Ask the experts: Trusting Statistics in Today's News
Dr Katie Chicot is a Senior Lecturer, Staff Tutor in Maths & Statistics at the Open University. Katie is involved in many projects which bring maths to the public and schools. She is the CEO of MathsWorldUK which aims to make the UK’s first Mathematics Discovery Centre. Katie has served on the council of the UK Mathematics Trust with whom she developed a brain teaser app called ‘Perplex’.
Kevin McConway taught and researched in statistics at the Open University from 1980 until he retired in 2016. Much of his research was in interdisciplinary teams across several subjects, applying statistics to issues in science. He also has a longstanding interest in the way statistics is reported in the media. Kevin is now Emeritus Professor of Applied Statistics at the OU. His work with the media continues; he helps train journalists on how to report statistics and science more generally, and is often called upon to comment on statistical issues for journalists.
Kaustubh Adhikari is a Lecturer in Statistics at the Open University. His research involves applying statistical tools to medical and genetics questions. Kaustubh obtained his PhD in Biostatistics from the Harvard University, working on mathematical modelling of epidemiology. He then worked as a postdoctoral researcher in statistical genetics at the University College London, before joining the OU.
Kaustubh’s research has explored the genetic and evolutionary basis of many aspects of our appearance, for example hair greying, face shape, or skin colour. A key component of his research is to increase the diversity of research participants around the world, in particular with under-represented ethnicities.
After Dr Chicot welcomes everyone and does introductions, Professor McConway will talk about:
What is statistics for in this context? Answer: particularly, statisticians know about dealing with variability and uncertainty. This can be a problem with press and TV reports about statistics, because statistical results aren’t often totally clear-cut – there’s doubt, and it can be hard to report about things that aren’t certain one way or the other.
So can you trust any of it? There are checks and balances. For example the UK government statistical service is regulated by the Office for Statistics Regulation, part of the UK Statistics Authority, which has been independent of government since 2008. The Office for National Statistics (ONS), which produces many key official statistics (including some on COVID-19), takes its independence very seriously, and has a tradition of transparency and openness. Not all government statistics come from the ONS, but the UK Statistics Authority still has an oversight role and doesn’t hesitate to criticise openly any Government spokespersons (including ministers) or Government departments that aren’t coming up to scratch. This doesn’t mean that everything is perfect, but on the whole you can trust ONS.
Then Dr Adhikari will go on to discuss: Scientific publication, pre-prints, and peer review; How these are being affected by the urgency of Covid-19; Use of flawed statistics in decision-making, and their media coverage; Difference between data analysis and modelling, particularly for forecasting.
He will also refer to his article on OpenLearn where he talks about COVID-19:Making decisions based on flawed statistics.