[ad_1]
A few years ago, two researchers selected the 50 most commonly used ingredients in a cookbook and badyzed the number of them badociated with a risk or benefit of cancer in several studies published in scientific journals.
The answer: 40 out of 50, a list that includes salt, flour, parsley and sugar. "All we eat is linked to cancer?", They irony, in their article published in 2013.
Your question is related to a known but persistent problem in the world of research: many studies use samples only too small to reach generalizable conclusions.
But the pressure on researchers, the competition between journals, and the insatiable appetite of the media for studies that announce revolutions or major breakthroughs, make these articles continue to be published.
"Most published articles, including in serious and weak journals," says AFP, one of the authors, John Ioannidis, professor of medicine at Stanford, who specializes in studies studies.
The researcher showed in an article in 2005 "because most studies published false".
Since then, he says, progress has been made.
Some journals require authors to provide their raw data and publish their protocol in advance. This transparency prevents researchers from changing their methods and data to find a result, no matter what. This also allows others to check or "replicate" the study.
When they are redone, the experiments rarely lead to the same results. A third of the 100 studies published in the three most prestigious psychology journals could be reproduced by researchers in an badysis published in 2015.
Medicine, epidemiology, clinical trials of drugs and nutritional studies do not work much better, insists Ioannidis , especially during rehearsals.
"In biomedical sciences and elsewhere, scientists do not train enough in statistics and methodology," he adds.
Many studies focus on a few individuals, making the result impossible to generalize to a total population, since it is unlikely that the selected participants will be representative.
– Coffee and Red Wine –
"The diet is one of the most regrettable areas," says Ioannidis, not only because of conflicts of interest with the agri-food industry. Researchers often look for correlations in huge databases without a starting point.
In addition, "measure an extremely difficult diet," he says. How to quantify exactly what people eat?
Even when the right method, with a random study where participants are chosen at random, the execution sometimes leaves something to be desired.
A famous 2013 study on the benefits of the Mediterranean Diet in relation to heart disease had to be withdrawn in June by the prestigious medical journal The New England Journal of Medicine because the participants were not recruited at hazard; the results have been revised downward.
So, what to choose in the avalanche of studies published every day?
Ioannidis recommends asking the following questions: Is an isolated study or strengthening of existing work? Small or large sample? a random experience? Who funded it? Are researchers transparent?
These precautions are fundamental in medicine, where bad studies contribute to the adoption of at best ineffective and, at worst, harmful treatments.
In his book Ending Medical Reversal, Vinayak Prasad and Adam Cifu cite terrifying examples of practices adopted on the basis of studies that were invalidated years later, such as the placement of stents in a cerebral artery. to reduce the risk of stroke. Ten years later, a rigorous study showed that the practice increased the risk of stroke.
The solution requires the collective adjustment of common criteria for research officers in magazines: universities, public funding agencies, laboratories. But all these entities are subject to competition.
"The system does not encourage people to go in the right direction," says Ivan Oransky, co-founder of Retraction Watch, which covers withdrawals of scientific articles. "We want to develop a culture in which we reward transparency."
The media also have their share of responsibility, because he believes they should better explain to their readers the uncertainties inherent in scientific research, and avoid sensationalism.
"The problem of the never-ending success of studies on coffee, chocolate and red wine," he complains. "We have to stop."
Source link
