Statisticians who designed exit poll are winners of election

News

  • Author: Statistics Views (Source: University of Warwick)
  • Date: 09 June 2017
  • Copyright: Image appears courtesy of Getty Images

Royal Statistical Society President Sir David Spiegelhalter has announced today via Twitter that statisticians are the real winners of the General Election. In 2015, a team of statisticians designed the exit poll model that correctly predicted a hung parliament and this success has now been repeated. One would expect a pat on the back from Nate Silver but he has chosen to tweet a complete history of British polling foul-ups in one chart. Pourquoi, monsieur?! Yes, 2015 was not spot on like 2005 and 2010 but as the University of Warwick's Statistics Department has rightly said, this exit poll was achieved using the same methods and the same team and should be considered a triumph for statistical methods.

thumbnail image: Statisticians who designed exit poll are winners of election

As the Department explains, 'The exit poll is based on a fairly small sample (100–200) of voters at each of a fairly small number (100-odd, out of 40,000 in all) of polling stations. There is always a risk that either the polling stations selected, or the particular voters who respond to the exit poll, will turn out to be untypical — in regard to their vote-changing since the previous election — of other voters or other polling stations. To some extent this can happen purely by chance, or "bad luck".

The statistical approach outlined above is carefully designed to minimise the impact of an unrepresentative sample, but the risk of a "bad" sample can never be completely eliminated. A reasonable question, then, is "How accurate should we expect predictions based on an exit poll to be?". This is a hard question to answer precisely. Roughly speaking, though, based on past experience:
◾A House of Commons majority prediction that is within 20 seats of the actual outcome is a reasonable aspiration from a well-conducted exit poll.
◾Getting the House of Commons majority right to within 10 seats should be regarded as an exceptionally accurate prediction.
◾Predicting the size of the largest party in the House of Commons exactly is highly unlikely — but, as evidenced by the results in 2005 and 2010, even such unlikely events do happen sometimes!'

In this link, exit polling is fully explained and full credit must go to the team of analysts: At recent elections the analysis team has included: John Curtice (University of Strathclyde, Dept of Government), David Firth (University of Warwick, Dept of Statistics), Steve Fisher (University of Oxford, Dept of Sociology), Jouni Kuha (LSE, Dept of Statistics), Clive Payne (Nuffield College Oxford) and Neil Shephard(University of Oxford, Dept of Economics).

Related Topics

Related Publications

Related Content

Site Footer

Address:

This website is provided by John Wiley & Sons Limited, The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ (Company No: 00641132, VAT No: 376766987)

Published features on StatisticsViews.com are checked for statistical accuracy by a panel from the European Network for Business and Industrial Statistics (ENBIS)   to whom Wiley and StatisticsViews.com express their gratitude. This panel are: Ron Kenett, David Steinberg, Shirley Coleman, Irena Ograjenšek, Fabrizio Ruggeri, Rainer Göb, Philippe Castagliola, Xavier Tort-Martorell, Bart De Ketelaere, Antonio Pievatolo, Martina Vandebroek, Lance Mitchell, Gilbert Saporta, Helmut Waldl and Stelios Psarakis.