In two previous post (1, 2), I highlighted a symposium that was held to improve the reproducibility of biomedical research. The published report includes a discussion on cultural factors that have contributed to the high prevalence of irreproducible research.

Culture and nature of science

Whether or not the questionable research practices described in the previous post are the result of ill intent or researcher ignorance, the result is the same.

Attendees of the symposium believed these practices were the result of a culture in which researchers feel they need to publish novel results while, at the same time, seem productive. The incentive structure rewards positive results over robust methods, which leads researchers to unconsciously work to succeed in an environment that undermines good science.

There is increasing competition for faculty positions. A key predictor of a researcher’s career progression is their publication record: the number of publications and the real or perceived impact factor of the journals where articles are published. This hyper-competitive research environment contributes to scientists preferring to publish positive findings. The pressure to attract funding and publish in journals with high-impact factors does not incentivise the best scientific practice, especially when salaries are dependent on factors such as securing grant funding.

There is an old saying that “science is self-correcting”. Unfortunately, this is only possible when science is done rigorously. For science to be self-correcting, studies need to be replicated, all research findings (positive or negative) need to be published, and published findings need to be challenged. Unfortunately, the emphasis on novelty means that direct replication is rare and whole lines of research can develop based on research that is not scientifically sound. This can lead to a false, distorted sense of certainty in one’s research.

Summary

The current scientific culture encourages researchers to embrace their biases and adopt, consciously or unconsciously, practices that generate significant results to publish more papers at the cost of scientific rigor. Many scientists are oblivious to these biases, pressures and practices, but once they are pointed out, these threats to sound science are impossible to ignore.