A pint of publicity, but a wee dram of supporting evidence?
27 February 2013

Here’s an interesting situation for researchers. The news today contains a lead item on new research purporting to demonstrate that surveys underestimate real alcohol consumption, compared to sales. Even an interview with the lead researcher, Sadie Boniface of UCL on the BBC Today programme on R4 at prime-time.

Maybe the overall finding is not really surprising to many market researchers, not least, for example, due to the phenomenon of socially desirable behaviour when answering questions in a field such as alcohol consumption. However, I wanted to know more – especially about how consumption had been linked to sales.

The BBC website contains little method information, but hints that the item might be linked to the European Journal of Public Health.

The Guardian coverage is no better. A press release on the UCL website provides a little of the methodology, but not enough. However it confirms that the paper is published in the EJoPH as hinted at by the BBC.

Well, a look at their website shows nothing! If it’s there it is well hidden, and certainly not in the latest issue of that journal, published at the beginning of February.

Maybe I’ve missed something, but I’m sure that many researchers would like to investigate the basis for the claims, and what methodology was used. 

Two years ago, the MRS worked with the RSS and CIPR to produce guidelines for the use of statistics by the media.

In my view, research attracting high profile coverage needs to be easily traceable to the source, but unfortunately that is not always the case, despite the best endeavours of bodies such as the MRS, RSS & CIPR.

The research is hopefully based on very sound evidence, but how do we know?