The figures for these Party member respondents are:

Definitely for Remain: 17 per cent ( + 1 per cent).

Definitely for Leave: 58 per cent ( – 2 per cent).

Leaning to Remain: 9 per cent ( + 1 per cent).

Leaning to Leave: 13 per cent. ( – 3 per cent).

Note, not for the first time, the consistency of the survey: the changes are so small as to signify little if anything.

– – – – –

The YouGov poll for the Times that I referred to yesterday may cast some light on the reliability of our survey. It asked about Party members’ referendum voting intentions: the question apparently was: “How will you vote in [the] referendum?”  In other words, it allowed for a response from those Party members who have already made up their minds, or don’t know.  It found –

Remain 31 per cent.

Leave 59 per cent.

Don’t know 11 per cent.

Readers will see that our survey’s Leave total is almost exactly the same as YouGov’s, but that our Remain total is almost half of theirs.  An obvious reflex reaction is that both got an accurate reading of the Leave total, but that one obtained an inaccurate one reading of the Remain total.  Since YouGov ran a poll and we ran a survey it would follow that the former’s Remain figure is far more likely to be correct.

However, it isn’t just the nature of the inquiries that differed: the questions did, too.  By asking respondents if they will definitely vote one way or the other, or else are leaning one way or the other, our survey has an additional level of nuance.  And it may be that the Leave vote among Party members is simply more solid and committed than the Remain one.  Readers will also note that if the two Remain totals in our survey are added together they reach 26 per cent – only five points short of the YouGov Remain total.  The one certainty is that there will be no shortage of polls and surveys between now and polling day, and therefore plenty of fresh material available for assessment.