Publication of pre-print and code for open access publication
Open Research objectives/practices
- Submission of scientific article to bioarchive as preprint prior to submission
- Open access research paper published
- Peer review process freely available after publication
- All code and data needed to replicate the analysis accessible.
Introduction
In May I was able to publish my first paper and first chapter of my PhD. After uploading a preprint upon submitting the article, we published the accepted article open-access. This, along with the available peer review, enables a transparent documentation of the published article and its review process. Furthermore, all data and code is available freely and permanently at dataverse. This code is a comprehensive documentation of all code used in creating variables and other data handling that occurred as well as the statistical models ran for the analysis (both raw code and outputs). This would be useful for other researchers aiming to carry out a similar analysis of their own but will also allow a thorough replication of our study to be done.
Motivation
Despite the prevalent use of R for both data handling and analysis in my research field, evolutionary biology, code and data is not always made available when papers are submitted. I believe that this is useful as we can learn from each others code but should primarily be done to help increase the replicability of findings. It is only when the code is made available that one can fully understand what has happened to the data in order to produce the findings reported. Ultimately, you cannot expect this behavior from others unless you do it yourself. As such, I hope that my small piece of action can help encourage other researches to do the same and help the field move towards a more transparent, fair and equal research system.
Lessons learned
Upon reflection, I believe there are two big barriers that remain for increasing open research practices. Firstly, publishing code (especially thorough and understandable code that can allow replication) takes a lot of time. This naturally means that in our highly-competitive environment, individuals and groups that do not follow these practices are favored. This is not right and works against open research practices. I believe it is only with all journals requiring the publishing of code that this can be solved.
Secondly, publishing code is often “scary”. For me, as an early career researcher, who is already worried about scrutiny and/or judgement from the scientific community, this is another part that can undergo the same judgement. This may seem trivial but I am sure I am not the only one and I believe that this is definitely a factor with individuals not publishing code. Removing stigmas around “what your code should look like” (ultimately if it is understandable and replicable then that’s all that matters) and having more experienced researchers lead by example and share their code would help combat this.
URLs, references and further information
| Last modified: | 01 November 2023 12.36 p.m. |