Tuesday 4 May 2010

That Belfast Telegraph poll

The Belfast Telegraph has put its flawed poll on the web. The poll, published today, but based on interviews conducted between 22-29 April, is out of date and thus quite pointless. Additionally the sample size was small – 3,200 people across Northern Ireland but less than 200 per constituency.

For what it's worth, though, the poll has the following 'results':

Overall – DUP 26%, Sinn Féin 25%, SDLP 17%, UCUNF 13%, Alliance 7%, independents 6%, TUV 5%, 1%.

Fermanagh and South Tyrone:

The Belfast Telegraph has Michelle Gildernew (44%) slightly ahead of Rodney Connor (43%).

Upper Bann:

John O'Dowd is within 4% of David Simpson.

South Down: Margaret Ritchie (43%) is well ahead of Caitriona Ruane (26%).

East Belfast: an easy win for Peter Robinson (42%) ahead of second-placed Ringland (26%).

North Antrim: Paisley 39%, Allister 31%

South Antrim: McCrea 36%, Empey 30%

North Down: an easy win for Sylvia Hermon.

The rest of the seats are reckoned to stay with their incumbents, without any shocks. In fact, if this poll is correct, all the seats will remain in their current party hands. However, to repeat – this is a week-old poll, and based on fairly loose methodology, so its resemblance to reality on Thursday is not likely to be perfect. It disagrees with the evidence of the bookies – who may also not be correct, but their 'results' are what people are prepared to stake their own money on!


menaiblog said...

Two quick points:

Northern Ireland polls are notoriously inaccurate & have a history of under estimating the vote of the so called extreme parties.

A sample of 200 leaves you with a margin of error of 7% - too high to be of any use in predicting a number of the constituencies you name.

Paddy Matthews said...

The poll, published today, but based on interviews conducted between 22-29 April, is out of date and thus quite pointless.

"Out of date"? In Northern Ireland?

Horseman, a lot of voting intentions are based on things that happened in 1994, 1981, 1921, 1912, or 1690. They're not going to change too radically over the course of four days :)

You can expect Sinn Féin to be hammering the FST figures for all they're worth and more for the next two days.

bangordub said...

Still, Interesting stuff.
I have a double bet on Michelle in FST amd Gerry in NB.
Don't forget that the Shinners support is notoriously underestimated in polls

hoboroad said...

O'Dowd is just 4% behind Simpson if only the TUV had put up a candidate in Upper Bann. I see UCUNF are trying to rubbish the poll. They say the guy who runs the polling company is a good friend of Jeffrey Donaldson and followed him out of the UUP into the DUP.

Paddy Matthews said...


Don't forget that the Shinners support is notoriously underestimated in polls

The headline figures actually look reasonable enough for all parties by the standard of NI polls. I'd be less certain about the constituency-level samples but most of them don't seem wildly out of line with what might be expected either.


O'Dowd in Upper Bann looks a bit too high to me, but doubtless the DUP will be milking the figures over the next two days as well.

Owen Polley was certainly trying to get some pre-emptive retaliation yesterday. His prediction that UCUNF could outpoll Allister in North Antrim isn't looking too good on the BelTel figures (Allister 31%, Armstrong, er, em, 4%).

hoboroad said...

Kenneth Clarke became the first shadow minister to talk openly about a possible deal.
“I think there’s a still a slim chance we can get an overall majority, which I would very much like to see,” the former chancellor said.
“It is very difficult to read because the electoral geography is quite local and it is a complicated breakdown of voting which is taking place. But my hunch is we’re still in with a chance of getting a parliamentary majority.
“In the end you can always do a deal with an Ulsterman, but it’s not the way to run a modern, sophisticated society.”

hoboroad said...


Anonymous said...

Can't understand why the communities split their votes. Dumb, Dumb, dumb.

Garrett Carr said...

"Electoral geography" http://newmapsofulster.blogspot.com/2010/04/election-maps.html