Linear growth
Years ago most research papers contained one main claim and some experimental results that supported that claim. Then started to contain one main claim and a functional consequence of such claim and the associated experiments to support it. Then things started to grow in size, the main hypothesis, followed by some claims, consequences, and some experiments trying to link everything together.
This exponential growth in size is not exclusive to high-tier journals but a common trend in scientific literature. This continuous pursuit of a new large and novel story about how something behaves just continuously raises the bar to the new aspiring scientist. To publish needs more results, then need more funding and more work time. Leading to unsustainable and unrealistic situations that most of the research groups will be unable to maintain. And in some cases where work migration status is tight to that job things can escalate quickly.
On one side, it’s ok to have more results and data coming from the same research group. However, it increases the burden on science professionals, as more work is needed to publish something. It also increases the risk associated with scientific publishing, as every journal publishes only novel data. And if someone publishes an independent replication of your work, then your work is not novel.
This results in another behavior, if a large continuous scientific tale can’t be told, the results can be sliced into smaller papers. However, if experiments are also split, then inadvertently the researcher is p hacking the results. As the number of hypotheses changes from the originally intended to test. Also by dividing the paper, the original hypothesis for the experiments will be changed to other ones that match better the narrative.
Nonetheless, if raw data is available in any of the different scenarios, some of the possible biases can be overcome. It also adds value to the paper as it’s easier to review and analyze and extend the work. Yet, a form to trace back the dataset needs to be provided for authors to incentivize the publication of both data and the expert interpretation of the data.
As the size and number of papers increase are also important to add value to the different outputs of scientific development. This offers visibility in the role and amount of work needed to publish scientific literature. Expert opinion is important, but data acquisition and labeling is a field that every day shows its importance. See you in the next one.
