Patrick Flynn
Poll was conducted 3–6 June 2024, with a sample size of 2,077 respondents. Responses were weighted by age, gender, region, education and ethnicity. A ‘squeeze’ question was asked to respondents who initially said they did not know who they would vote for. Squeeze respondents who said they would not vote were then removed from the sample, with others allocated to their chosen party.
Turnout was predicted using a turnout model, which uses British Election Study data to estimate a respondent’s likelihood of voting based on their demographics and how they responded to questions on attitudes towards voting. A manual override was also applied, with those who say they are 0/10 likely to vote automatically assigned a 0% likelihood of voting.
While we finalise research pertaining to false recall of 2019 general election vote, we have chosen not to currently weight by recalled past vote. Given how many voters have switched party since 2019, we expect there to be a higher likelihood of false recall, which could skew our results. Furthermore, given the age divides in each party’s 2019 electorate, there is a significant disparity in expected mortality levels for previous voters of the two major parties. Weighting recalled vote based on the exact vote share of the last general election would not account for this fact. Any future past vote weight for this tracker will attempt to take into account false recall and the number of past voters of each party still in the electorate.
Data tables are available here.
