But a second YouGov poll shows no change
As well as the YouGov poll in the Telegraph this morning, there is also one in the People tomorrow. The topline figures are CON 40%, LAB 31%, LDEM 18%.
I've not put any changes in the vote, since this poll would have been carried out at roughly the same times as the Telegraph poll (I forgot to note down the actual dates, and the People haven't published them yet - Nigel Nelson just twittered the topline figures). The difference between this and the other poll, which had figures of CON 38%, LAB 31%, LDEM 18% will be just normal sample variation, and compared to YouGov's previous poll in mid-January this one shows no difference whatsover.
It's a reminder that a lot of movement in polls is just down to sample error, and doesn't signify anything whatsover. If a poll shows just a point or two's difference, then unless it is part of a broader trend reflected in several polls it doesn't mean much.
