“Flashy results for fancy journals”. Or: “Rewarding researchers for the number of papers they publish results in a ‘natural selection’ of sloppy science”.

I’ve found this kind of statements, which seem to be based on serious research, in several recent articles that claim that science is currently suffering due to bad incentives for researchers.

Scientists, they say, are not evaluated and awarded by the quality of their questions and methods, but by publishing many high-impact papers with novel findings. Their number of publications is usually positively related to the amount of grant funding they get, which is the other basic measure of “success”.

The greatest discoveries take time, generally decades. But this requires carrying out studies that may fail. We can learn a lot from research whose results are not those we expected. But the short-term funding schemes prevent scientists from taking risks. They are incentivized instead to get positive results they can publish, even to spin their results to appeal to the public.

Some people say that all this is producing unethical science. In the field of biomedical research, there would be an “epidemic of statistical significance”, since the overwhelming majority of papers that include a p-value in the abstract present significant results.

I found this expression that was new for me: “p-hacking”. It refers to the practice used by some researchers of testing their data against many hypotheses, in order to report only those that have statistically significant results.

What is your opinion about this?

Do you think that researchers have bad incentives?

If so, do you have some ideas about how to solve the problem?

More information in these articles:

https://www.sciencenews.org/blog/scicurious/blame-bad-incentives-bad-science

https://www.vox.com/2016/7/14/12016710/science-challeges-research-funding-peer-review-process

More José Eduardo Jorge's questions See All
Similar questions and discussions