Ipsos-MORI

[gibbon]

Today we have three opinion polls(so far). One shows and increased Labour lead, another shows a slight Conservative vote, and another showing an increase in the Labour vote, notwithstanding that a number of councillors have resigned from the Labour Party over his handling of the was in the Middle East. All very interesting. I suspect that electorally the Middle East war is not the highest issue in voters minds than some people think.

[The Bishop]

Except that the poll series cited here concludes in *June* 2023, so whilst there is some other evidence for your claim it doesn't include this.

[andrewp]

Voting intention (v Nov)
Labour 41% (-5)
Conservative 24% (-1)
Lib Dems 13% (+1)
Greens 9% (+3)
Reform 7% (+3)
Other 6% (-)

[Davıd Boothroyd]

Even without the changes this looks a bit iffy

[The Bishop]

If you only do polling once a month or so, it is liable to bounce around like this.

[Raddy]

This poll had all the makings of ' The 10 best places to eat in UK' or 'The 10 best walks in the UK', so popular with London based journalists. They usually consist of places near where they live in London, places near where their parents or friends live and places where they have a second home or where they holiday.

This poll by Ipsos Mori for the Standard seems to be similar to the voting demographic of the London region perhaps with a bit of Scotland thrown in.

I'm sure there are rules about who gets polled, but does anybody actually audit the figures, or does the student mark his own work.

[The Bishop]

I mean, we do have an actual British Polling Council and nearly all our pollsters sign up to its rules.

Which include being transparent about methodology and breaking down the national figures into regional sub-samples.

Frankly this comment simply reeks of "know all sounding off in their local pub" ignorance.

[CatholicLeft]

Jan 30, 2024 11:36:04 GMT Raddy said:

This poll had all the makings of ' The 10 best places to eat in UK' or 'The 10 best walks in the UK', so popular with London based journalists. They usually consist of places near where they live in London, places near where their parents or friends live and places where they have a second home or where they holiday.

This poll by Ipsos Mori for the Standard seems to be similar to the voting demographic of the London region perhaps with a bit of Scotland thrown in.

I'm sure there are rules about who gets polled, but does anybody actually audit the figures, or does the student mark his own work.

None of this makes any sense when you understand how polling by Ipsos-Mori works.
It is likely that the stark figures are due to the last polling being one end of the MoE in relation to other polls at the time, with it being the other end of the MoE for where the polls stand at the moment.

[Raddy]

Jan 30, 2024 11:47:09 GMT The Bishop said:

I mean, we do have an actual British Polling Council and nearly all our pollsters sign up to its rules.

Which include being transparent about methodology and breaking down the national figures into regional sub-samples.

Frankly this comment simply reeks of "know all sounding off in their local pub" ignorance.

Oh that's alright then!
Is this British Polling Council independent?

I prostrate myself as a supplicant before the giants of polling knowledge on here such as yourself.

Perhaps you can enlighten me as to the general breakdown and geographic location of the 1000 or so people who participated in this poll to educate this 'pub bore'.

[batman]

I’m sure this poll was carried out scientifically enough, but it is highly improbable that Labour would get 49% nationwide even now, and that the LDs would get as little as 7%.

[bigfatron]

Jan 30, 2024 13:22:53 GMT Raddy said:

Jan 30, 2024 11:47:09 GMT The Bishop said:

I mean, we do have an actual British Polling Council and nearly all our pollsters sign up to its rules.

Which include being transparent about methodology and breaking down the national figures into regional sub-samples.

Frankly this comment simply reeks of "know all sounding off in their local pub" ignorance.

Oh that's alright then!
Is this British Polling Council independent?

I prostrate myself as a supplicant before the giants of polling knowledge on here such as yourself.

Perhaps you can enlighten me as to the general breakdown and geographic location of the 1000 or so people who participated in this poll to educate this 'pub bore'.

They will be in the detailed data tables that the pollster publishes, usually a day or so after the poll summary is released.

Your comment was not unreasonable based on the regular release of 'polls' by the Express etc which are merely snapshots of their readers' views, however this sub-forum is devoted to polls conducted scientifically and according to BPC rules, which are generally pretty sound (although we do occasionally see some dubious leading questions,  which will be commented on if found)

I've learned a lot on here about polling and politics; there are some who can be a bit grumpy, but many of the participants are really well-informed on their specialist subjects!

