Challenges and Advantages of "Open Science"
In his essay How I learned to embrace open science in the July 11, 2024, online issue of Science, PhD student Albert W. Li describes his experiences with practicing "open science."
As a student in China, he had been discouraged from being too open about his research plans, methodologies, and data. His professors gave the typical reasons: someone else might steal his ideas before publication, or early findings might be misinterpreted.
Things changed when he moved to the United States. His professors encouraged openness. However, some peers were still skeptical and had to be convinced of what Li believes are the “tangible benefits” of practicing open science:
When I began making my data and methods accessible, I found my collaborations quickly grew and my colleagues provided insights that significantly enriched my work. Open access to my research led to more citations and increased the impact of my studies. And I knew I was doing the right thing for research integrity.
It is important to understand that in how research has been conducted over the centuries there has not always been a stark distinction between "open" and "closed" states regarding research. Researchers have always shared findings before publication, both in person, socially, and via conferences where researchers would often reveal findings early on to claim ownership and primacy of findings as well as to garner feedback and suggestions.
The potential for "theft" has always existed, no matter how secretive researchers are at different stages of the research process. Science is a competitive enterprise. One only has to read a book like Howard Markel's The Secret of Life: Rosalind Franklin, James Watson, Francis Crick, and the Discovery of DNA's Double Helix to see how cutthroat research can be—and the potential impacts of intellectual “theft.”
But that was then; this is now. PhD student Li and his support for "open science" are operating in a fundamentally different communication world from previous generations. Research collaboration is now simpler, cheaper, faster, and more widespread. Tools are available to share data in real-time. Journals and the peer review system still operate as a "gold standard" conferring approval, status, and prestige, but by the time the journal process has operated, researchers may have already moved on to something else.
As positive as I am about openness and collaboration, some words of caution are in order.
First, bad actors will always exist. Actual theft—of data, ideas, or methods—will always be possible. They may be uncovered faster but by then, harm may have already been done. Errors and false results may have already propagated throughout the system.
Second, the current system, which bestows approval and credit based on publication, encourages volume, and not all of this volume is of high quality. In short, it's easier than ever before to publish junk research.
Third, author Li writes from the perspective of a human researcher collaborating and communicating with other human researchers. Unmentioned in his editorial is consideration of the role of artificial intelligence (AI) tools in the research process and how that research is communicated. For example, if sources of data are scraped and analyzed by LLM tools operating under the instruction of remote researchers unconnected with the source data, how will the data source be acknowledged—and how will AI-based findings be described and credited? Another concern will be guarding against “fake data” given how easily synthetic data can be generated for uses such as system testing.
On balance, I certainly believe that "open science" has significant net benefits, but researchers must always be on the lookout for ways to validate both research credibility and validity.
Fortunately, establishing trust and reputation—often based on personal relationships— will continue to be strong system controls.
Copyright (c) 2024 by Dennis D. McDonald. The image at the top was created by Microsoft Image Creator on July 18, 2024 based on the prompt, “male and female scientists sharing data openly with many other researchers operating around the world.”