o2r @ Open Science Conference 2017, Berlin

Due to the overall topic of our project, we felt the Open Science Conference (#osc2017) taking place this week in Berlin would be a great chance to share our ideas and meet like-minded folks. We were happy about the notification that our poster was accepted and even made it into the top ten (of altogether 57 submissions), which allowed o2r team member Markus to give a three-minute lightning talk and present a project poster. Both days included interesting talks given by international speakers (see full programme) and in this post Markus reports on the trip. The first day covered several topics related to o2r, for example, data infrastructures (see European Open Science Cloud). Speakers also mentioned social challenges such as a new reward system and incentives required to motivate scientists to conduct Open Science – a key issue in the Executable Research Compendium-concept as well. In times of fake news and the credibility crisis, keynote speaker Prof. Johannes Vogel strongly encouraged in his opening talk to set a good example in the field of Open Science and convincingly put scientists in charge of the issue.

A few people I talked to liked the idea of making the dataset the actual publication and the paper being “only” the supplementary material. It might be interesting to play around with some thoughts on that: Will institutes focus on publishing datasets instead of papers? Is “data collector” a new job title?

The lightning talks and the poster session were a success. Several visitors were keen to ask questions and to get explanations on technical and conceptual details. I hope that I was able to answer all of them in sufficient detail. If you think I didn’t, please don’t hesitate to ask me or in case of doubts, my colleagues Docker Daniel and Metadata Marc. You should also take a look at Conquaire, an interesting project in the context of reproducible research.

One highlight was a visit by Open Science Radio, who also published a short interview on opening reproducible research.

In the evening, we had a wonderful dinner next to dinosaurs (I am not talking about the scientists 😉) organized by Museum für Naturkunde Berlin, a museum of natural science. In this impressive atmosphere, we were able to network a bit and to continue discussions.

The second day was rather education-driven. However, we do also want to enhance and extend the understanding of scientists when examining a paper by using our ERC. Why not addressing students, too? We still dream of a reproducible and interactive atlas.

It was interesting to see that the great majority of guests and speakers focused on open data when discussing challenges in Open Science. Mentioning source-code was rather the exception although reproducibility was perceived as being part of Open Science. For this reason, I think that our contribution to the conference was relevant as we treat (open) code and software as being equally important. I mentioned this aspect in my lightning talk, too, and tried to highlight the importance of source code during the poster presentation. One might argue that open code is implicitly included in open data or open methodology. However, we should not rely on vague interpretations and make explicit what is required to rerun analyses. In the future, submitting, for example, analysis scripts should be as mandatory as it is demanded for datasets.

To conclude, here a few take home messages:

1. Rewards and incentives that motivate to conduct Open Science are key issues
2. We have to engage people from society to increase trust in scientific results (tackle credibility crisis)
3. Problems are social – not technical. BUT: we have to provide scientists with working examples, otherwise they don’t know why to use it and how.
4. Open Science strongly focuses on data and educational aspects.

P.S. The next time you read about guidelines, recommendations on open data, try to replace it by source code. The argument still works, right?