Skip to main content


Statistical (il)literacy and expert fatigue

In this country it is perfectly acceptable to say “I don’t do numbers”.  It’s a strange cultural quirk, which would be completely unimaginable in China or Vietnam.  Imagine if someone were to say “I don’t do letters – I can’t read at all”.  This would not be socially acceptable at all – being able to read is a basic skill, so why is numeracy different?  It’s almost a badge of honour to say that you’re terrible at maths.  I have friends who teach maths and they frequently have to explain the value of maths to disgruntled parents who think that their progeny don’t need even a basic level of numeracy.   

We seem to have a similar issue with listening to expert opinion.  According to Michael Gove (and he doesn’t seem to be the only one) people “have had enough of experts”.  Again, it’s a very odd cultural quirk, to think that people who have spent a large chunk of their life studying something have a no more valid or informed opinion than anyone else.  Admittedly, it can be hard to work out if someone is an expert or not, and perhaps this is part of the reason that people often prefer to disregard experts in favour of “gut instinct”.

As someone who conducts research for a living I frequently experience this distrust of “experts”.  The fact that my expertise lies in statistics seems to heighten this as people have a fundamental mistrust of statistics.  I mentioned, in passing, to a teacher that teenage pregnancies in the UK are at an all-time low.  The teacher was not impressed as according to her own experience she knew that teenage pregnancies were increasing.  I said that nationally this is certainly not the case, but perhaps Brighton and Hove (where she teaches) buck the national trend.  I checked the figures online and confirmed that teenage pregnancies have been falling there too, but the teacher refused to believe me.  Her individual experience was sufficient to override objective data about the population.  She could not conceive that her experience might be unrepresentative and told me the data must be wrong.  She told me that I should stop relying so heavily on “data and statistics”. 

A similar situation occurred when talking to a student who had run a small survey.  She had asked about 50 women in a small African country when they first had sex, and the median response was 25.  A large national survey of nearly 14,000 women had included the same question and the median response was 19.  The student was convinced that the national survey must be wrong, and wanted me (as a statistician) to explain why.  I explained to her that the national survey (though far from infallible) was designed to be representative, while she had asked only a small number of women in a single city.  She had not attempted to sample the women in a coherent manner, asking people based largely on convenience.  Thus, while the national survey wasn’t necessarily correct, her small survey provided no reason to suspect a problem with the large, national one.  The student listened, but was not entirely convinced - the power of her individual experience outweighed the statistical evidence.  This was a PhD student at a top University, so even the highest levels of education do not guarantee statistical literacy.

Suspicion of experts and statistics pervades all areas of life (or at least my life).  Recently, I was asked, with some opprobrium, why I had chosen to breastfeed beyond 6 months; I responded that the WHO recommends breastfeeding for at least the first 2 years, only to be told by the questioner that “nobody cares what the WHO recommends.”  I could see no reason not to continue breastfeeding when the weight of evidence that breastfeeding benefits both mother and child is huge; it reduces child mortality and morbidity, boosts intelligence, decreases the mother’s risk of breast cancer, and potentially her risk of ovarian cancer and diabetes too.   With attitudes like those of my questioner it is little wonder that we have spectacularly low breastfeeding rates in the UK.  In fact, our overall breastfeeding record is the worst in the world, even though one academic estimated increased breastfeeding could save the NHS £40m a year.  Unfortunately, despite the strength of evidence, the government is “particularly weak” in terms of commitment to breastfeeding; it is highly unlikely that the UK will reap the rewards of increased breastfeeding anytime soon. 

My point is that we desperately need of a change in attitudes.  It’s time that we valued expert opinion for what it is – I’m not suggesting that we blindly accept any particular expert’s opinions, but that we give it an appropriate weight.  It’s also time that we learned to value statistical literacy and the ability to understand and evaluate statistics for ourselves.  Statistics can be misappropriated, but if we learn to be critical of what we are told then there is no reason to disregard all data and statistics in favour of personal, anecdotal experiences.

About the Author:

Dr Melanie Channon (nee Frost) is a Research Fellow at the Oxford Institute of Population Ageing working on the Collen Programme. Melanie is a trained demographer and social statistician, and her primary areas of research interest are the drivers of fertility transition in developing countries, son preference, and gender statistics. She has expertise in the demography of both Asia and Africa, with a focus on Nepal and South Asia.

Comments Welcome:

We welcome your comments on this or any of the Institute's blog posts. Please feel free to email comments to be posted on your behalf to or use the Disqus facility linked below.

Opinions of the blogger is their own and not endorsed by the Institute

Comments Welcome: We welcome your comments on this or any of the Institute's blog posts. Please feel free to email comments to be posted on your behalf to or use the Disqus facility linked below.