Why do scientists not tackle the issue of reproducibility?

Labstep
Labstep
Aug 22, 2017 · 2 min read

Reproducibility is a taboo topic amongst scientists. The main idea behind experiment reproducibility is that the same published results will be obtained when the protocols are repeated by another scientist. It is validation that the claims for the experiment are true and the results are reliable enough to be useful for further research. Despite the large amount of published work, 50% of researchers agree that there is a significant reproducibility crisis.

Reproducibility is a problem. More than 70% of researchers fail to reproduce another scientist’s experiments. Moreover, over 50% of researchers failed to reproduce their own experiments. It is estimated that as 90% of academic literature is not reproducible and therefore potentially unreliable.

Within science, this issue is not widely discussed. Even though most researchers have failed to reproduce experiments, fewer than 20% of researchers of researchers say that they have been contacted by another researcher unable to reproduce their work. Researchers do not want to risk appearing incompetent or accusatory. There is also a great air of secrecy with their own projects and do not want to reveal too much.

A scientist’s career is closely tied to their number of journal publications. These same journals can be reluctant to publish negative findings. Researchers are also almost always competing for coveted limited grants and positions. Researchers therefore face a situation of ‘publish or perish’ where they will prioritise publication over reliability. They are so focused on publishing that they make mistakes in experiments, fail to record data, miss key protocol details, poorly use of statistics and, deal with the bureaucracy in a laboratory. Often, there is variability in reagents and difficulty in replicating specialised techniques. Researchers use data and results that align to their experiment aims and ideas. This saves time, money and keeps them on track for publication.

Scientists produce vast amounts of experimental data but little is used for further work, due to reproducibility issues or negative results. This system is flawed. Scientists should move away from the focus of publishing. A strategy for enabling research to be more easily reproduced is to practice open science: the content and process of producing the results need to be made transparent and accessible. Every experimental description should be recorded in great detail, peer reviews and feedback should be encouraged, collaboration between institutes celebrated, and data sets shared.

Comprehensive experimental details should be disclosed so that alterations to protocols should be recorded in great detail to help improve the quality and transparency in research reporting. The usage of new web-based management tools and software can help researchers conduct open science. These online tools allow attachment of data from scans, videos or audio recordings, as well as traditional traditional graphs and tables. More importantly, scientists can be more open about their detailed procedures conducted throughout their research. It also improves reproducibility as detailed histories and descriptions of the experiments are kept on cloud which can be easily accessible and shared.

The reproducibility of experimental data can make or break its potential. Transitioning into this culture of open science through better recording of experimental procedures will improve reproducibility.

)

Labstep

Written by

Labstep

Labstep is a free tool to help the scientific community! Share interactive protocols that automatically track your progress. Learn more at www.labstep.com.

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade