Science and Technology Group 17 June 2011 John Moss-Jones
Report on 'Risk in the Media', a presentation by Prof David Spiegelhalter at Cheltenham Science Festival 2011.
David Spiegelhalter is Winton Professor of the Public Understanding of Risk at University of Cambridge [60% of time] and also Senior Scientist in the Medical Research Council Biostatistics Unit.
He heads up a small team at Cambridge focusing on the appropriate use of quantitative methods in dealing with risk and uncertainty in the lives of individuals and society. The work falls into the broad category of 'public understanding of science', while work with schools can be considered as 'maths outreach'. The team tries to take a view of the subject that extends beyond the application of probability and statistics, acknowledging that there are deeper uncertainties that cannot be easily put into a formal framework, and that social and psychological issues necessarily play a vital role.
The problem is that the general public has great problems with understanding cause and effect, statistics [indeed numbers in general], and risk. The press capitalises on this. The presentation was a rapid and witty run through the many ways the media alters meanings of their reports often using numbers and charts. Taking notes was a nightmare. Some points he made:
The papers love to scare us, and it seems many of us love to be scared ie we buy and read papers. Illness is a scary subject and many media stories are calling up data to show reasons for getting cancer, a heart attack, a stroke, road accidents and so on.
Headlines: Hidden danger in fruit juices
or Mobile phones can give you cancer. In fact on this one the International Agency for Research on Cancer set the ball rolling, by classifying radio waves that mobiles use into category 2B: 'possibly carcinogenic to humans'. Papers then decided how much they wanted to scare us.
Category 2B contains 266 different agents, most of them being chemicals with long names, which most of us, and almost certainly the reporters, have never heard of. But the list does include factors which people would know: coffee, talcum powder, petrol exhaust fumes and DDT.
The Times went for "Mobile phone phone cancer risk is at coffee level'. The Guardian went with talcum powder. Both these comparisons are benign-ish - likewise their surrounding story.
The Express links mobiles with car exhaust fumes, which we all know is nasty stuff, therefore? mobiles are dangerous.
The overall message here is that, when you're reading a story about risk (or indeed about anything else), you should always ask yourself,
"Why are they telling me this?" and "What are they not telling me here?" That is, be suitably sceptical!
Incidentally there have been a lot of stories linking coffee with cancer. And yet at least 44 studies have shown there is no correlation between coffee intake and cancer.
Papers often make a lot of 'Absence of evidence is not evidence of absence'. But the prof pointed out that absence of evidence does suggest there is a low risk of presence.
It is posssible to alter the meaning people take from a report by framing the story in positive, or in negative terms.
E.g 'Half of all hospitals have above average mortality rates'
or ''Half of all hospitals have below average mortality rates'
Framing was used by the papers in the mobile case above.
Be wary of graphics
Papers regularly show comparisons between situations using shapes - or numbers superimposed on shapes. Note most people perceive the areas of the charts as the key indicator and do not take account of the actual numbers.
The rarer things get, the more press coverage occurs when they happen
E.g.: Railways are by far the safest way to travel, [in UK there has not being a single pasenger death in a train crash since 2007] yet if there are passenger deaths in a rail crash it receives huge coverage.
However about 300 people die on the railways each year - c. 250 suicides and c. 50 trespassers. Largely unreported.
Muddling absolute and relative risk
Papers often give a relative risk but do not give absolute risk. Papers will say for that your risk of something bad happening to you is doubled by some activity - drinking/eating burgers/ fish and chips etc but do not tell you that even when doubled your risk is trivial from the activity.
An interesting concept is the Micromort which is a unit of risk measuring a one-in-a-million probability of death.
Activities that increase the death risk by one micromort, and their associated cause of death:
smoking 1.4 cigarettes (cancer, heart disease)
drinking 0.5 liter of wine (cirrhosis of the liver)
spending 1 hour in a coal mine (black lung disease)
spending 3 hours in a coal mine (accident))
living 2 days in New York or Boston (air pollution)
living 2 months in Denver (cancer from cosmic radiation)
living 2 months with a smoker (cancer, heart disease)
drinking Miami water for 1 year (cancer from chloroform)
Living 150 years within 20 miles of a nuclear power plant (cancer from radiation)
eating 100 charcoal-broiled steaks (cancer from benzopyrene)
eating 40 tablespoons of peanut butter (liver cancer from Aflatoxin B)
eating 1000 bananas, (cancer from radioactive 1 kBED of Potassium-40)
traveling 6 minutes by canoe (accident)
traveling 6 miles (10 km) by motorbike (accident)
traveling 230 miles (370 km) by car (accident)
traveling 6000 miles (9656 km) by train (accident)
flying 1000 miles (1609 km) by jet (accident)
flying 6000 miles (9656 km) by jet (cancer from cosmic radiation)
receiving one 10mrem chest X-ray in a good hospital (cancer from radiation)
taking 1 ecstasy tablet
Hang gliding involves a risk of eight micromorts per trip while scuba diving involves five and a parachute jump (in the US) is about 17.
Conclusion - be careful of everything thing you read in the papers, especially if numbers are involved. Even if you are smart and wary you are probably being misinformed!