Professor John Robertson OBA

YouGov, two days ago, published their latest poll of support for Scottish independence and found a 3% No lead, the first NO lead since June and the first major lead for NO since, again, their own two polls in March 2025, at a time where other pollsters find steady and increasing YES leads in the context of increasing frustration with Westminster governments. YouGov have not found a YES lead since April 2024. Survation have not found on since March 2021.

How does this happen? Surely they all use the same methods to ensure reliability?

Nope.

In their last poll, YouGov made sure their sample had 540 No voters and 436 Yes voters from September 2014 regardless of the fact that 500 000 mainly No voters had passed and the same number of mainly Yes voters had joined the electorate. Similarly, in their last poll, Survation adjusted their sample from 417 to 313 in favour of Yes voters in 2014, to 404 to 326 in favour of No voters. The smaller gap of 81 in the Survation poll compared to that of 106 in the YouGov poll may explain the latter’s larger NO lead.