Science moves forward by collaboration, specifically when researchers verify others’ scientific findings. Unfortunately, a large and growing proportion of studies published across various disciplines may be unreliable due to challenges with research and publication practices, largely because of mounting pressures to publish. While the perception and awareness of a “reproducibility crisis” across the scientific community varies, as highlighted by a survey in Nature, it is clear that promoting transparency and accountability in science is necessary to ensure the credibility of the science you read… as well as the research you publish.
Discussion regarding scientific integrity and responsibility in research, and by extension, problems with publication-driven science, recently peaked in response to Nobel Prize winner Frances Arnold’s retraction of a published article from the prestigious journal Science. Arnold stated that not only was the work irreproducible, but raw data and other critical entries essential to the final analysis were absent from the lab notebook.
Reaction to Arnold’s decision to publicly retract the article was largely met with praise, with many scientists echoing her message that results must be reproducible to be credible, and that it is is critical to be scientifically transparent in research. While this event garnered significant attention by the scientific community and public alike, we must ask: how can we solve the reproducibility crisis?
Efforts to promote transparency and reproducibility are ongoing, many of which are directly related to research ethics. For example, new incentives are encouraging scientists to share all research findings, not only the most exciting results, thus limiting selective reporting. Pre-registration of experiments allows researchers to declare their research questions, methods, and hypotheses, and when they have completed their experiments, publish all results regardless of a positive or negative outcome. Furthermore, funders, publishers, and journals are collectively working together to improve reproducibility; for instance, Nature has implemented reproducibility checklists as part of peer review.
In the end, the replication of important findings—by yourself and other, independent investigators—is fundamental to the accumulation and verification of scientific results. By ensuring the reproducibility of your own research, you promote the scientific community’s core mission of delivering reputable discoveries.