Brian Moynihan, Global Graduate Meat Department, Bord Bia – Irish Food Board
The recent EAT Lancet report grabbed headlines when it was published. It contained multiple recommendations for drastic dietary changes. It was a PR headache for certain sectors of the food industry. The studies which formed the basis of this particular report included work from more than 30 scientists at some of the most prestigious food and nutrition institutes in the world – making it all the more credible. However, a reputable name and institute does not always equal quality research. All research is subject to some degree of uncertainty- it’s the measure of this degree that separates research from good research. Recently it hasn’t been adding up in the food industry.
Brian Wansink is a top food researcher from the famous Food and Brand Lab at Cornell University, his work has been cited more than 20,000 times. He is responsible for some of the most renowned studies and concepts of consumer behaviour. When six of his studies were retracted in one day last year, it caused quite a stir in academic circles and the food industry as a whole. He has now had 15 studies retracted due to questionable research methods and he is due to be removed from his post effective 30 June 2019.
This scandal has shone a light on what may be a much wider problem in the food research industry and indeed academic research across all disciplines. The psychology research field came under scrutiny recently when it suffered a “replication crisis”, meaning that the results of published studies could not be replicated when the studies were repeated by peers or other research groups.
Publish or Perish
Scientists are concerned about the current state of academia. They are competing for a finite amount of funding, it helps to have as many published studies as possible in the battle for this funding. Time too is valuable, which means that scientists may be desperate to discover any publishable findings from painstaking research. Best practice can sometimes be forgotten or misused in the heat of this battle.
Currently, a P Value of .05 is required to have a study published in a top food journal. It is argued that this value allows far too much room for manipulation of results and too high a possibility of getting a false positive (25% according to statistical experts) – something which can be exploited quite easily. The solution to this may be to demand higher standards, making it virtually impossible for a poor quality study to be published.
At what cost?
This could be seen as stunting progress as it would dramatically increase the time and money needed to run a study worthy of being published and consequently would mean fewer studies and fewer verified findings.
It is important that we educate ourselves on the pitfalls of scientific research methods. We know that not all research is created equal. Bias, corruption, negligence and poor practice are always a danger in research. However, even objective and exhaustive research carried out to the current required standards is still very much susceptible to misleading findings. The media needs to take responsibility for how these findings are communicated to the public. It might be tthe ime that they come with a caveat, if Wansink has taught us anything it is that sometimes scientific fact is more suggestive than definitive.