The ability of phone polls to accurately predict the outcome of the EU referendum has been called into question by YouGov, which states they include too high a percentage of graduates.
To give some context, there's been wide discrepancies between online and phone polling in the buildup to the vote, with phone polls putting Remain comprehensively ahead but online polls suggesting a dead heat between Leave and Remain.
So, which is right?
Well YouGov, which runs online polls, has conducted some research and says that phone polls include far more graduates, which may have skewed the findings.
It's an issue because there are demographic correlations on how people intend to vote, such as age, geography and education.
Those who have been in education longer typically feel more favourably toward the EU.
So, YouGov says, phone polls basically include too many people in the societal group most likely to back Remain. However, it concedes that it is basing its calculations on Populus polls alone, as other organisations conducting phone polls do not ask about their respondents educational level.
As we reported in more depth yesterday, the type of polling used is producing vastly different results, and determining which is more accurate will only be known when the result is out.
What will be critical either way, however, is turnout. Leave voters typically have more enthusiasm, which should get them out. That is, if you are someone who dislikes the EU with a passion, you probably aren’t going to forget to vote.
Meanwhile, the highly educated typically turn out, and will be more likely to back Remain, but turnout of younger people, who tend to be more pro-EU, is a concern for pro-Remain campaigners.
It's not surprising then that both sides have portrayed the risks of staying and leaving the EU, ushering their sides to just turn up and vote on the day.