There was a flutter of activity across Twitter and blogs the other day, in response to some reports that suggested kids’ increasing TV viewing was having a detrimental effect on mental health. According to the reports, TV viewing should be limited for children even into their teens and banned altogether for under-threes. The issues highlighted here will be familiar to detractors of Bad Science and Bad Reporting, but I wanted to record some thoughts for posterity.
I first saw the story in The Guardian and it was also picked up by BBC News, The Independent, The Telegraph, Daily Mail, Metro and many other outlets. Whilst it’s an interesting and worthwhile area of study, the paper published in the journal Archives Of Disease In Childhood and the subsequent press statements, had a few problems that undermine the stark headlines.
The paper was not an original research paper, but an opinion piece that looked back at some previous research. The chief agitator in this is Aric Sigman, a psychologist whose method of ‘cherry-picking’ evidence Ben Goldacre has had much to say about in the past. ‘Cherry-picking’ is essentially picking the bits of evidence that support a particular claim, whilst ignoring other evidence that doesn’t. As Goldacre points out, a better way to analyse previous research is to perform a ‘systematic review‘. These reviews say exactly how the literature was searched and compiled, which means it is more free from bias and allows others to reproduce it.
As for this specific case, Pete Etchells at SciLogs does a good job at highlighting the problems with the selective nature of the analysis and why it’s important to understand the cause of something before issuing guidance on fixes. I worry that many developmental outcomes – such as empathy, attention, educational performance – are lumped in under the banner of ‘mental health’, but that is probably for someone more qualified to comment on. Professor Dorothy Bishop‘s remarks in the Guardian article are salient too – if Sigman’s concerns are to do with kids just sitting for long periods, you shouldn’t advocate reading books for too long.
My first thought on reading the reports was that the conclusions seem to be based entirely on correlative studies, so it’s hard to determine cause-and-effect. What if children who watch more TV are also more likely to have inattentive parents? You may still see an association between more TV watching and developmental problems if these are both caused in some way by inattentive parenting, but enforcing a reduction in TV time wouldn’t do anything – getting parents to interact more at other times would have the most effect. (For the record, this is just an example of ‘correlation does not imply causation’ and I’m not suggesting this is supported by the evidence!)
On a more general but related point, there is a real problem with defining ‘screen time’, because you’re essentially describing a medium and not an activity. The Mind Hacks blog (written by KCL psychologist Vaughan Bell and Sheffield University psychology lecturer Tom Stafford) has written about this in relation to internet use. Bell has also written about how there have been worries throughout modern history over new technology. Even ‘education’ was once considered a risk to mental health.
As for TV, there are clearly different types of programmes kids can watch – some are aimed at learning and education, some are musical and participatory, some are interactive, and so on. And there are also different contexts in which to watch TV – alone, with parents talking things through, in the background whilst doing other things, etc. Understanding whether different types of TV interaction have different effects or whether other factors in the child’s environment tend to lead to a particular sort of behaviour, are critical in getting to the root of the issue.
The evidence just isn’t strong or reliable enough to make the sort of alarmist claims Sigman has made. And this is why it is again so disappointing to see the same blanket coverage across much of the press, with little in the way of a proper critique (Prof Bishop’s comments aside). It was once again left to bloggers and commentators on social media to provide a more discerning look at the issues.
I want to emphasise that I’m not dismissing these issues, and there may well be problems caused by excessive use – however that’s defined – of certain types of ‘screen time’ (as Etchells notes too). But it’s important to know what you’re measuring and understand the nuances. It is also crucial to have proper evidence before issuing supposed evidence-based guidance.