Reproducibility

A scientific work is only credible if it can be reproduced, checked and repeated by others in as much detail as possible - one speaks of reproducibility .

The results of statistical calculations, (geo)data analyses and simulations etc. can only be reproduced if the underlying digital resources such as data, software, program code and input parameters are available in unchanged form and with all dependencies. A purely textual, illustrated description, as is the case with traditional scientific publications, is often no longer sufficient for digital, data- and computer-supported science. Scientific journals such as the Journal of Statistical Software have therefore started to add digital attachments to their articles by default and make them openly available together with the respective article.

In the meantime, there are also so-called data journals in which research data can be published. These are interdisciplinary journals:

• Scientific data: http://www.nature.com/sdata
• Data Science Journal: http://datascience.codata.org
• Data in Brief: http://www.journals.elsevier.com/data-in-brief
• Data: http://www.mdpi.com/journal/data