The Methods of Qriously and Number Cruncher Politics

What methods did these two companies use?

Previously, I wrote a summary of the different methods used by ten market and social research companies.

This article focuses on the two companies who conducted a single poll each during the 2019 election campaign: Qriously and Number Cruncher Politics.

Two companies

This section looks at the different methods used by the two companies.


Survey mode: Internet river (ads in mobile applications)
Target population: United Kingdom (all adults)
Weightings: Age, gender, region, constituency type, and education level.
Turnout model: Self-reported turnout likelihood of 9/10 or 10/10.
Vote intention question:

If the election were held today, who would you vote for in your local constituency?

Respondents are offered the Scottish National Party and Plaid Cymru, if they live in Scotland or Wales, respectively. If the Brexit Party is not standing in the constituency, people are additionally asked:

If the Brexit Party does not stand in your constituency, who would you vote for?

Those who are not eligible or not registered to vote are removed.

Number Cruncher Politics

Survey mode: Internet river and internet panels
Target population: Great Britain (eligible voters)
Weightings: age, gender, education, ethnicity, and region.
Turnout model: Turnout weighting takes into account both stated likelihood to vote on a 0–10 scale (or, when applicable, a voter’s declaration that they have already voted by post) and the empirical relationship between stated and actual likelihood to vote, to assign a turnout probability.
Vote intention question:

If you did vote, which party or candidate do you think you would vote for?

Only those are say they are eligible to vote and expressed a vote intention are included in the headline estimate, weighted by turnout probability.

Internet river sampling

An internet river sample is gathered via intercepting advertisements on specific websites.

Respondents are invited to participate in a survey. Once agreed, people answer a few qualifying questions, and then routed to a survey.

This sampling method draws respondents like fish from the internet ‘river’, and then (usually) throwing them back. Sometimes, at the end of a ‘river’ survey, respondents may be invited to participate in an internet panel.

This is not to be confused with River Broadband.

Like internet panels, survey research costs are low as no interviewers are involved — the surveys are self-administered.

Quality of internet samples may be an issue.

The UK and GB vote shares are slightly different. (Image: Google Sheets)

However, both Qriously and Number Cruncher Politics estimated vote intention close to the actual results. In the UK, the Conservative lead was 11.5 points. Qriously slightly overestimated this lead.

Whilst Number Cruncher Politics is not a member of the British Polling Council, it had the third lowest five-party error among the 12 companies.

This blog looks at the use of statistics in Britain and beyond. It is written by RSS Statistical Ambassador and Chartered Statistician @anthonybmasters.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store