Reproducibility Hackathon – ReproHack (DSC-2022-07)


11.02.2021


10 AM - 4 PM

Workshop for PhD students and Postdocs


Speaker:
Dr. Sandra Zänkert (Data Steward)
Data Science Center

Location:
On campus (room tba)

The workshop will be held in
English.

REGISTER HERE

If you have any questions regarding our workshops, please feel free to write us an E-MAIL.






« Back
During a ReproHack , short for Reproducibility Hackathon, you attempt to reproduce published research from a list of proposed papers with publicly available associated code and data. You get to work with other people’s material in a low-pressure environment. In this half-day event, you will be able to choose a paper from a wide range of research areas including medicine, biology or psychology but also geography, physics or computer science (find the full list of papers here ). We record and discuss our experiences on a number of key aspects, including reproducibility, transparency and reusability of materials which we will also learn to formulate useful and appreciating feedback to the authors. You can also bring and try to reproduce a paper from your own field of research. But first, please contact the author to get their consent to formulate feedback on the reproducibility of their work.

EVENT DESCRIPTION

In recent years, there has been a growing interest in the importance and value of reproducible research. But what does it take to achieve reproducibility of research results for others? As you might know, it can be quite a challenge to prepare not only data, but also the code for the analysis, as well as the preliminary and final results in such a way that our future selves remember what we did only six months ago. Or have you even ever tried to reproduce the results of a colleague or research partner, perhaps even on a different computer system?

During our reproducibility hackathon you will have the opportunity to get your hands on some data, code and results from researchers who aimed at making their research reproducible and who ask for your feedback on whether they succeeded.
Together, we will learn how to provide this feedback in an appreciative and constructive manner and thereby providing the opportunity for improvement.
Our hackathon is also an opportunity for researchers to help others learn from their work by submitting their papers, code, data and documentation for reproduction and review (more information below).

We will pre-select some easy to reproduce papers from the ReproHack homepage. But of course, you are free to explore the complete list of papers . We will mainly focus on R and python, but you are welcome to work with your own ‘tools’.

WHY ATTEND?

Reproduction of research is a beneficial activity in itself. During the hackathon, you will also learn about different approaches to make research results reproducible. This offers you the opportunity to develop customised strategies to make your projects reproducible. You will also have the opportunity to apply and expand your R and Python skills with the support of DSC staff.

TARGET AUDIENCE

This hackathon is for beginners, as well as people who have experience with reproducibility practices. The event is open for researchers from all disciplines of the University of Bremen and other institutions of the U Bremen Research Alliance (15 – 20 participants).
Please note:This is a hands-on workshop, so please bring your own computer. Basic R or Python programming skills are required to take part in this workshop. To create the right hackathon atmosphere, drinks and snacks will be provided.

Time Event
10:00 – 10:30 ReproHack Introduction (DSC)
10:30 – 10:45 Get to know each other – Ice breaker session
10:45 – 11:00 Choose paper, find groups
11:00 – 12:30 1st round of Reprohacking
12:30 – 13:15 Lunch Break
13:15 – 13:30 Get together, reflection on 1st round of Reprohacking
13:30 – 15:00 2nd round of Reprohacking
15:00 – 15:30 Formulate Feedback
15:30 – 16:00 Get together, Conclusion
  1. Submit a paper for us to reproduce!
    • We invite researchers to submit their paper (pre-print or published), including code, data and documentation necessary to reproduce and review the results. This provides the opportunity to receive constructive feedback on the documentation and reproducibility of your results. You can submit your paper via mail or directly on the ReproHack Hub. You can associate your paper with the DSC Reprohack Event or submit it openly. Click here to submit a paper. You can also consult the author guidelines for more information.
    • To get an impression you can browse the full list of submitted papers .

  2. Join us for the event and review and reproduce a paper
    • Join us for the hands-on ReproHack, work with other people’s code and data and learn more about reproducibility on the way! You can work alone or join together in groups.

Please note that we will mostly be able to support you with R and Python related code and analysis. But of course, you are free to work with any code or language you have access to. Due to time constraints of our event, we recommend to refrain from complex machine learning analysis.