The Use and abuse of statistics

STATOIL & INSTITUTE OF IDEAS COME TO METRIC

Statoil will host a lively, informative and interactive debate in association with the Institute of Ideas at Metric Nightclub on Monday 29th October from 5:00pm.
The event will be hosted by Statoil's Katie Jackson, senior vice president, corporate mergers & acquisitions and will be followed by a unique networking opportunity where you can talk with senior Statoil people and take inspiration, ideas and tips for your future career.

For all those attending the debate, free food and drink will be available as well as an opportunity to win the all-new to of the range iPad with retina display.

 The first part of the evening (a debate on the “use and abuse of Statistics”) is now fully subscribed, but the organisers are still taking registrations for a reserve list so if you still want to attend this, or the networking event you should register straight away at https://statoi.cvent.com/imperial

 Join the Debate

While the study of numerical data allows us to see patterns, understand problems and spot underlying causes, should the public be warned that it’s not that simple?

From Hans Rosling’s surprise hit TV show The Joy of Stats to the new fashion for evidence-based policy, there seems to be a new love affair with statistics. Are we now ready to give full credence to statistics or should we still show caution, as Mark Twain points out: “there are lies, damned lies and then there are statistics”?

 While the study of numerical data allows us to see patterns, understand problems and spot underlying causes, should the public be warned that it’s not that simple? After all, the premise of a study might be more important than statistical analysis or presentation of results.

 We too rarely question the assumptions that underlie the figures and often forget that correlation does not always lead to causation. In subjects where causality is indistinct or multifaceted, such as in biology and the social sciences, statistics can be a powerful descriptive and analytical tool. But we must remember that helping to describe and analyse a phenomenon is not the same as proving it. Statistics are often used to beguile us into believing that the past can predict the future. But people are not data points and humans behave in very unpredictable ways.

 Statistics are at the heart of a recent row between campaigners and scientists in the debate over hydraulic fracturing. Josh Fox, the Oscar-nominated director of "Gasland" has claimed that breast cancer rates rose in the Barnett shale region of Texas where extensive gas drilling activity has taken place. He says “In Texas, as throughout the United States, cancer rates fell — except in one place— in the Barnett Shale." Fox has subsequently been accused of misinterpreting causal relationships by a wide range of cancer specialists. Simon Craddock Lee, a professor of medical anthropology at the University of Texas, called the claims "a classic case of the ecological fallacy" because they falsely suggest that breast cancer is linked to just one factor. In fact, diet, lifestyle and access to health care also play key roles.

 How then to understand the role that statistics should play in public life and debate? Can the use of statistics ensure that we see the world as it really is, not just as we imagine it to be, or are we in danger of disengaging our critical faculties when presented with ‘the numbers’? In the context of greater availability of data and an emphasis on transparency, can statistics help us hold politicians and the powerful to account? Is it our mathematical literacy or our acumen and ability to debate that we most need to hone?

 We look forward to seeing on Monday!

IMPERIAL COLLEGE UNION
Beit Quadrangle
Prince Consort Road
London SW7 2BB
Registered Charity No: 1151241
Tel: 020 7594 8060
Fax: 020 7594 8065
Email: union@imperial.ac.uk
Twitter: @icunion
VAT Reg No. GB 240 5617 84
imperialcollegeunion.org