The Norm Chronicles: Stories by numbers and danger is a collaboration by journalist and former producer of BBC Radio 4’s More or Less Michael Blastland and Professor David Spiegelhalter, Winton Professor for the Public Understanding of Risk, which was published at the end of May this year.
Norm, Prudence and Kelvin are the characters in this unbiased tale of the sensationalised risks that we frequently read about in the media – Norm, your average man who leads an average life of risk, Prudence, who suffers from acute anxiety of the apparent dangers around her, and Kelvin, a reckless risk-taker without a care for the consequences. In some way, we can all relate to these characters which are counter-balanced against the actual numbers when they are placed in various situations by Blastland and Spiegelhalter. Each chapter discusses an area of hazard, using both anecdotes and mathematics.
One Amazon reviewer drily comments that this book should be made compulsory reading for buyers of the Daily Mail and Daily Express, and as found in the recent debate ‘Perils of Perception’, held by the Royal Statistical Society and Ipsos Mori, people tend to overestimate statistics, whether it be the rate of violent crime or teenage pregnancy, both of which have been falling in recent years according to the Office for National Statistics but this is not news. Rather Blastland and Spiegelhalter encourage us to think for ourselves and beware of the agendas of others whom are trying to make us think in certain ways.
Statistics Views caught up with Professor Spiegelhalter during the Joint Statistical Meetings last month in Montreal and with Michael Blastland after, to ask about their collaboration and their thoughts about its reception by critics and the general public. You can follow Norm on the microsite that accompanies the book – http://thenormchronicles.com/.
1. What inspired you both to write the book? How did the writing process begin?
MB: It’s partly Andrew Franklin’s fault, the boss at Profile books, who said ‘there’s someone I think you should write with – a man called David Spiegelhalter.’ Funny you should say that, I said. I knew David and had been talking to him, coincidentally, about a book that he might write. We liked the idea of working together, so that part was easy. The harder part was what to write about. For me, it begins with the fancy that popular science writing is a noble ambition, that ‘simple’ is often a lot harder to do than it looks, especially finding inventive ways of making ideas truly accessible and more intuitive. Bayesian statistics hasn’t exactly cracked the pop market, so the idea of doing something on the subjective elements in probability was fun and challenging – though we never use the B-word. Out of that evolved this idea to represent the risk data as clearly and innovatively as possible for lots of different hazards in life – using Micromorts and Microlives – but to combine it with the necessary subjectivity, and organise the data around a single life. And that’s when it got genuinely far out as we decided to use fiction.
DS: As Michael said, we knew each other already and got on really well and did various joint presentations to BBC journalists and worked on More or Less together. We knew we fitted together rather well and made a good double act. It’s worked absolutely brilliantly. I suppose I put down most of the numbers, although Michael did do some as well who is very good at numbers. He’s an English Literature graduate and he wrote most of the fiction – these vignettes of various characters but he also did a lot of work on the psychological behavioural economics, and for which incidentally, he’s currently working on a programme for the BBC. It meant that we could talk about what ‘the risks’ are and we actually demolish that concept at the end of the book – we say there is no such thing as ‘the risk’. But we do talk about numbers and what rates have been in the past and then we put it together in terms of what it feels like as people to be confronted with this kind of information and these kinds of threats.
2. Who should read The Norm Chronicles and why?
MB: Anyone who has ever wondered what will happen next.
DS: People suggest that they’re going to buy it for their dad! It’s a dad’s book?! It wasn’t set out to be a dad’s book but that’s how it works out. My dad is 90 and he liked it, so there’s a good recommendation! It’s got a lot of information, including a lot of numbers but it also tells some quite reasonable stories. I put most of the history in because I really like historical stories. It puts it in quite a human context as well. You’ve got to be a little bit nerdy. You have to like numbers otherwise reading it can be quite difficult, although some people with a background in the arts have read it and enjoyed the stories. What’s amazing we’ve found is that people can get something completely different from reading it than another person has – we got some very good write-ups from journalists and from various different countries and they tend to pick on one thing and just go for that. Dominic Lawson did a review for the Sunday Times and he just picked on the media communication of statistics whereas others just picked on some of the numbers we put in. Maybe if we’d made the book stronger in terms of one single message, then possibly that might have appealed to a wider readership.
The vast majority of people think that violent crime is going up in the UK when in actual fact, it’s been going down for the past twenty years. There is this paradox where the more rare an event happens, the more coverage it receives – so people think something is very dangerous when in fact, it is not.
