Welcome to the Opinium Election Polling Centre.
Opinium accurately predicted the outcome of the 2012 London Mayoral Elections and were one of the top pollsters in the London Mayoral Elections in 2008, the 2009 EU Elections and the UK General Election of 2010. In September 2012 Opinium began working with the Political team at the Observer Newspaper and now release a fortnightly Opinium/Observer Political Poll.
Find out how we calculate our political forecast:
This survey is conducted online using Opinium’s highly engaged consumer panel of UK adults.
This research is run from a representative sample of adults (aged 18+) in England, Scotland and Wales
- The sample is scientifically defined from pre-collected registration data containing gender, age (18-34, 35-54, and 55+) and region (North East, North West, Yorkshire and Humberside, East Midlands, West Midlands, East of England, London, South East, South West, Wales, and Scotland) to match the latest published ONS figures
- The sample variables are defined within our panel management software and then randomised using a purpose built visual basic macro
- Opinium also takes into account differential response rates from the different demographics groups (e.g. those aged under 35, typically have a response rate of 1/3rd of those aged over 55), to ensure the sample is representative
- To test the robustness of the sample we have run additional controls, measuring such things as the respondents’ propensity to vote for each major party, their past vote, and confirmed that the results were generally similar to some major mass surveys such as Eurobarometer and the European Elections Study
- Data is weighted to the profile of all adults aged 18+ (including households without internet access) according to gender, age, region, working status and social class
- Age, gender and region weighting figures are taken from the published ONS statistics. Social class weighting figures are taken from NRS published figures. Working status weighting figures are derived from a combination of ONS published figures and our own research
- Our weighting matrix has been developed in conjunction with Professor Michael Bruter of the LSE
All those that state they intend to 'definitely' or 'probably' vote for a specific party are included in the final calculation, and are all treated as equal. This is due to the fact that they aim of this survey is to accurately estimate the current voting intentions of the public, rather than predict the levels of voter turnout. This approach maximises the effective sample size, giving a base of around 1,400 responses for each wave. This sets the margin of error at 2.4% (or better) for a party with a 30% share of vote