Just Divide Everything They Say by Eight

An early-era Monty Python sketch has a scene in a bed shop in which all the staff are perfectly normal, but for the fact that you need to divide what they say by ten, or multiply it by two, or not say the word ‘mattress’.

And today, as often before, reality has matched the Pythons .

From the open-access journal PLoS Medicine comes an article showing that industry-supported soft-drink science is nearly eight times more likely to conclude that its sponsor’s product are beneficial compared to independently funded studies.

The study’s authors conclude:

We agree that financial conflict is not the only cause of bias. Indeed, the hallmark of modern scientific method, the a priori hypothesis, indicates a preconceived notion of how an experiment will unfold. Moreover, long-standing scientific viewpoints, career considerations, and even political opinions might color study design or interpretation. However, these types of individual bias tend to cancel themselves out among large groups of scientists over the long term. While one investigator’s career may rise on a cherished theory, another’s may rise by debunking that theory. We contend that financial conflict of interest is qualitatively different, producing selective bias that acts consistently in one direction over time…

The authors suggest a range of remedies – restricting the publication of industry-funded science, voluntary rejection of industry funding, etc. In the meantime, consumers should be taking a very large grain of salt when reading about the therapeutic effects of drinks and unguents funded by the food industry. Theirs is the best science money can by and, like democracy, its purchase by the few corrupts it absolutely.

And, whatever you do, don’t mention ‘funding bias’.