Thursday, September 4, 2014

What we know about upcoming polls

People are always asking me "when is the next poll due?", and my answer is usually that I don't have a clue.  In fact, my answer is no different today - but we do have an unusual degree of clarity about a few polls that can be expected to appear over the next week or so.

TNS-BMRB say their next poll will be released in the middle of next week.  Their fieldwork is always more out-of-date by the time of publication than any other firm, but hopefully this timetable means that a decent chunk of the fieldwork will have been done after the second leaders' debate.

YouGov seemingly have two polls coming up in quick succession - one on Sunday for the Sunday Times, and one early next week for the Sun.  (Unless the original plan for a poll in the Sun was scrapped in favour of the Sunday poll.)  I must say I find it profoundly depressing that the Sunday Times have seemingly ditched their previous house pollsters Panelbase, who were presumably too Yes-friendly for them at this crucial stage of the campaign.  Let's hope no sneaky McDougall-esque apple-and-oranges comparisons will be made in Sunday's paper.

Panelbase have had another poll in the field this week.  Our old friend Mike "can't be arsed" Smithson claims to have been told that this one will be published, which on the face of it is inconsistent with the client being either Yes Scotland or the SNP, because they would presumably wait to see the results before deciding whether to publish.  My first thought was that it must be a Sunday Times poll, but that was before I heard the news about the ST switching to YouGov.  So a bit of a mystery.

Meanwhile, a dog's breakfast of a poll has been reported in the Telegraph, apparently conducted by an outfit I've never heard of before called TubeMogul.  Every instinct in my body tells me it's a voodoo poll with no scientific validity (ie. self-selecting respondents), but I can't say that for sure because no meaningful information is given about methodology.  At face value, though, it's an excellent poll for Yes - there's a 50/50 split among women and a 44/56 split among men, which presumably averages out as Yes 47%, No 53% - exactly the same as reported by both YouGov and Survation.  Needless to say, the Telegraph have somehow managed to convert this into a "disaster for Salmond" narrative by pretending that only the tiny subsample of 16 and 17 year olds actually matters!

17 comments:

  1. TubeMogul there, highlighting once again the Yes campaign's "man problem". Um, yeah...

    ReplyDelete
  2. Nice to see they actually balloted their members before declaring, unlike the No-supporting GMB.

    Having said that, a 52/48 split is closer than I'd have expected.

    ReplyDelete
  3. @Keaton
    but it precisely reflects the prediction Ken MacDonald (he of Radio Scotland's excellent - and therefore axed - Headlines programme) made of the final result.

    Ken never said whether he thought it would be 52/48 in favour of independence, but given his independent mind,I'm guessing that was what he meant!

    ReplyDelete
  4. James - I am sure that your skepticism re the TubeMogul poll is correct. But isn't there a "self-selection" problem with all internet polls?

    ReplyDelete
  5. I thought the idea behind the Panelbase poll is that the SNP already know the results and are waiting for the best time to release it, i.e tomorrow's Daily Record.

    ReplyDelete
  6. These pro 'Bwitish' union polls like the one done in the Torygraph rag is called Predictive Programming by the London Media Industrial Complex, constant Polls show 99% of Scots saying no !!! in other words, don't bother voting, resistance is futile, the British state knows best. When the reality is the Poll is bent, they interviewed 5 people in Alistair Darling's kitchen, probably his brother, sister, dog, cat, goldfish.

    ReplyDelete
  7. "But isn't there a "self-selection" problem with all internet polls?"

    In a literal sense, yes, but there's a world of difference between weighted online polls from the likes of Survation and ICM, and "we'll just take whoever answers" polls.

    ReplyDelete
  8. ''i.e tomorrow's Daily Record.''

    Surely you mean Daily Retard, 20 pages of television, gossip, insignificant trivilities and 20 pages of mind numbingly tedious football / bread and circuses at the back of that rag.

    Well, you gotta keep the native population brainwashed, completely ignorant and misinformed of important matters like the McCrone Report or Elm Street Guest House etc etc. '' Wickedness in High Places.''

    ReplyDelete
  9. When I see the Daily Record my IQ suddenly goes into a tail spin, strange that, isn't it ?

    ReplyDelete
  10. I'm not sure it's really more focussed on trivialities than any other tabloid.

    ReplyDelete
  11. They're supposed to be a NEWSpaper, I think the clue is in the title, but the Daily Retard is a Liebour / Pro Bwitish unionist propaganda rag, so obviously they want a dumb down Scots population.

    ReplyDelete
  12. Salmond will be guest editing the paper tomorrow, so I'm sure it'll be slightly better than normal. To be honest, it mainly seems like No folks who are scared/aware of this mysterious Panelbase poll, and they're all assuming that Salmond is going to splash with it on the front page of the Record tomorrow if its shows a tie or Yes lead. Seems unlikely to me.

    ReplyDelete
  13. The nutters in the RMT were disavowed by labour because they started giving funds to tommy sheridan's party 10 years ago. That showed good jusgement didn't it :-)

    They lost members hand over fist in scotland particularly over that (they only have 2,500 members in scotland vs 80,000 in the whole of the uk), to the point where the only members left are the militant socialist worker party types. Many moved to the TGWU (now part of Unite). I'm surprised yes only got 52% (exc dks)

    ReplyDelete
  14. Oh!!!
    Ladbrokes have suspended betting on Yes being ahead in the next PanelBase pool.

    ReplyDelete
  15. Don't get excited, Patrick - they were probably reacting to a flood of bets based on speculation on Twitter (of the very sort we've been indulging in here!).

    ReplyDelete
  16. Survation have tweeted that they have a poll coming on 11/9/14 for Daily Record

    ReplyDelete
  17. If the Panelbase poll was not commissioned by the Sunday Times then it was most probably the SNP or the Yes Campaign that commissioned it. I would have expected the poll therefore to be used in Friday's Yes Scotland guest editorship of the Daily Record. But there has been no mention of it - just rumours about its findings on the net. Could this mean that Yes Scotland have pulled it? A possible reason for this might be that it doesn't show the widely anticipated yes lead.

    ReplyDelete