“Lies, damn lies and statistics”: How to pick apart science headlines


The writer Mark Twain once wrote: “there are just three types of lies: ‘lies, damn lies, and statistics“. I couldn’t have put it better myself.

For a number of reasons, rubbish appears in our newspapers; MMR scares, vitamin supplements and fad diets to give just a few examples. With a recent study in PLoS finding health spin occurring in 41% of abstracts, 46% of press releases and 51% of news items, it seems sensationalising research is all too tempting, but why is this the case? It needs a definite answer. It’s no wonder after spending years on their research, researchers would want to big their research up, or at least play down the negatives, but is that really all? Between the lab bench and the newspaper, misunderstanding and the desire to print a big scoop can often lead to science articles being exaggerated. In the media, science is often presented as unintelligible to us layfolk; a complicated system of big ideas and even bigger words given to us by evil geniuses in labcoats. It really isn’t.

Science simply doesn’t work in great leaps forward on the back of a new discovery. It takes an army of nerds in white coats to gradually chip away at new ideas before they finally become established, if at all. This rigorousness is good for science, but for journalists wanting to report the next big thing, it becomes a constant battle for the research presented to remain credible, and to have any limitations acknowledged. Ultimately, one research paper, however strong, is meaningless without further replicable tests. Now I’m not going to square the blame solely on journalists, but I do want to clear up a few things, especially what to look out for when we read science.

Everyone in journalism knows what sells: Money, sex and health. Now I don’t think I’m really qualified to talk economics; I’m an undergrad biology student, a pretty nerdy one, but still. Perhaps for the same reason I’m not sure I’m qualified for the second topic either. But when it comes to the third (at least the science behind it); Thunderbirds are go.

Stories presented in the media are often backed up with pretty poor evidence; for stressed journalists without the time or knowledge to critically assess scientific information, it is hardly surprising that they would simply parrot press releases that don’t really make the grade. We have to look closer at the research that they got this information from, and when you do you often see same recurring problems.

Chance findings reported as positive effects, small sample sizes or inappropriate groups, and stories based on weak observational studies are just a few things you can pick out. For example, if you are studying the effect of sauna use on sperm quality, checking the tackle of ten Scandinavian men might not give you an accurate picture of what is going on in the trousers of Brixton. Yet reading the tabloid coverage of it you might be none the wiser. There are no two ways about it; it is simply misleading to print research without reference to any weaknesses in the information.

When you look deeper into these stories, you often find some hidden personal or political biases creeping in. We can only imagine what was going through the mind of this Daily Mail journalist when he sent, “Oral sex helps women fight depression” to the printers, but I’m pretty sure it wasn’t world peace. Though, when you then read, “Semen is good for you!” in The Sun, I would say the Mail probably took the moral high ground in that case.

So people exaggerate? Big deal? Well yes actually, it is a pretty big deal. Let’s take the most famous example, some of you might have seen the zombie corpse of which recently rise from the grave on the front page of The Independent. It was 1998, and Dr Andrew Wakefield PhD (now just Andrew Wakefield) published a paper in ‘The Lancet’ outlining a peculiar example of what he saw as cause and effect: the MMR vaccine was directly correlated with the onset of autism in children.

His paper involved 12 children, all of which had behavioural problems and all of which had received the MMR jab. There was no comparison to non-autistic children, no comparison to children who hadn’t had the jab to check for other factors, and all the children came from the same hospital treating behavioural and intestinal problems. Can you see a problem here? Furthermore the parents of 10 of the 12 children had already filed lawsuits over the MMR vaccine before 1998, which they said had given their children the autism. Their lawyers gave Wakefield a total of £435,643 to ‘investigate’ the autism and for his role in the cases against MMR. Under pretences like this, it’s hardly surprising he found a link.

The link has been shown to be false time, time, and time again and it’s been taken as a prime example of crooked research, unreasonably peddled in press releases, and taken up by journalists either not understanding (inexplicably) the limitations of the research, or perhaps not prepared to let something like facts get in the way of a good story. Even today papers continue to print it, in the name of ‘balance’, forgetting that freedom of speech should always come alongside with the freedom to ignore.

I think for me, this is where it really becomes a moral issue. I think it’s wrong to sell people vitamin tablets and fish oil, under the pretence it will let them live another ten years. I think it’s wrong to worry people needlessly by making up ridiculous headlines. But perhaps most of all, I think it’s wrong to tell someone that if they vaccinated their children against some of the worst childhood killers, they were personally responsible for their child later developing autism.

If we are going to have a grown-up and legitimate discussion about what science is and what the results means to us, we need clarity and understanding. The onus is on each and every one of us to critically assess what is placed in front of us during our morning coffee, and if it doesn’t make the grade, simply say no. We will not read mulch at this level.

Andrew Balmer

Image: NS Newsflash on Flickr

Leave a Reply