Re-formatting the EURef polling table so it strips out the don’t knows
This should make comparison easier
The election is a month today and postal votes to overseas addresses are starting to be mailed out. To coincide I’m switching my regular polling table to election mode stripping out the don’t knows and just give the figures of those who had a voting intention.
Where a pollster has not provided this information I have set my spreadsheet so it calculates this itself.
Showing net numbers is, after all, how general election polls a presented and it allows us to compare different pollsters on the same basis. It also has the effect of inflating the Leave or Remain lead.
I’m expecting the number of published polls to increase as we get closer. This week there will be the first Survation phone survey for some time which it is to produce regularly in conjunction with the IG spread betting firm.
Since last Monday’s ICM online comparison survey all the other surveys have been pointing to IN leads.
My table list polls in chronological order of their fieldwork. This means that if new polls are published with a fieldwork date before the others in the table then it will not appear at the top. YouGov’s modal test fits into that category as well as the latest TNS poll.
The next online surveys could be interesting. Will, like YouGov and Opinium, show a narrowing of the gap with the phone polls?