RSS head: Statisticians no longer ‘party bores’ in Covid era

“Statisticians used to be the party bores – people would glaze over when we explained what we did – but now people want to know more,” Sylvia Richardson, the new president of the UK’s Royal Statistical Society (RSS), said about the transformation in perceptions of her discipline over the past year.

The organisation has paid keen attention to the use and abuse of numbers, particularly how information is gathered, presented and used to inform public policy, for almost two centuries. Yet the Covid pandemic has made its mission seem more relevant than ever, with statistics not only dominating daily life – in the form of endless Covid figures, graphs and charts – but also being contested like never before.

“We have always been seen as ‘back office’, but the pandemic has reminded people about how important the use of statistics is for their lives, and for having a healthy democracy,” said Professor Richardson, director of the Medical Research Council (MRC) Biostatistics Unit at the University of Cambridge, whose two-year term as president started in January. The pandemic, she said, will “help to attract more people to our discipline, even if it’s an unfortunate way for it to happen”.

In normal times, the RSS is active in the UK policy sphere, advising on everything from how inflation is calculated to the failings of the teaching excellence framework, but that work went into overdrive with its creation of a Covid-19 task force, chaired by Professor Richardson, last March.

“Some members are experts in vaccines, some in epidemiology and some in data transparency, and we’ve been pretty active, particularly around reviewing how health data is presented,” she explained.

These interventions were helpful in resolving some early disputes, including how Downing Street presented daily death tallies, which, until 26 April, showed hospital figures only, she said.

“If you looked at the dashboard from March last year, things are [now] much better presented – slides are clearly labelled, definitions used more consistently and figures are put into context,” said Professor Richardson.

Trying to “encourage the very best analysis and develop proposals on those grounds”, as Professor Richardson describes the society’s work, might not sound controversial, but political expediency often means that policies are not always underpinned by strong evidence bases.

Source: 

Getty

This month, lateral flow tests – on which the government has spent £1 billion – are being used to test children for Covid as they return to school despite the society’s working group on these tests having repeatedly raised serious questions about their accuracy. One task force member, Jon Deeks, professor of biostatistics at the University of Birmingham, has argued that they provide a false sense of assurance to staff and students.

The RSS’ intervention on this issue may not have altered policy, but Professor Richardson hopes that its rapid work on this issue has helped to correct some misapprehensions about the work of statisticians.

“We are sometimes seen as taking years to collect data and interpret it. [But] we’re showing how things have progressed and how experiments can be rapidly designed and carried out quickly while maintaining high levels of trust in the work,” she said.

The work of the RSS’ task force in, for instance, examining the relative risks and rewards of using lateral flow tests also highlights how statistics has become a huge battleground for many recent academic arguments, debunking the concept that the discipline is a dry and dusty affair inhabited solely by featureless bean counters. The RSS has not shied away from some of the more controversial debates: in September, it hosted an online debate about whether the UK census should collect data by sex or gender, or both, with Professor Richardson keen to continue discussion of this and other fraught issues in future.

“We’ve seen huge public debate around how data science has been used in the pandemic and how vaccine trials can be done quickly but rigorously. But we’re also seeing a lot of interest in other ethical questions around data,” explained Professor Richardson.

“At the society, we’ve organised a lot of debates on open access to data, personal privacy and the responsible use of statistics. But we’re seeing new issues arise all the time, like the fairness of algorithms used to decide A levels last year, which we commented on.”

Despite its venerable history dating back to 1834 and past luminaries including Florence Nightingale, the society’s first female member, computing pioneer Charles Babbage and NHS founder William Beveridge, the RSS seems in no mood to be looking backwards.

“We’re anything but old-fashioned,” insisted Professor Richardson, who believes the society is needed more than ever. “We are the voice of good statistical practice and public understanding of data – we have so much to offer at this point in time.”

jack.grove@timeshighereducation.com