This article is from the source 'guardian' and was first published or seen on . It last changed over 40 days ago and won't be checked again for changes.

You can find the current article at its original source at https://www.theguardian.com/politics/2017/jun/08/seen-the-poll-showing-a-labour-lead-its-worth-checking-the-methodology

The article has changed 6 times. There is an RSS feed of changes available.

Version 0 Version 1
Seen the poll showing a Labour lead? It's worth checking the methodology Seen the poll that shows a Labour victory? It's worth checking the methodology
(about 5 hours later)
As the people of Britain cast their votes in the general election, a poll putting Jeremy Corbyn’s Labour party in the lead for the first time is being widely shared on social media. Published by Wired on Wednesday, it gives Labour a 2.8 percentage point lead over the Conservatives.As the people of Britain cast their votes in the general election, a poll putting Jeremy Corbyn’s Labour party in the lead for the first time is being widely shared on social media. Published by Wired on Wednesday, it gives Labour a 2.8 percentage point lead over the Conservatives.
However, it is worth exercising some caution around the figures. The company that carried out the survey is not part of the British Polling Council. It is an ad-tech firm called Qriously, without a specific history of political polling.However, it is worth exercising some caution around the figures. The company that carried out the survey is not part of the British Polling Council. It is an ad-tech firm called Qriously, without a specific history of political polling.
The company told Wired that since the Labour lead is within its 3.2 point margin of error, it believes the election is still too close to call.The company told Wired that since the Labour lead is within its 3.2 point margin of error, it believes the election is still too close to call.
Asked about the sample size and methodology, Qriously said the data was collected via mobile phone between 4 and 7 June, and it surveyed 2,213 UK adults, including 1,905 registered voters and 1,279 likely voters. It claims to use machine learning to turn that into a representative sample. The findings gave an outcome of 41.3% to Labour and 38.5% for the Tories.Asked about the sample size and methodology, Qriously said the data was collected via mobile phone between 4 and 7 June, and it surveyed 2,213 UK adults, including 1,905 registered voters and 1,279 likely voters. It claims to use machine learning to turn that into a representative sample. The findings gave an outcome of 41.3% to Labour and 38.5% for the Tories.
So this isn’t a face-to-face or telephone survey, as conducted by other pollsters. Essentially, the company uses algorithms to buy advertising space within mobile phone apps, and shows a question instead of a banner ad. So the figure putting Corbyn in the lead is in effect based on a self-selecting sample doing an online poll.So this isn’t a face-to-face or telephone survey, as conducted by other pollsters. Essentially, the company uses algorithms to buy advertising space within mobile phone apps, and shows a question instead of a banner ad. So the figure putting Corbyn in the lead is in effect based on a self-selecting sample doing an online poll.
For comparison, the Guardian’s poll tracker, which has been following the results from several polling companies, puts the Conservatives ahead with 43.91%, and Labour second on 35.93%.For comparison, the Guardian’s poll tracker, which has been following the results from several polling companies, puts the Conservatives ahead with 43.91%, and Labour second on 35.93%.
Of the surveys conducted in the last few days of the campaign, Survation and SurveyMonkey show the narrowest leads for the Conservatives over Labour, at +1 and +5 respectively. ICM shows the biggest gap, placing the Tories 10 points ahead. But all the surveys make Theresa May’s Conservatives the favourites to win the election.Of the surveys conducted in the last few days of the campaign, Survation and SurveyMonkey show the narrowest leads for the Conservatives over Labour, at +1 and +5 respectively. ICM shows the biggest gap, placing the Tories 10 points ahead. But all the surveys make Theresa May’s Conservatives the favourites to win the election.
The chief executive of Qriously, Christopher Kahler, told the Guardian: “There seem to be some questions around the representativeness – we’re finding that while there is, of course, a bias in the smartphone population, so many people have smartphones that there is a representative subset within it.The chief executive of Qriously, Christopher Kahler, told the Guardian: “There seem to be some questions around the representativeness – we’re finding that while there is, of course, a bias in the smartphone population, so many people have smartphones that there is a representative subset within it.
“Every method has a bias; it’s a question of which has the least. Our bet is that in the long run, random smartphone users will be more representative of the general population than relatively small online panellists who are paid to answer questions, and certainly than people who only have landlines. It’s early days, but we believe this direction makes the most sense.”“Every method has a bias; it’s a question of which has the least. Our bet is that in the long run, random smartphone users will be more representative of the general population than relatively small online panellists who are paid to answer questions, and certainly than people who only have landlines. It’s early days, but we believe this direction makes the most sense.”
After failing to correctly predict the results of the 2015 general election, the EU referendum and November’s US presidential election, polling companies are under increasing pressure to get the result of this election right, and prove they can be trusted as a barometer of public opinion.After failing to correctly predict the results of the 2015 general election, the EU referendum and November’s US presidential election, polling companies are under increasing pressure to get the result of this election right, and prove they can be trusted as a barometer of public opinion.