Thursday, 17 November 2016

Journalism which misrepresents academic research is bad for society, says top researcher.

A friend of mine wrote a paper on the happiness of children. It got picked up in the media with headlines such as

Although the study reports positive effects of attending nursery it also found strong positive effects for parents spending time and engaging with their children. The authors conclude that engagement is important – whether this was at home or in a nursery was not so important. The headline deliberately misrepresents what the academic paper tells us in order to create interest.

The media misrepresenting research is nothing new. One only has to look at the amount of things the Daily Mail has said can kill or cure cancer to understand this. In my own field of economics, Simon Wren Lewis has written extensively on how the media misrepresents economists views on austerity. Getting the media to sign up to an independent press regulator is hard enough so perhaps we should try a different approach?

What if news outlets or individuals journalists could sign up to a “Reporting Research Charter”. It could be like the blue tick in twitter or a “fair trade” stamp. Perhaps in online articles when you click on the logo it would take you straight to the charter for more information. Crucially, it would be entirely voluntary. The hope is that readers start demanding a better standard of journalism through refusing to read or cite articles without this accreditation.

What would be in this “Reporting Research Charter”?

This is something that needs thinking about in depth but a few suggestions
  1. Effects given in relative sizes: e.g. “doubles risk of cancer” may actually be from 1 in a million to 1 in half a million.
  2. Contextualise the research: Is this going against the vast body of literature?  Having two people debate both sides can actually result in bias rather than erode it (see climate change).
  3.  Contextualise the effects: e.g.  “bacon causes cancer” . OK, but the effects are very small compared to smoking,
  4.  Headline accuracy:  The headline should reflect the research. This may be hard as often editors select headlines for new pieces.
  5. Separating opinions from facts: e.g. A researcher thinks it may be possible for a link between apples and cancer but there is no evidence for it does not mean “apples may cause cancer says top researcher”

Regarding my last point, I made the headline of this blog up deliberately to highlight this point. I do not have any evidence for my hypothesis although it is my opinion and I am a researcher (OK, I lied about being a “top” researcher). It is my hypothesis that there may be negative effects for society and I think we should try and think about ways to test this hypothesis. In this way you could accuse me of hypocrisy in the sense that I am suggesting to stop something which I do not know if it is bad for society or not. 

My arguments foundations, however, are not built on my hypothesis. The reason I am arguing for a charter is that I want to know what I am reading is an accurate representation of the truth and I think journalists have a moral obligation to report research accurately. Hopefully some journalists and readers will agree with me.