Saturday, 07 June 2014 09:53
By Martin Kelly
Results of a newly published survey have revealed that the gap between those backing Scottish independence and those againts is just seven points.
The Populas poll, carried out on behalf of the Financial Times, asked people whether they wanted Scotland to remain part of the UK or leave and become independent.
Respondents were asked to choose between the following options:
- “I hope that Scotland votes to remain part of the UK.”
- “I hope that Scotland votes to leave the UK and become an independent country.”
- “I don’t have a strong view either way.”
Forty seven per cent backed remaining part of the UK, forty per cent opted to leave and become independent and thirteen per cent did not know.
When those who didn’t know were stripped out, the survey showed support for independence on 46% with those against on 54%.
Welcoming the result, SNP Business Convener Derek MacKay said:
“This is a very welcome poll which puts support for becoming an independent country at 46 per cent, excluding ‘don’t knows’ – a Yes vote is well within Scotland’s grasp.
“Today and every day between now and 18th September, Yes will be speaking to people on doorsteps and high streets about the gains off independence – the momentum is with Yes.”
The survey was carried out between 28th May – 6th and had a sample size 548 people. Whilst the numbers are statistically smaller than the optimum 1000, the Yes campaign will take further heart from the unusual formatting of the questions which included the phrases “Scotland votes to remain part of the UK” and “Scotland votes to leave the UK”.
The inclusion of such phrases is widely recognised as having an influence on the results of polls, in this case possibly inflating support for No whilst simultaneously reducing support for Yes.
Comments
2014-06-07 10:31
2014-06-07 10:51
“When Mr Mudie announced he was to retire as an MP, Labour leader Ed Miliband described him as a “fantastic MP” who would be a “huge loss” to the House of Commons.“
Pity he sits for Leeds East and is standing down at the 2015 election.
Also on the BBC this morning is Lego dropped from Treasury ‘Buzzfeed’ – see bbc.com/…/…
HMG have currently removed the images from their site and kept the lies. Pity is wasn’t the other way around. Perhaps they’re negotiating a deal with the Muppet franchise.
2014-06-07 11:02
2014-06-07 11:23
The sample here is pretty small. I defy anyone to use a survey of 500 people and use any methodology to accurately arrive at what the 3.2+ million who will vote in the referendum think.
What seems to be absolutely clear to me is that we are going to be in a position right up to the referendum where the 10-20% don’t knows (sometimes it is higher) will dictate the result of the referendum, meaning that the predictive value of all the polls is to a large degree negated by this significant unknown element. That this figure has remained largely resolute shows the opportunity YES has to convert people right up to 18.9.14.
2014-06-07 17:03
The referendum question is so straightforward it baffles me that so few of the polling companies use it verbatim.
They are afraid to use the simple question because a poll based on it gives higher values for YES than the muddy water Questions they actually ask.
A 500 poll sample gives results of +-4.5% range at the 95% confidence limit as opposed to +-3% for a 1000 sample.
James Kelly over at scotgoespop.blogspot.co.uk/ has it about right when he says:
Quote:
Five years from now, most Scots will regret it if they have voted to become an independent country : 49.9%
Five years from now, most Scots will regret it if they have voted to stay part of the UK : 50.1%
Which is tantamount to an Indy Ref result of 50.1% YES 49.9% NO.
And that is now before the big push by YES comes on stream.
2014-06-07 12:04
2014-06-07 12:17
2014-06-07 12:24
If a simple yes/no type question results in a 45:55% split in the response, then for a +/- 5% tolerance at the 95% confidence level you need a sample size (irrespective of population size) of 381; 4% 594, 3% 1056, 2% 2376 and 1% 9504. If that is correct, then this small sample size would have a tolerance of about 4.5 % points as against 3% on a more normal sample size of just over 1,000).
It would be interesting if anybody better qualified can corroborate that, and do please delete this post if it sounds erroneous.
2014-06-07 12:42
Derek Bateman worked in the Scottish media for the Scotsman, the Herald, The Observer, the Sunday Times and as an international programme maker for STV and for 20 years at the BBC reporting and presenting current affairs on radio and television.
Into Bookmarks-Favourites.
batemanbroadcasting.com/…/