Instead of aggregated results tables, we’ve decided to share the raw non-identifiable data of our election results so that anybody can download it and play around. We hope this will encourage more transparency and openness around polling data.
How to recreate our results:
We considered only registered voters (“Yes” to RegisteredtoVoteGeneralElection), and then only voters who give a “10 – Very likely” to LikelihoodtoVote. The data for VOTE-INTENTION then gives our baseline numbers. You can remove “I’m still undecided” and “I wouldn’t vote” to reach our final call.
You can download the raw data here: http://qrious.ly/GEUK2017
To weight the data and make it representative, use the WEIGHT column, which adjusts the data to match census / government data on gender, age, region and income (Link to weights HERE).
Sidenote: If you are looking to join a growing team and love to work with data, we’re looking for a Data Scientist and a Mobile Fieldwork Project Manager to join us in London! If this sounds like you, we want to hear from you! Below are the links to each job description.