Quantum Errors…
Well, last weeks RedC poll was pulled (understandably) as a result of the political upheavals making it pointless. While some ppl had already tweeted about being polled, it made little sense to continue spending money on a poll that wouldn’t have much value.
This appears not to have bothered the Sindo, who have ‘commissioned’ another poll from Quantum ‘Research’. Most of you will be aware of my reservations regarding QR ‘polls’….
First off, there was someone on politics.ie claiming to be one of those polled, so it’s possible there was a poll of sorts actually conducted (although that doesn’t prove anything). However, with only 200 respondents claimed, the m.o.e. would be over 6% even if true, so it’s credibility is poor before you even look at the results….
And what are they? Well, not quite as mad as I teased here but still not too easy to believe. Excluding Don’t Knows (of only 14%, surprisingly) the figures are as follows (spreadsheet seat projections in brackets)
FF 9.3% (5 seats)
FG 31.4% (62 seats)
LP 33.7% (69 seats)
SF 9.3% (9 seats)
GP 3.5% (2 seats)
OTH 12.8% (19 seats)
Where to start? Well FF are about two-thirds of their previous record low, and SF who were tied with them in that poll are tied with them in this also. FF down 5% and SF down by the same in the same poll? Can’t see that…..
Also LP are ahead of FG. Now while RedC have been over-stating FG relative to other companies, and under-stating LP by the same measure, it’ hardly credible that LP could be 15% behind FG in one and over 2% ahead of them in another credible company. Particularly hard to believe is that FG would be about 4% lower in a poll that FF (and SF) are also 5% lower.
No doubt they’ve their reasons to publish this nonsense …. don’t be surprised it a future ‘poll’ they commission shows a ‘swing’ that they make a lot of. In the meantime, let’s wait for the next post-Martin/Lenhen/Hanafin/O’Cuiv poll from MRBI, RedC or Lansdowne.
Irish Polling Report 
Does your method make it possible to do an upper and lower estimate on the total number of seats using the increased margin of error?
(I’m not not up on the stats of this, so the next bit may be pure shite!)
If the m.o.e. is 6%, and FF is on 9.3%, then does it make sense to say the upper estimate would be 15.3%? and if so, how many seats might that give them?
Tomboktu
January 26, 2011 at 1:19 am
It’s be very difficult as every projection needs a figure for all the parties, so if I had e.g. FF on 15.3% I’d have to decide who to take it off.
I’d say we’ll have new figures soon enough anyways – and must as I’d like to think LP are ahead of FG, there’s little basis to take this poll seriously. Next RedC or MRBI will be interesting, but even that will be skewed by the immediate ‘honeymoon period’ for Martin.
Dotski
January 26, 2011 at 8:15 pm
Hi Dotski, can’t see another way to contact you, I was just wondering if you could help me out with something. Do you know what the electoral numbers are for the new constituencey of Limerick City? In comparison to the old “Limerick East” especially. I’m trying to get a feel for what people (like the FG candidates) will be losing with the loss of that bit of countryside.
HandsofBlue
January 26, 2011 at 8:30 pm
I’ve not got box specific numbers, I just have the figures as to what is moving in terms of population and general information that it’s good territory for FG (Kieran O’Donnell being particularly hit by it)
Dotski
January 26, 2011 at 9:10 pm
So, the electorate for LE was just under 77’000 last time, any idea what it is roughly now? Like, what percentgae of the population has been split? Any idea? I’m having trouble finding any solid info elsewhere.
(I was stunned O’ Donnell didn’t try to get into “County Limerick” this time. He’s really up against it in the city. I mean, the two Labour candidates are former mayors. What chance doese have?)
HandsofBlue
January 26, 2011 at 9:57 pm
.FF down 5% and SF down by the same in the same poll? Can’t see that….?
mp3lemon
January 29, 2011 at 2:22 pm