– David Spiegelhalter
3. Why is this book of particular interest now? Do you think it is more relevant now due to the emerging interest in statistics and data analysis?
DS: I think that it is. There is a growing interest in evidence in data and statistics and what goes along with that, is the misuse in the understanding and evidence of data from statistics. These two narratives are running parallel alongside each other and are absolutely fascinating. With the growth of Big Data and the claim at least by governments that they depend on evidence, which is a pretty spurious claim most of the time but at least they pay lip service to it. Some agencies are getting extremely good at their quantitative analysis. There is really serious attention paid by many government bodies and organizations to data and statistics – what that means of course is that you’ve got to communicate the message of those statistics. What a lot of this is about is that we can manipulate those stories to be told in a certain way. We try to suggest some reasonably transparent ways in order to communicate such numbers. Very often, people with the numbers have also got an agenda and they want to present the numbers in a particular way to make you think along those lines. So we talk a lot about the framing and the way in which stories can affect you.
MB: People seem pretty much eternally interested in crime, money, their children, their health, sex and the like. So I’m not sure I’d claim hot topicality as such, rather that the book addresses all those every-day, here and now forevers, if you see what I mean.
4. What were your main objectives during the writing process? What did you set out to achieve in reaching your readers?
MB: We wanted to make it fun to do and fun to read. That was hugely important. I think we also felt that much media treatment of risk or danger was pretty hopeless – and we thought we should be able to do a bit better. But the main aim has always been to give people ways of thinking about danger and chance, together with the best data we could gather, so that they can make decisions for themselves. We were emphatic about not being prescriptive. That has some roots in Bayesian thought. But in essence we like to say ‘here’s the data, here are some useful ways of thinking about it… then do what you damn well please.’
DS: As Michael says, jokes! We wanted it to be fun to read. We have these archetypal characters called Norm, Prudence and Kelvin which Michael really invented. They are not meant to represent real people and they are cartoon-like. They represent aspects of ourselves, e.g. Norm who tries to be rational human-being and is inevitably doomed. So we had to make him a statistician! That’s the idea as he desperately tries to weigh up the evidence and numbers, and of course, fails hopelessly, partly because he is buffeted around by Prudence and Kelvin. Kelvin is a ludicrously reckless risk-taker who doesn’t take any notice of the evidence and just goes for it and Prudence, who is ludicrously cautious, worried about everything that she sees in the news and reads in newspapers.
5. Were there areas of the book that you found more challenging to write, and if so, why?
MB: The fiction was tough, not least because it was new to us both, but it also has to do several jobs at once. 1) give narrative shape to the whole. 2) illustrate a different psychological point about risk perception in each episode as well as 3) explore a different type of danger, and 4) Norm has to be progressively challenged by the problem of achieving an objective account of probability. And all this ideally has to happen without the mechanics being too obvious, so that you can simply read each episode and find it a plausibly human – and with luck entertaining – story.
…the main aim has always been to give people ways of thinking about danger and chance, together with the best data we could gather, so that they can make decisions for themselves. We were emphatic about not being prescriptive. That has some roots in Bayesian thought. But in essence we like to say ‘here’s the data, here are some useful ways of thinking about it… then do what you damn well please.’
– Michael Blastland
DS: The main challenge was deciding which stories to leave out, for example we left out climate change. We have material in there on people’s attitudes towards GM foods and nanotechnology but we don’t talk about the risks involved, partly because they are very difficult to put into numbers. We have concentrated on areas where there is data available. We enjoyed writing about the more extreme side of risk – sex, drugs, rock n’ roll – all the fun stuff, and I also enjoy writing about transport. I reckon it’s a lavatory book – to be read in short bursts in the bathroom! Short chapters on people’s lives working through various topics and that made it reasonably straightforward to write.
6. Which audience is the book aimed for?
MB: There are several. One is the numerate audience that already likes pop science and will enjoy – we hope – the Bayesian ideas, the statistical examples and principles, and the data, much of which is pretty arresting (discovering that an all body CT scan exposes you to about as much radiation as standing a mile and a half from the Hiroshima A bomb, for example). The other – a longer shot – is the more artistically inclined audience that often finds scientific accounts of human behaviour to be lacking something, well… human. We tried to give the human factor equal billing. There is also history and psychology in the mix. But this is only partly with the audience(s) in mind. It is also because it seems to us that the proper discussion of risk requires a combination of many different perspectives. Some in each audience will not like that – the idea that their own perspective could be usefully enriched by the enemy – but we feel it strongly.
7. With the characters Norm, Prudence and Kelvin, does the reader achieve a greater understanding of risk through reading the book?
