Who responded to the survey?
In the 36 hours before the debate respondents were recruited from a nationally representative sample. The sample were screened based on whether they intended to watch the debate. Then, during the debate, those that had indicated that they intended to watch the debate were sent an invitation to answer the survey. The link became active immediately after the closing remarks had finished. Respondents were again asked whether they had watched the debate – those who had not watched the debate were not allowed to answer the survey.

Is the sample representative of the GB population?
No. Whilst the invitations to answer the survey were sent out to a national sample, those that did not watch the debate were not allowed to answer the survey, which means that the final sample will not be representative of the GB population. However, the sample represents those who watched the debate, but with a weighting adjustment to match the demographics of the national population (by age, sex, region, socio-economic group and past vote)

How was the data weighted?
Data was weighted by age, sex, region, socio-economic group and past vote.