Scotland politics

Scotland poll tracker methodology

Opinion polling can be a nuanced business. Here are some key things to bear in mind when reading the polls.

Margins of error

No poll can be 100% correct 100% of the time. Polling companies generally claim that 95% of the time, a poll of 1,000 people will be accurate within a margin of error of +/-3%.

This means that a figure in the poll could be up to three percentage points higher or lower than that shown.

So in referendum polling if "Yes" responses said 32% and "No" came in at 38%, there is a chance they could both be 35%.

It is, however, more likely that the figures will be 1% out rather than 3%.

Weighting

Another issue is how to ensure the sample is representative of the general population. To achieve this, polling companies "weight" their data to match the demographic profile of Scotland.

At its most basic level, this means that if a poll of 1,000 people is made up of 550 men and 450 women, it is unrepresentative because it does not reflect the profile of Scotland (48% male, 52% female)

So the answers of female respondents will be given slightly more weight (in this case they will each count as 1.133 people) to give them a representative impact on the final findings.

Conversely, the men will be weighted to each count as 0.891 people.

The same procedure is routinely carried out for age group, social class and region.

Many polling companies - including Ipsos-Mori and TNS-BMRB - also filter results by asking questions within polls on respondents' certainty to vote.

Polls included

The poll tracker includes all companies performing regular opinion surveys on Scottish independence which appear in major UK media outlets.

It does not include polls conducted on the behalf of any political parties, or for websites with explicitly stated pro or anti-independence positions.

We will only include polls based on the revised referendum wording from January 2013: "Should Scotland be an independent country?"

Our sample period begins in September 2013 and will run right up just before polling day on 18 September 2014.

The date of each poll in the poll tracker is usually the last date in which fieldwork was carried out.

You can read more about the companies' various methods below.

YouGov

Question: "If there was a referendum tomorrow on Scotland's future and this was the question, how would you vote? 'Should Scotland be an independent country?'"

Sample size: 1,000 Scottish adults.

Interview method: YouGov conducts its polling online from a panel of respondents recruited using standard advertising. Respondents are given a username and password and can only respond to each poll once.

Margin of error: +/- 3%

Weighting: Completed survey data is statistically weighted to the national profile of all adults aged 18+ (including people without internet access). YouGov weights by age, gender, social class, region, party identity and the readership of individual newspapers.

YouGov methodology in detail

Survation

Question: "In the referendum, voters will be asked, 'Should Scotland be an independent country'. If this referendum were held today, do you think you would vote 'Yes' or 'No'?"

Sample size: 1,000 Scottish adults.

Interview method: Online interviews, telephone and face-to-face.

Margin of error: +/- 3%

Weighting: Data is weighted to the profile of all adults aged 18+ and by age, sex and region.

Survation methodology in detail

TNS-BMRB

Question: "There will be a referendum on Scottish Independence on the 18th of September 2014. How do you intend to vote in response to the question: 'Should Scotland be an independent country?'"

Same size: 1,000 Scottish adults (16+).

Interview method: Face-to-face "computer-assisted personal interviewing" in respondents' homes.

Margin of error: +/- 3%

Weighting: Data is weighted to the profile of all adults aged 16+, based on population profile estimates from the BARB (Broadcasters' Audience Research Board) Establishment Report 2011, and mid year population estimates from the 2011 Census.

Weighting is carried out in relation to working status within gender, age, social grade and electoral region.

Data is also weighted by recalled vote to match turnout and share of constituency vote from the 2011 Scottish Parliamentary election.

More information on Scotland polling from TNS-BMRB

ICM

Question: "In the referendum on independence for Scotland on 18th September 2014 voters will be asked, 'Should Scotland be an independent country'. Do you think you will vote Yes or No?"

Sample size: Approx. 1,000 adults aged 18 and over

Method: Online poll. Participants are a demographically representative sample of Scots, in proportion to population distribution and to specific demographics including gender and age.

Margin of error: +/- 3.2%

Weighting: Data weighted to the profile of all Scottish adults aged 16+. The results were weighted by sex, age, social grade and region. Targets were derived from 2011 Census. A question on how the respondents voted in the 2011 Scottish Parliament election was weighted for past votes.

Ipsos-Mori

Question: "As you may know, there will be a referendum on Scotland's constitutional future in September 2014. If the referendum was held tomorrow, how would you vote in response to the following question: 'Should Scotland be an independent country?'"

Sample size: Approximately 1,000 adults (16+)

Method: Telephone, using computer-assisted telephone interviewing. To ensure the achieved sample is broadly representative of the Scottish adult population, sample quotas are set on age, sex, working status and region.

Margin of error: +/- 3%

Weighting: Data is weighted by: age, sex and working status (all using census data); tenure (using Scottish Household Survey data) and public-private sector employment (using Scottish Government Quarterly Public Sector Employment series data).

Panelbase

Question: "Should Scotland be an independent country?"

Sample size: Approximately 1,000 Scottish adults aged 16+

Method: Online poll

Margin of error: +/- 3%

Weighting: All the results were weighted by age, sex, household tenure and socio-economic group, generally determined by the occupation of the chief income earner in each household. Responses about voting in the 2011 Holyrood election were weighted back to the actual election result. Referendum voting intentions were filtered on the assumption that eight out of 10 people were likely to vote in the referendum.

Sampling and weighting are generally done in line with census data and also the National Readership Survey, which covers over 250 of Britain's major newspapers and magazines, showing the size and nature of the audiences they achieve.