Here, Sky Data explains how it carries out its work, whose opinions it seeks and how its polls are used.
Tables for all Sky Data's previous polls for Sky News can be found at the bottom of the page.
What is Sky Data?
Sky Data conducts nationally representative polls. We work primarily for Sky News, though other clients include the University of Oxford, London School of Economics and the Economic and Social Research Council. Sky Data is a member of the British Polling Council and abide by its rules.
Who are we polling?
We use the millions of Sky customers who have provided consent for us to contact them with regard to research as a form of research panel. We contact our respondents by email or SMS.
How do we make it nationally representative?
As with all pollsters, we select the sample and weight the data to be in line with the demographic profile of the UK - so that we have the right proportion of people of different genders, age groups, regions, social grades, etc.
Do Sky customers have different political attitudes to non-Sky customers?
No - when we control for demographics as above, Sky customers are no different to other members of the public.
We have conducted extensive testing of our results comparing with "gold standard" academic polls - the British Social Attitudes survey and the British Election Study - and our results compare favourably with other mainstream pollsters.
How can a sample of only a thousand people reflect the opinion of millions of Britons?
George Gallup - the man who pioneered modern opinion polling in the USA in the 1930s - used the analogy of tasting if a bowl of soup has enough salt.
If you have mixed it well, you can tell with one spoon - you don't need to eat the whole bowl.
In this analogy, the "mixing" is the sampling and weighting mentioned above - so long as the sample is representative of the overall population, statistical theory dictates that a poll of 1,000 people will be within 3% of the true result if you asked every single Briton 95% of the time.
What about those votes in news articles or on social media that allow anyone to click on them?
Those sort of polls boost engagement with stories and are fun to take part in - Sky News publishes them - but they do not reflect public opinion.
Not even if loads of people have taken part?
No. An example of this is the creation story of opinion polling - George Gallup taking on the then-influential magazine, the Literary Digest.
The Literary Digest polled a huge but unrepresentative sample of 2.4 million people ahead of the 1936 US Presidential election, taking every respondent they could get their hands on, and predicted a big win for Alf Landon.
Gallup polled a far smaller but representative sample, with quotas for the likes of gender, age and social grade, predicting a big win for Franklin D Roosevelt.
As you'll have guessed, Gallup was right - and the Literary Digest was extremely wrong.
Even a survey of millions will be way off if it has proportionally too many of some types of people and too few of others - and much smaller samples that do take that into account can be very accurate.
Why do some of your results not add up to 100%?
Often results will add up to 99% or 101%. This is due to rounding, for example if 49.5% said "yes" and 50.5% said "no", it would round to 50% and 51%.
Sometimes we will also show results excluding people who select options such as "don't know" or "prefer not to say" - we always publish a full breakdown of results online on the Sky News website and tweet them out from @SkyData.
Below you can find the tables for all Sky Data's past polling for Sky News.
Online polls
January 2020 - Boris Johnson's performance
October 2019 - Party leader satisfaction
03/04/19 - Other Brexit questions
27/03/19 - Other Brexit questions
January 2019 - State of the nation
January 2019 - Racism in football
November 2018 - Housing for IPPR
October 2018 - British views on Northern Ireland and Brexit
September 2018 - Conference season
June 2018 - Crime and police funding
July 2018 - Brexit, Corbyn anti-Semitism
May 2018 - Republic of Ireland
May 2018 - Nostalgia with Demos
March 2018 - Preferred leader and party on key issues, Russia
January 2018 - Trump, NHS & social care, plastics
December 2017 - Markle, Christmas
November 2017 - Universal Credit, plastics, intergenerational justice, sexual harassment
August 2017 - Brexit, racism, university
July 2017 - leaders and policies
February 2017 - Brexit, NHS & social care
Late November 2016 - plastics, communities, Brexit
Early November 2016 - Clinton & Trump, Budget
September 2016 - best leader and party on key issues
August 2016 - Corbyn vs Smith, refugees, electoral reform
April 2016 - Brexit, junior doctor strikes
January 2016 - EU renegotiations, junior doctors, gay clergy
December 2015 - Trump, Christmas, 2016 predictions
November 2015 - local services, Budget, climate change
October 2015 - tax credits, Corbyn, Putin, Assad, China
September 2015 - Corbyn policies, Syria, immigration, racism, trains
August 2015 - Labour leadership, Robots, Queen
July 2015 - economic growth and standards of living
June 2015 - Brexit, sex changes
May 2015 - Labour lessons, Brexit, Scottish nationalism
April/May 2015 - policies and leaders
SMS polls
European Parliament elections 05/04/19
Brexit national humiliation 20/03/19
Second Meaningful Vote 12/03/19
Queen Brexit comments - 25/01/19
Brexit crisis, confidence vote 15/01/19
Brexit 'break faith' - 17/12/18
Brexit 'PR exercise' - 11/12/18
Brexit worth being poorer - 28/11/18
International Men's Day - 19/11/18
Boris Johnson on burkas - 08/08/17
EU Referendum cheating - 17/07/18
Brexit post Chequers - 09/07/18
Social media bullying - 06/07/18
Brexit dividend for the NHS - 18/06/18
Prince Charles Commonwealth - 19/04/18
Post-Syria airstrikes - 16/04/18
Pre-Syria airstrikes - 11/04/18
Facebook data breach - 21/03/18
Social care and council tax - 08/02/18
Returning IS fighters - 01/11/17
Texting while driving - 24/10/17
No deal vs bad deal - 11/10/17
When should May resign - 07/10/17
May conference speech - 05/10/17
Royal children private school - 06/09/17
May resign good or bad for Conservatives - 30/08/17
Military intervention in Syria - 06/04/17
May Brexit trust, Scottish independence - 29/03/17
Care if Scotland independent - 22/03/17
Should there be indyref2 - 13/03/17
National insurance, early election - 08/03/17
Trump travel ban UK - 30/01/17
Legalising cannabis - 21/11/16
Trump impact UK & US - 10/11/16
Article 50 trigger without MPs - 03/11/16
Scottish independence - 12/10/16
Military intervention Aleppo - 11/10/16
Single market vs freedom of movement - 30/09/16
Junior doctor strikes - 02/09/16
Interest rates post-rise - 05/08/16
Banks vs under mattress - 03/08/16
Snap election, May vs Corbyn - 11/07/16
EU ref campaigns impact - 22/06/16
Republican referendum - 09/06/16
Economic turmoil vs immigration - 03/06/16
Obama EU post-visit - 25/04/16
Cameron Panama papers - 08/04/16
Nationalise Port Talbot steelworks - 01/04/16
Compassionate Conservatives? - 21/03/16
EU ref scaremongering - 07/03/16