,
3 tweets,
2 min read
Read on Twitter
That is not the only difference between the two companies' approaches to polling - @drjennings , Mark Pickup and I took a comprehensive look at pollsters' choices and their efects here:
sotonpolitics.org/2019/09/10/hou…
sotonpolitics.org/2019/09/10/hou…
@drjennings It is also unwise to focus on leads rather than shares as all polls are subject to error and the error on lead estimates is much bigger (because it is the difference between two estimates, both with error ranges)
@drjennings Turnout estimation is certainly once challenge, but so is false recall. There is suggestive evidence that many of those who voted Labour in 2017 now claim they did not. Correcting for that reduces Labour's estimated share. Suspect Owen will be less likely to focus on that issue.
