Latest Polls and Latest Odds

[Savage Henry]

Good report for once on ITN News about the missing million voters who are re-registering on the Electoral Roll for the referendum vote.

For whatever reason these people just stopped voting and dropped off the electoral register. These people won't show up on opinion polls, so how will they vote?

As expected the turnout for the referendum vote will be high - 75%-85% as referenda always produce a high turnout.

My own theory is many of these missing voters just gave up on voting for parties. They would rather not vote at all and perhaps apathy has set in.

But they feel a referendum vote is different and meaningful. Will they go to the trouble of re-registering just to vote to keep things as they are? Or will they vote for something different?

Perhaps they just moved abroad and intend to move back, temporarily, for this referendum and haven't bothered to register until now? There's no real reason to suspect they are any more or less likely to vote one way or the other than "domiciled" voters.

[renton]

Ever see those videos from South By Southwest asking about imaginary bands? Hipster banging on about these made-up bands. Not one of the respondents says "I've never heard of these guys". For a poll using standard methodology to return a 15% "don't know" response flies in the face of statistical probability and human nature. People don't, as a rule, like to give no answer. They like to give an answer which is expected by whoever they think is conducting the poll.

Now, the respondents may well have had the impression that this was an SNP/Yes driven poll. But if not, given the significantly higher "yes" vote, I'd suggest this might, at best, be an outlier. Of course, we don't know the methodology or the questions asked or the preamble, but I'd be inclined to discount this poll on the grounds of the size of the undecided vote alone.

Except that every pollster is seeing DKs at around that level or higher (and has been since 2012), much higher in some cases, 25% odds for TNS. Besides it's not a case of giving "No answer" it's a case of saying I haven't made my mind up yet - two different scenarios. The preamble is given in the NNS article and to my mind is fairly neutral - there is also no reason to suggest the methodology is different from other PB polls carried out for the Times. I'm also unsure how a selected sample would guess it was an 'SNP' poll, or why that would make them more inclined to say Yes. Finally, Survation had a poll at 39.3% this month, so 40% in PB is not exactly a stretch.

Edited March 21, 2014 by renton

[Savage Henry]

Except that every pollster is seeing DKs at around that level or higher (and has been since 2012), much higher in some cases, 25% odds for TNS. Besides it's not a case of giving "No answer" it's a case of saying I haven't made my mind up yet - two different scenarios. The preamble is given in the NNS article and to my mind is fairly neutral - there is also no reason to suggest the methodology is different from other PB polls carried out for the Times. I'm also unsure how a selected sample would guess it was an 'SNP' poll, or why that would make them more inclined to say Yes. Finally, Survation had a poll at 39.3% this month, so 40% in PB is not exactly a stretch.

You don't see why people are more likely to say "yes" to a vote they believe is conducted by the SNP? Really?

[Newbornbairn]

As expected the turnout for the referendum vote will be high - 75%-85% as referenda always produce a high turnout.

Got any proof for that claim?

[renton]

You don't see why people are more likely to say "yes" to a vote they believe is conducted by the SNP? Really?

Yeah, why would they?

Also, while picking me up on that, what's your take on dismissing a poll as an outlier becuase of the number of DKs, when every single other poll shows the same?

[vikingTON]

It's perfectly obvious that SH hasn't really been paying attention to the rest of the polling data. Those who have and are indeed paid to do so, like Curtice, haven't raised the above ludicrous points, because they're not utterly rattled by the headline figures.

Thanks for playing anyway.

[Colkitto]

Perhaps they just moved abroad and intend to move back, temporarily, for this referendum and haven't bothered to register until now? There's no real reason to suspect they are any more or less likely to vote one way or the other than "domiciled" voters.

What a million people moved abroad and moved back to vote?

[Colkitto]

Got any proof for that claim?

No.

[Confidemus]

You don't see why people are more likely to say "yes" to a vote they believe is conducted by the SNP? Really?

What a ludicrous suggestion.

If Captain Birdseye polled you asking what your favourite food was, would you say fish fingers?

Of course you wouldn't. Unless you actually do favour fish fingers above all other foods, which I doubt.

[ICTJohnboy]

What a ludicrous suggestion.

If Captain Birdseye polled you asking what your favourite food was, would you say fish fingers?

Of course you wouldn't. Unless you actually do favour fish fingers above all other foods, which I doubt.

I would.

I love fish fingers.

[Savage Henry]

What a ludicrous suggestion.

If Captain Birdseye polled you asking what your favourite food was, would you say fish fingers?

Of course you wouldn't. Unless you actually do favour fish fingers above all other foods, which I doubt.

Fish fingers are delicious.

The point being people in surveys avoid saying anything they think will be unpopular with the census taker.