[John Chanin]

Ipsos Mori are interesting because they are old time pollsters who still use phone surveys rather than an internet panel. The problem is that response rates are now very low because of the flood of scammers (I get around 10 scam calls a week on my landline and now run all calls to the answerphone). This arguably means that it is difficult for them to get a representative sample, but difficult in a different way from panel surveys. They do reach people who aren't interested in signing up to a panel, or indeed are still not internet connected at all, which may mean they find more low information respondents. I also believe that they don't make 'adjustments' other than to balance their sample demographically. So no guesses about how don't knows might vote, if they vote at all. As a result of all this there is no herding either.

Of course this survey is likely to be an outlier, but Ipsos Mori is an honest and well respected pollster.

[mattbewilson]

the big swing from Lib Dems to Labour looks like sampling error

[woollyliberal]

This looks like an odd poll all round. The low Reform and Lib Dem scores are out of line with everyone else, plus the unusually large swings.

[batman]

Jan 30, 2024 15:02:51 GMT mattbewilson said:

the big swing from Lib Dems to Labour looks like sampling error

to a large extent yes, but there may be a germ of near-future truth there. It would not be surprising if the two main parties start to squeeze the less major ones as the election gets near, but not by this much this soon or this quickly.

[iain]

Jan 30, 2024 15:59:31 GMT batman said:

Jan 30, 2024 15:02:51 GMT mattbewilson said:

the big swing from Lib Dems to Labour looks like sampling error

to a large extent yes, but there may be a germ of near-future truth there. It would not be surprising if the two main parties start to squeeze the less major ones as the election gets near, but not by this much this soon or this quickly.

Worth remembering that smaller parties traditionally have traditionally expanded their support during election campaigns, when the media is compelled to give them semi-equal treatment and, in the case of the Lib Dems, tactical voting feeds more into poll-response. That notably hasn't happened for the Lib Dems at the past three elections, but IMO there are clear reasons for that (in the public eye already during coalition, in 2017/19 they had attracted geographically unconcentrated anti-Brexit voters who could be squeezed during a campaign) and this time round, when the Lib Dems are all but forgotten, we are more likely to get the old pattern.

I could be wrong, but let's not assume that squeeze is necessarily the most likely outcome as we approach an election.

[Sibboleth]

John is correct: Ipsos-MORI is a traditional polling firm which means that its figures will always fluctuate (and quite randomly) more than normal, especially as its sample sizes are fairly small. But there are also significant advantages to relying less on assumptions, as we've seen from them in quite a few elections over the years (and, notably, they become less bouncy as polling day nears and people start to switch on more). Anyway, they are a credible firm and less prone to ethically compromising behaviour than is the norm.

[doktorb🏳️‍🌈🏳️‍⚧️]

Jan 30, 2024 13:22:53 GMT Raddy said:

Jan 30, 2024 11:47:09 GMT The Bishop said:

I mean, we do have an actual British Polling Council and nearly all our pollsters sign up to its rules.

Which include being transparent about methodology and breaking down the national figures into regional sub-samples.

Frankly this comment simply reeks of "know all sounding off in their local pub" ignorance.

Oh that's alright then!
Is this British Polling Council independent?

I prostrate myself as a supplicant before the giants of polling knowledge on here such as yourself.

Perhaps you can enlighten me as to the general breakdown and geographic location of the 1000 or so people who participated in this poll to educate this 'pub bore'.

I've been on this forum on and off for years and know the following to be true:

A. This sub forum tends to only support and promote BPC affiliated pollsters. I recall a few upstarts appeared with a heap of scepticism about their veracity. We're not amateurs, here, we can spot when the water is both tight and slack.

B. The BPC is all we've got, and they're pretty good as an ombudsman. 

C. Opinion polls will always have outliers. The biggest of all is the general election, the results from which annoy *everyone*

[gibbon]

IPSOS-MORI also do online polling as well as telephone polling.

[willpower3]

Jan 30, 2024 13:33:03 GMT batman said:

I’m sure this poll was carried out scientifically enough, but it is highly improbable that Labour would get 49% nationwide even now, and that the LDs would get as little as 7%.

Is it? Labour often seem to poll in the mid-40s when the Greens and Reform have a combined percentage of around 20% (which seems improbably high). Both this Labour percentage and the Tory percentage (which is notably higher than what e.g. YouGov have been showing recently) aren't outside the realms of possibility in an election.