| layout | title | keywords | ||||||
|---|---|---|---|---|---|---|---|---|
page |
Reproducible Research @ AGILE '20 |
|
- The reproducibility committee updated the review process schedule in collaboration with the publisher and created a report on the results and lessons learned from the 2020 reproducibility review. Find the updated process description and the report at https://osf.io/7rjpe/.
- Daniel reported on the status of the AGILE conference reproducibility review at the online AGILE General Meeting on June 23, 2020. See the slides and the slides PDF with full speaker notes for details.
- 6 papers of 22 accepted manuscripts were successfully (partially) reproduced. This corresponds to 27% percent of accepted papers.
- The first results and lessons learned of the reproducibility reviews at AGILE were presented at the 15th Munin Conference on Scholarly Publishing 2020, organised by UiT, The Arctic University of Norway; see abstract, slides and a video recording of the talk Improving reproducibility of geospatial conference papers - lessons learned from a first implementation of reproducibility reviews at https://doi.org/10.7557/5.5601. Watch the video recording on the proceedings website or on Youtube (below). <iframe width="560" height="315" src="https://www.youtube-nocookie.com/embed/NUTdcrRKTpU" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe>
- Daniel Nüst, Chair (University of Münster, Germany)
- Frank Ostermann (University of Twente, The Netherlands)
- Carlos Granell (Universitat of Jaume I, Spain)
- Alexander Kmoch (University of Tartu, Estonia)
Workshops at the 23rd AGILE International Conference on Geographic Information Science, Chania, Crete, Greece on 16 June, 2020.
This year, the team of the workshop series planned to organise two half-day workshops. In the morning, an introduction to reproducible research by means of the AGILE Reproducible Paper Guidelines. In the afternoon, a practical application of these guidelines. With this split, we hoped to address different interests and skill levels adequately.
It is not planned yet to hold the same workshops at the next year's conference. If you're curious, take a look at the submitted workshop proposals (1 and 2) about our objectives, reasoning, and planned outcomes, or read the short workshop descriptions below.
Introduction to reproducible research by means of the new AGILE Reproducible Paper Guidelines (cancelled)
Reproducibility, replicability, and transparency are widely recognised as crucial topics forscholarly research. This half day workshop teaches participants concepts and best practicesfor reproducible research. The workshop prepares participants for the future of research publications by means of presenting the AGILE Reproducible Paper Guidelines in detail, demonstrating common tools for reproducible research, and discussing participants' questions and own experiences in the light of reproducibility.
In the GIScience community, the AGILE conference is leading the endeavour towards higher reproducibility with establishing the AGILE Reproducible Paper Guidelines for the 2020 conference. The guidelines include detailed information about rationale, motivation and vision of reproducible publications at AGILE Conferences. They also link numerous resources for detailed and specific practices, e.g. for social media data, personal data, or workflows in specific programming languages. After the workshop, participants will have a good understanding of the rationale behind the guidelines and the different aspects of data, code, and environments that matter for reproducibility. They will have an overview of the practical challenges in modern data- and code-based scholarly communication, and an understanding of the authors' and reviewers' perspectives.
The practical aspects of reproducibility are covered in a second workshop in the afternoon.
Reproducibility of research is not a black and white question. A spectrum of reproducibility exists, only some works fall short of the ideal due to relevant limitations but most due to lack of knowledge, incentives, and skills. In this workshop, participants learn the tools and habits to improve reproducibility of their work, and how to create and publish knowledge in an open and transparent way. The workshop covers literate programming with Jupyter and R Markdown notebooks, and proactive management and sharing of computing environments (hardware, software dependencies, containerisation).
Participants benefit most from the workshop if they are familiar with scripted workflows in either R or Python. During the workshop the diversity of tools and practical tips'n'tricks are explored, including practical application of the most useful ones. The skills acquired at the workshop will allow participants to work closer to the "ideal" level of reproducibility as presented in the AGILE Reproducible Paper Guidelines (see the blue column for "ideal" practices), greatly improve efficiency and transparency in their daily work habits, and enable new ways to collaborate with others. Participants will have the opportunity to share own experiences and learn from the mutual exchange.
The workshop assumes familiarity with the basics of reproducible research. These are introduced in depth in a workshop in the morning or can be acquired beforehand, e.g. by studying the AGILE Reproducible Paper Guidelines.
- Daniel Nüst (ifgi), daniel.nuest@uni-muenster.de (main contact), @nuest
- Carlos Granell (Jaume I)
- Barbara Hofer (Z_GIS)
- Frank Ostermann (ITC), f.o.ostermann@utwente.nl, @foost
