The idea of ​​reproduction in science is not something new. The disputes between Isaac Newton and Joh Falmsteed centuries ago clearly showed that reproduction and replication were necessary mechanisms for the progress of science. In Newton’s words,

“these and all your communications will be useless to me unless you can propose some practicable way or other of supplying me with observations … I want not your calculations, but your observations only.”

It is undeniable that modern science relies heavily on technology and computing. Many scientific disciplines can be described to some extent as “computational”. Some require large-scale computational infrastructures to carry out their research, others need access to heterogeneous data sets in digital format, and others need analytical processes to obtain the results. In one way or another, the difference from Newton’s epoch is that we are talking about computational reproducibility. Instead of asking for “raw observations on paper”, Newton in our day would ask for “data, scripts, and the computing environment in which Falmsteed had performed the calculations”. It has been a radical change in the medium, there is no doubt, but the idea behind the reproduction remains intact: transparency and integrity in the scientific process.

However, the way research is disseminated has not changed radically since Newton’s time. The content of a paper remains essentially as it was centuries ago, long before the introduction of computing and technology in research methodologies. To scientists of past centuries, the current appearance of a scientific article would be very familiar. But they would have serious problems reproducing a today’s journal article. In part, of course, due to total ignorance of modern technology. On the other hand, however, because the data and resources to understand the results described in the journal article are not included. “Something is missing here! The author says that the data is processed but where is the data?” . Perhaps, they would write a letter like Newton did and wait 6 months for a response.

What is the difference between a traditional paper and a reproducible one? In its simplest form, one more paragraph. A paragraph that indicates where data, code, documentation, instructions, scripts, computational infrastructure and, ultimately, any resources that have been used to reach the results and conclusions set out in the written article can be found. A Data and Software Availability (DASA) section in the methods section of a journal article does this. Easy right?

The important thing is not to include the DASA section in a journal article. Well, don’t get me wrong, include it whenever you can! What is really important is the systematization of describing in great detail the entire scientific process carried out until its culmination in the published scientific article. The DASA section is just the tip of the iceberg, it tells the reader “here is my work“, where to find and access the resources needed for reproduction. All that is not seen, the rest of the iceberg, which is immense compared to the DASA section, are the reproducible research practices and habits that researchers have been acquiring and applying in their daily routine.

You, as a researcher, cannot think of reproducibility a few days before submitting an article for review, such as such as collecting co-author biographies. Being reproducible begins much earlier, when the research project begins, during data collection, during analysis, during writing of the article; in short, throughout the life of the project. Being reproducible implies a change in the process. The key is the process, not the product. By optimizing the process in a reproducible way, you will generate reproducible products (papers) over and over again.

The Reproducible Research Practices course, a training course for doctoral students at UJI, was not intended to teach you to write DASA sections in your journal articles. That’s a side effect. The main objective of the course was to promote a mind shift in your research process and work routine, so that you, as a researcher, assimilate the basic reproducible practices and recommendations as your own habits.