registered reports which are a type of journal article that details a study’s protocol that does get peer reviewed before data is collected and if reviewed favourably, is given an in-principal acceptance regardless of the study outcomes.
Other discussions centred around the utility of preregistration in exploratory versus confirmatory research, whether preregistration can improve our theories, and how the process of conducting multiple but slightly varied analyses and selecting the most desired outcome (also referred to the ‘garden of forking paths’) affects the claims we make.
The overall sentiment from the meeting was that while preregistration does not solve all the issues that have arisen from QRPs, it ultimately leads to more transparency of the research process, accountability on the part of the researchers conducting the research, and it facilitates deeper engagement with one’s own research prior to any collection or analysis of data.
Since attending the meeting, I have taken away valuable insights that have made me critically reflect on my own research choices, and from a practice perspective, I have downloaded the OSF preregistration template and am documenting the plans for a research project.
Given the strides that have been taken toward improving the transparency, credibility and reproducibility of research, researchers at Cambridge need to consider whether preregistration plans should be included as another type of output that can be deposited on the institutional repository, Apollo. We have recently added Methods and preprints as output types which have broadened the options for sharing and which align with open research practices. Including preregistration could be a valuable and timely addition.
References
Stefan, A. M., & Schönbrodt, F. D. (2023). Big little lies: a compendium and simulation of p-hacking strategies. Royal Society Open Science, 10(2), 220346. https://doi.org/10.1098/rsos.220346
Simmons, J. P., Nelson, L. D., & Simonsohn, U. (2011). False-positive psychology: undisclosed flexibility in data collection and analysis allows presenting anything as significant. Psychological Science, 22(11), 1359-1366. https://doi.org/10.1177/0956797611417632
