Scientific Practice

Scientific results are often not reproducible, polished or not even published. How can we maintain and improve good scientific practice?


Workshop Scientific Practice - Conclusions

We should create awareness for the non-reproducibility-topic. A pillar for improvements would be an intense mentoring of young scientists by the PI. Master and PhD thesis should normally start by reproducing experiments they want to build their work on; this reproduction should be published and recognized. To improve the publication system, we should have full PhD theses instead of paper theses; including negative results. The thesis should be public and searchable. The broadest possible solution: All scientific work is published in one journal only; all researchers have the right to publish the same limited number of articles per year. The peer review could shift from pre- to postpublication.

After the workshop, the following issues were additionally raised:

  • Peer-review should be made a lawful place where whistleblowers are protected and rewarded.
  • Research proposals should be divided into different categories based on their riskiness whereby some funding should be allocated to low risk research such as reproducing published reports.
  • All scientific data should be made available with very few exceptions.
  • Irreproducibility should be regarded as normal and no longer be treated as a taboo. To this end, an open space should be created where scientists can publish their reproducibility efforts and different parties (including authors of the original study) can enter into an open, positive dialogue. These reproducibility efforts should be properly accredited and the process fast and rewarding.

Workshop held at the congress «We Scientists Shape Science», January 2017, Bern. Conclusions are based on the debate in the workshop and on online comments by the participants.


News on Reproducibility

News on Integrity

Join us on Twitter

we scientists shape science

#wescientists

More on Scientific Practice

ScienceGeist