It's been said before, but... the random sample + UNS really is dead.
The yougov-type massive panel that they know everything about then measure how opinions are swinging around modelling can potentially figure out the correlations between voters and project FPTP seat numbers with way more accuracy.
Comments
The yougov-type massive panel that they know everything about then measure how opinions are swinging around modelling can potentially figure out the correlations between voters and project FPTP seat numbers with way more accuracy.
'tis the future.
Related;
https://www.theregister.co.uk/2017/06/13/election_poll_accuracy/