If captain Birdseye came around to

my house and said, "do you like my fish fingers", I'd probably say yes just to get the boring b*****d off my doorstep.

Likewise, any presumption that a poll is being conducted by one side or other of the campaign is likely to skewer results. Most People will sooner lie than give a true answer they think will make them look different or antagonistic.

Look at people who are asked about voting Tory. Significantly poorer in surveys than actual voting.

[renton]

Fish fingers are delicious.

The point being people in surveys avoid saying anything they think will be unpopular with the census taker.

If captain Birdseye came around to

my house and said, "do you like my fish fingers", I'd probably say yes just to get the boring b*****d off my doorstep.

Likewise, any presumption that a poll is being conducted by one side or other of the campaign is likely to skewer results. Most People will sooner lie than give a true answer they think will make them look different or antagonistic.

Look at people who are asked about voting Tory. Significantly poorer in surveys than actual voting.

Which had nothing to do with the poll being intepreted as coming from the Labour party, and everythign to do with accept social norms about the amoral evilness of the tory party (and which is factored into polling these days, showing that such effects can be screened for)

Still, what about those high levels of DKs huh?

[vikingTON]

The point being people in surveys avoid saying anything they think will be unpopular with the census taker.

If captain Birdseye came around to

my house and said, "do you like my fish fingers", I'd probably say yes just to get the boring b*****d off my doorstep.

Do you have any evidence that the respondents were aware that the poll had been commissioned by the SNP? If not, then it's not actually relevant to the results. Which is probably why a qualified psephologist wading through Scottish voting intentions didn't 'think' of this utterly obvious, yet factually incorrect non-point of yours.

Better luck next time.

[ecto]

3 weeks ago the Yes vote polled at 32%, that was a Ipsos/ Mori poll, now a panelbase poll has them at 40%, in the last 18 months the polls have shown a small 4% improvement for Yes Scotland, very unlikely that there would be a 8% rise in 3 weeks, would suggest the truth lies somewhere in-between,

[1320Lichtie]

3 weeks ago the Yes vote polled at 32%, that was a Ipsos/ Mori poll, now a panelbase poll has them at 40%, in the last 18 months the polls have shown a small 4% improvement for Yes Scotland, very unlikely that there would be a 8% rise in 3 weeks, would suggest the truth lies somewhere in-between,

No I think it shows what a lot of shite the polls actually are. Anyone that reads too much into them is being silly.

[vikingTON]

3 weeks ago the Yes vote polled at 32%, that was a Ipsos/ Mori poll, now a panelbase poll has them at 40%, in the last 18 months the polls have shown a small 4% improvement for Yes Scotland, very unlikely that there would be a 8% rise in 3 weeks, would suggest the truth lies somewhere in-between,

Nice try at conflating to different methodologies to cover up BitterTogether's fail: 1/10

[Lambie's Pigeon Feed]

Perhaps they just moved abroad and intend to move back, temporarily, for this referendum and haven't bothered to register until now? There's no real reason to suspect they are any more or less likely to vote one way or the other than "domiciled" voters.

Working class people are more likely to vote Yes,

You think someone so disengaged and disenfranchised by the main political parties in the UK is registering to vote for the first time to enshrine the Westminster rule that has failed them to the extent they can't even be fucked to vote Labour?

[ecto]

No I think it shows what a lot of shite the polls actually are. Anyone that reads too much into them is being silly.

Nice try at conflating to different methodologies to cover up BitterTogether's fail: 1/10

Yes Scotland love polls, and they read loads into them, in fact like to quote them, when I mean quote them I mean, big up the good ones, an as for the methodology argument, you could say that about all polls that are done, but on this forum only poor polls for Yes Scotland are dismissed on these grounds, just my view

Edited March 22, 2014 by ecto

[vikingTON]

Yes Scotland love polls, and they read loads into them, in fact like to quote them, when I mean quote them I mean, big up the good ones, an as for the methodology argument, you could say that about all polls that are done, but on this forum only poor polls for Yes Scotland are dismissed on these grounds, just my view

No, you tried to dismiss a swing as unrealistic, despite the fact that the two polls in question were from two different companies. They don't reflect a swing; the change of in house polling makes a swing. And it has been categorically towards Yes since the White Paper.

[ecto]

No, you tried to dismiss a swing as unrealistic, despite the fact that the two polls in question were from two different companies. They don't reflect a swing; the change of in house polling makes a swing. And it has been categorically towards Yes since the White Paper.

It has moved 2% on panelbase towards Yes 38% to 40%, No losing the 2%, where as in Ipsos/MORI Yes vote has lost 2% going from 34% to 32%, No staying steady at 57%, YouGov showing Yes only up 3%, 32% to 35%, No only losing 1% now at 52%, so what does this say?

Edited March 22, 2014 by ecto