Results of PLOS experiments to increase sharing and discovery of research data
For PLOS, increasing data-sharing rates—and especially increasing the amount of data shared in a repository—is a high priority.
Research data is a vital part of the scientific record, essential to both understanding and reproducing published research. And data repositories are the most effective and impactful way to share research data. Not only is deposited data safer and more discoverable, articles with data in a repository have a 25% higher citation rate on average.
With support from the Wellcome Trust, we’ve been experimenting with two solutions designed to increase awareness about data repositories and promote data repository use among both authors and readers. One solution didn’t achieve its expected outcome in the context we tested it (a “negative” result) while the other shows promise as a tool for increasing engagement with deposited data. The mixed outcomes are an example of why it’s so important to share all research results regardless of their outcome – whether “positive” or “negative” results. We hope that our experiences, what we’ve learned, and above all the data and results, can help the scholarly communications community to develop new and better solutions to meet the challenges we all face, and advance Open Science.
Read on for a quick summary of the studies we conducted. Or get the full details from our new preprint on Figshare, and explore the data for yourself.
PLOS data repository experiments, 2021-2023
Experiment 1: Using iconography to highlight data shared in a repository
Question
Does highlighting data in a repository linked to published research articles with an eye-catching graphic in the form of an Accessible Data icon increase access to the associated dataset?
Hypothesis
We hypothesized that:
- the Accessible Data icon would be associated with a statistically significant increase in engagement with datasets, and
- the icon would motivate authors to deposit their data in a repository in order to have the icon applied to their article
Implementation
An Accessible Data icon automatically appeared on any article published after 2014 that included a link in its Data Availability Statement to one of three popular data repositories (Dryad, Figshare, and OSF) beginning on March 29 2022. Readers can see and click on the icon from any eligible PLOS journal article and be directed straight to the data in the repository.
Data
Usage data relating to the number of Accessible Data icon link clicks, and internal data from Figshare, one of three selected repositories. A survey of 4,898 researchers, and follow-up interviews with 12 researchers.
Results
In the first 12 months of the experiment (April 2022-March 2023) we recorded more than 20,000 reader clicks on the icon across all PLOS properties. Through analysis of 543 Figshare datasets linked to PLOS articles, we observed that in the 12 months prior to the launch of the icon, the average number of views received per month was 2.5, rising to 3.0 in the 12 months following the launch (a statistically significant relative increase of 20%).
Survey respondents who were aware of the Accessible Data feature and who had used a repository were asked to what extent the feature influenced their decision to use a repository: 51% answered that they were influenced either somewhat or strongly by the existence of the feature.
Respondents were also asked about the impact of the Accessible Data feature on their future likelihood of submitting to a repository: 40% were more or much more likely to use a repository.
Experiment 2: Integrating Dryad into the Editorial Manager submission system for PLOS Pathogens to ease repository use
Question
If authors have the option to submit their data to a repository as part of journal submission, will repository use increase?
Hypothesis
We predicted an absolute increase of 10 percentage points in the use of data repositories by PLOS Pathogens authors, from the 2020 baseline of 25% to 35%.
Implementation
A new option appeared on the Attached Files screen in the PLOS Pathogens Editorial Manager submission system on October 5 2021, allowing authors to upload research data to Dryad. Authors were redirected into the Dryad system via a popup window where they entered the required information and files and received a Dryad DOI, which they could then include in the Data Availability Statement later in the submission process.
Details and instructions were added to the journal submission guidelines, Editorial Manager submission system, and PLOS website, and the offering was promoted through established channels including email, social media, and advertising.
Data
Usage data from the Editorial Manager submission system and Dryad. Survey data from 654 researchers who submitted to PLOS Pathogens.
Results
About 2% of submitting PLOS Pathogens authors (44 submitting author groups) used the integrated repository feature. PLOS Pathogens’ repository use for the first half of 2023 was 33%, however this appears to be part of an ongoing organic increase that began prior to the experiment.
Among the 574 survey respondents who did not share or plan to share data using the Dryad integration, the biggest reasons included lack of awareness (35%), uncertainty in how to use the integration (26%), and that data had already been uploaded to another repository (21%).
Our next steps
As a result of these experiments and our findings, we’ve made the decision to discontinue the PLOS Pathogens integration with Dryad. PLOS will continue its publisher membership of Dryad and our authors are still encouraged to use it to share data in their own workflows. We look forward to exploring other ways to increase use of repositories. Because of the promising results from the Accessible Data icon, we decided to extend the service to include an additional six repositories, for nine total. We look forward to seeing—and sharing—the outcomes from this next phase of work.