Monday, January 16, 2017

Experimental Reproducibility Drama

Experimental reproducibility is the only thing an investigator can guarantee about his study. Science relies on reproducibility to gain confidence in ideas. It’s what makes science respectable, worthy of trust and value. Which basically means that if it’s non-reproducible, y’all are bogus!

Publish or perish, the official motto of this modern scientific era. This idealism puts in risk the value of scientific experiments due to professional pressure, competition and stress, affecting the quality of their work. The rush to have a science epiphany dominates over the mundane activity research is built on, repetition. These days it’s all about quantity over quality. 

Through the years there has been reported numerous papers with data that can’t be reproduced. Many graduate students come in clutch, taking this opportunity to use it in their advantage and publish their “own” work driven by a deception bias. What they do is very clever: take a paper, repeat the experiments, and if the results are different than the ones reported… Congratulations you have yourself a publication! Which leads to an interesting phenomenon, students getting a PhD based on their ability to prove that someone else is wrong. Yes, this will get them papers, which is a requirement in all PhD programs. But students should be rewarded because of the novelty and creativity they’ve put into their work, not because they are trying to reproduce data. It’s not their responsibility to do this nor they should get a degree by doing it.

In the article “Trouble at lab”, the author states that in fact there are some organizations that pay to reproduce findings already published, like Laura and John Arnold Foundation a charity that looked at 50 of the highest-impact cancer findings published between 2010 and 2012, and Blog Syn, a website run by graduate students dedicated to reproducing chemical reactions reported in papers.

If groups that do this repetition science exist, should PhD programs allow their grad students to keep publishing papers doing the same thing? Is it a techs or grad students work?

No comments:

Post a Comment