MB: That’s the idea. The characters are caricatures to some extent – and we have fun with them – but they’re intended to represent a plausible span of recognisable behaviours and attitudes to encourage reflection on the subjectivity of risk. They are, obviously, very different, but it’s less obvious that any of them is unhinged from reality.
DS: I hope the reader does. I’d like to think people read it and think “Oh, I didn’t know that!” I think you can achieve a deeper understanding of risk but the book definitely doesn’t tell you what to do. As Michael says, we think it is really important that people be presented with information in a transparent, balanced way and they then make every effort they can to work out the possible harms and benefits of what their actions might be. Having done that, they should do whatever they damn well feel like, like skydiving, as I did on the BBC4 documentary ‘Tails You Win’! Just go for it and do it. We don’t go for the reckless and we don’t go for the over-cautious. Actually, we do like Norm but we do have those aspects of our personality – we can be anxious or reckless and we should not let either dominate. You are going to be pulled around by your feelings and that is OK, it’s what makes us human. At the same time, you don’t want to be overwhelmed by it and in particular, you want to avoid being manipulated by those with agendas. I guess that is our main theme – to advise people to deconstruct what they hear in the news, statements that politicians make, to enable to think, “Hang on, let’s sit down and work out what this really means. Let’s compare with history, other events and people.” I think it’s just fascinating when I hear, “These countries with their terrible traffic accidents!” and also with the recent criticism of hospital standards and we examine historical evidence and point out that it was just the same in the UK in 1957 and then again, in 1963. What has changed is that we have slightly more advanced.
8. I recently attended Perils of Perception, a panel hosted by the Royal Statistical Society and Ipsos Mori, where it was shown that the public massively overestimates statistics, such as migration and the amount of land in the UK that is actually built on. Do you think it is the same for risk, do people overestimate the risk?
MB: Actually, 29% of people overestimate risk by an average of 87% and the remaining 32% underestimate by about 205%. Which nonsense is another way of saying that people might vary hugely in their estimation of the odds, sometimes they are wildly off and sometimes pretty accurate. So yes, they can be right or wrong about the average probabilities, at least as far as we know them. But the odds alone are only a part of the problem. What people mean when they say something’s risky is also often packed with other values and preferences for things like a feeling of control (at the wheel of a car rather than as a passenger in a plane, for example). On top of which, their own behaviour can make, say, crossing the road, more or less risky than the current average probabilities of being run over. So risk is also contingent. For these and other reasons, a large part of people’s sense of risk has to be subjective, and that’s partly why we go so far as to say that objective risk for an individual thought of as an independent property of the world out there, doesn’t exist. We prefer to think of risk as typically more like an uncertain bet on a horse using scraps of imperfect information mixed with your own judgement: the horse might come in, or it might not…
DS: People do overestimate risk and it’s hardly surprising given the media coverage. The vast majority of people think that violent crime is going up in the UK when in actual fact, it’s been going down for the past twenty years. There is this paradox where the more rare an event happens, the more coverage it receives – so people think something is very dangerous when in fact, it is not. There has not been a train crash in this country for six years but at some point there will be one and some people will make the stupid comment that our trains are dangerous which is absolutely absurd, when they are very safe. As people are reliant on information, they pick up on what the news says which can give a distorted view of the events as they are only interested in the news. We talk about that a lot in the book and I’m not really blaming the media. The whole point about news is that news is different, that’s what news – the one child that gets abducted compared the millions who don’t and that makes people terrified.
9. How did you first become aware of statistics as a discipline?
DS: I did maths at Oxford, and at first enjoyed the pure maths and couldn’t stand the statistics – we learnt it with absolutely no connection with problems or the real world, and no data whatsoever. But then the pure maths got too tricky, and we got taught Bayesian statistics which was beautiful and sensible, and then I went to UCL where I found that statistical inference was both important and exciting.
MB: By the failures of journalism – including my own – to pay enough attention to the pitfalls of simple causal inference. Storytelling is an essential way of organising our ideas, feelings and perceptions about the world. And it’s not as if I’ve ceased to believe in the truth of good fiction. But some of the stories that journalism tells seem to lack any acquaintance with the science of evidence. They just chuck impressions together with scraps of detail and fashion them into a plausible narrative arc to tell us that we’re all doomed or saved or whatever. Good statistics is a habit of thinking about how our storytelling instincts can lead us astray and why – as much as it is a set of technical tools – and it made a tremendous impression on me, thankfully not too late in my career to do things like More or Less… or Norm.