Lies, damned lies and statistics
20 May 2013
With the UK Statistics Authority’s criticism of the DWP's use of figures on welfare benefits, misrepresentation of statistics appears to be back on the agenda.
New research, conducted by Ipsos MORI for the Royal Statistical Society and King’s College London has identified a lack of confidence in politicians using official statistics accurately when talking about their policies – only seven percent felt they did so.
Politicians were also the least trusted when providing information (8 percent). Scientists (74 percent) and academics (63 percent) are the most trusted.
Some 61 percent say they are confident they understand statistics in the context of government spending cuts. When asked if people could explain the difference between ‘debt’ and ‘deficit’, 82 percent choose the correct definition.
However, only half demonstrated that they actually do understand the definitions in the following question: ‘True or False: the national debt will always go down if the deficit is decreasing’. Some 28 percent answered ‘true’ and 20 percent didn’t know – while 52 percent correctly said this is false.
This confidence is partly reflected in the British public’s ability to answer simple mathematical questions – 92 percent correctly say that 50 is 25 percent of 200. However their ability to deal with probabilities is much lower – only a quarter of the public knows that the probability of getting two heads in two tosses of a coin is 25 percent.
The research also revealed that the public appears to put little value on the understanding of numbers. People are four times more likely to say they would be proud of their children if they excelled in reading and writing (55 percent) than if they were very good at numbers (13 percent). Only 6 percent would be embarrassed to admit poor numeracy skills while 15 percent would be uncomfortable admitting to poor reading and writing skills.
The survey findings also provide a useful snapshot of how the public makes decisions. For example, 52 percent of the population still believe that politicians draw conclusions based on principle rather than evidence. This mirrors how the British public makes their own decisions. When forming opinions on government performance numbers are not the top priority. It is anecdotal evidence and personal experience rather than statistics which seem to play a decisive role (by 46 percent to 9 percent).
Denise Lievesley, head of the School of Social Science and Public Policy at King’s, says those who are excited by the opportunities provided by Big Data believe that the numbers speak for themselves. "This is not the case," she adds. "Even large datasets are the product of human design and we have to understand the context and limitations of such data in order to draw valid conclusions. The science of statistics is even more important in such an environment. Statisticians have an important role to play in equipping people to make use of evidence."
Hetan Shah, executive director of the Royal Statistical Society, says our economy now requires most people to handle data and numbers. "It is really important that as a nation we up our data skills before we get left behind," he says. "The starting point is schools - government is looking again at school curricula and now is the time to really strengthen the quantitative and data handling skills of our young people.
"Secondly, we need all policy makers and politicians to get basic statistical training, so they know how to assess and use evidence. And finally we need journalists to get savvy about data, so we have fewer 'bacon cures cancer' type headlines. The RSS is running a campaign called 'getstats' to promote each of these strands."
Bobby Duffy, managing director of Ipsos MORI Social Research Institute, believes it has never been more important for the public to understand and critically review statistical information on how our economy and society is doing.
"It is therefore worrying that the public has so little confidence in the use of statistics, that their understanding of basic concepts of probability and the difference between debt and deficit is so shaky - and that relatively little value seems to be placed on statistical information or the skills needed to make sense of it," he adds.
Contact Details and Archive...