The Ipsos MORI pollster, in the National today, has Yes on 53% with a 6 point lead.
In November they had Yes on 51% and an 8 point lead.
These two, plus one for Find Out Now, were a minority in a sea of 11 No leads since October 2023.
Overall the picture (above) has been dominated by No leads from several pollsters but there’s a pattern going back to August 2015, when, in a sea of No leads , Ipsos MORI had a 9 point lead for Yes.
From there forward to the present, Ipsos MORI have had 14 Yes leads, 4 No leads and 2 ties. No other major pollster has found a fraction of that number of Yes leads. Their last No lead, only 1%, was in June 2022.
What happened in the weeks before that poll surrounded again by a sea of polls with No leads? Police Scotland raided Nicola Sturgeon’s house, arrested her husband and confiscated a camper van from his mother’s drive.
Analysts have suggested that weighting samples based on the 2014 referendum can make a No lead more likely but Ipsos MORI seem to do that. Online polls as opposed to telephone polls are thought to favour Yes but Ipsos MORI use telephone polls
So, readers, any ideas? If they keep getting the same results are they perhaps correct?
How accurate have Ipsos MORI been?
In 2019, the actual share of the vote was Con 44.7, Lab 33, LD 11.8 and Ipsos MORI predicted 44, 33 and 12.
in 2017, they were second best, after Survation, 0.7% out.
Talking-up Scotland 'Albatraz' 
Weighting is certainly a factor in the Indy Polls: I recall an item in now- gone blogsite of the Indy-supporting Scottish Polls guru (whose name I now can’t recall!) in which he pointed out that, years after 2014, all but one of the major pollsters were still applying weighting based on the 2014 vote. Ipsos define ‘weighting’ as largely demographic – age , sex, region, education etc. There’s no clue on the Ipsos website about whether they use 2014 demographics or more up-to-date ones, but obviously recent demographics will reflect the increasing proportion of young pro-Indy voters as the much older, more anti-Indy voters inevitably decline.
Other factors that are critical in poll results are:
1) Those sampled: how representative were they of the whole Scottish electorate? and
2) The wording of the questions: were they clearly and unambiguously worded, did they imply a preferred outcome?
Coinneach
LikeLiked by 1 person
James Kelly
https://scotgoespop.blogspot.com
LikeLiked by 1 person
Yes, I think James Kelly is pretty savvy when it comes to interpreting polling data.
Alasdair Macdonald.
LikeLiked by 1 person
Ipsos samples are random, unlike other pollsters who use self-selecting panels and are therefore likely to be biased towards those with a strong interest in politics including those strongly anti- or pro- particular political viewpoints (this can obviously work both ways on issues like Scottish Independence).
Because they do it by telephone there are verbal cues that are stronger than online ones and therefore cause those polled to think carefully rather than click on the first thing they see that they agree with in an online poll, in order to rush through – ie telephone polls are based on a slightly more considered opinion.
Census data for 2022 is only available for age, sex and region so far; data for the other population variables used by Ipsos to ensure a representative sample – ‘social grade’, tenure, working status, work sector, education and country of birth – will not be available until after May 2024, and are therefore likely to be based on 2011 census data.
There is a problem with the way some polls weight by Indyref, because it was 10 years ago and there is more likelihood of false recall and the proportion of samples is declining (eg I didn’t have a vote in 2014 because I didn’t live in Scotland at the time but am in the highest turnout older groups for future elections).
No one knows yet what the impact of the requirement for Photo ID will be on turnout in the next UK general election but it’s not likely to be good.
LikeLiked by 1 person
Support in a Referendum must be up to 60% now. Nearly there.
People who support Independence need to get out and vote for Independence supporting Parties. Voting unionists will not get Independence. They are doing all they can to try and stop Independence and self governance. Make the world a better place. Vote for Independence.
Stop illegal wars, tax evasion, Trident, redundant weaponry. Brexit, the waste of Scottish resources on HS2, Hickley Point.
All the other bad projects wasting £Billions of public (Scottish) monies. Paying Westminster debt for loans and repayments not borrowed or spent in Scotland.
LikeLike