Ipsos political polling methodology

Ipsos’s current political polling design is based upon a telephone quota sample using both landline telephone numbers (generated by random digit dialling) and mobile numbers (a mix of random digit dialling and targeted sample for some hard to reach groups), weighted to demographic population totals, aimed at achieving a representative sample of all GB residents aged 18+ (not Northern Ireland). A two-part voting intention question is used with prompting on major parties and a follow-up “squeeze” question of those who were undecided or refused to the first question, and a turnout filter is used based on two questions (current stated likelihood to vote and previous voting behaviour).

Quotas are set on gender, age, region, whether working full time, ethnicity and educational qualifications. Data are weighted to match demographic population targets from reliable sources (ONS population estimates, National Readership Survey, Labour Force Survey, etc). These weights are applied simultaneously by rim-weighting. Data are weighted by age within gender, work status within gender, ethnicity, car in household within social grade, work sector (public sector/other) within Government Office Region, housing tenure, newspaper readership, and educational qualifications.

Turnout correction

Respondents are included in the final totals if

1. They say they are at least 9 out of 10 certain to vote, and
2. Say that they “always” or “usually” vote in general elections or that “it depends”.

In the final prediction poll of a campaign, this is supplemented by including those who say they have already voted by post, and excluding those who say they are not registered to vote. In final polls for previous elections, but not in 2019, a voting intention was also imputed for those respondents who refused to say how they intended to vote (based on how their newspaper readership compared to those who did give a voting intention).

Remember that all polls are subject to a wide range of potential sources of error. On the basis of the historical record of the polls at recent general elections, there is a 9 in 10 chance that the true value of a party’s support lies within 4 points of the estimates provided by an individual poll, and a 2 in 3 chance that they lie within 2 points. This is especially important to keep in mind when calculating party lead figures.

Reallocations based on Brexit Party/Liberal Democrats/Green Party/Plaid Cymru not standing in every constituency in the 2019 election

After being asked the regular voting intention questions, participants were then asked: And if there was no [INSERT PARTY NAME] candidate in your constituency, how would you vote? In constituencies where a participant’s first preference party were not standing a candidate, their answers have been reallocated to their second party preference.

We continually review our methods, and may make further refinements and changes to our methodology in the future.

Article Updated 2022