Keith Humphreys is given prominence in the journal Addiction to make powerful points about the harm that can be done when scientists are too eager to share new findings. Too eager means bypassing the peer review process that can highlight research shortcomings and often limits the potential for misunderstandings. He points to the recent headlines which suggested that cocaine was no more addictive than Oreo biscuits (cookies across the Atlantic).
A few days after the sensational story began spreading on internet news sites and blogs, other scientists were contacted by responsible journalists for their views of the study. This quickly made apparent that the research in question came from a press release arising from an undergraduate student project that had not been published or even presented at a conference. The contents of the press release were passed along uncritically by many news outlets. Facebook, Twitter, Reddit and the like did the rest.
There was a response, but too late:
Then a round of news stories and commentaries appeared which noted serious flaws in the research and questioned its astonishing conclusions. Following the backlash, the college news service changed the sensational title of the press release and added a new line to its text: ‘The results are preliminary and subject to further scientific review’. Indeed, but at that point the closing of the stable door was but a distant sound in the ears of the horse that had galloped away.
Humphreys argues that this kind of premature reporting damages the reputation of scientists and the authenticity of good research. It also makes the science vulnerable to malign influence and misuse.
What might be done to dissuade less scrupulous scientists from trumpeting their version of the research findings prior to publication? Humphreys suggests journal editors could simply refuse to publish when researchers behave badly and go public prior to peer review and publication. That would leave the foul stench of self-trumpeting firmly in the orbit of the researchers and might prevent future transgressions.