Polling is in fact an incredibly complex subject – the recent US Elections demonstrated this. You would have thought that going and asking someone what they think was straightforward. Far from it.
This “poll“, which is about as flawed as a poll possibly could be (but no doubt innocently so) is a perfect demonstration of this – but the same warning applies to any poll or survey. I have no idea who carried it out, and actually the pollsters themselves did a perfectly good job – but the questions they were asked to pose caused the entire poll to be utterly redundant. Rubbish in, rubbish out.
Firstly, you need to ask questions which are in fact viable. This poll, whose questions were apparently set by the “Conservative Party” (but I have no idea by whom or by which section of it) and encompassed “the east of Northern Ireland” asked:
Do you believe there should be an official opposition at Stormont, similar to Westminster or the Dail [sic]?
This is not a valid question. Northern Ireland has compulsory power-sharing because it is fundamentally socially divided along sectarian lines, so a straightforward majority/minority government/opposition system (which is what they have at “Westminster or the Dail”) is simply not an option.
Not only is the question illegitimate, but it also asks questions of the knowledge of the people who asked it – since it is an obvious impossibility to anyone with any knowledge and understanding of Northern Ireland. Do not trust evidence presented by people who do not understand the very people who are the target of the survey!
Another question asks: Several highly respected economic reports state that reducing Corporation Tax from 24% to about 12.5% will attract significant local and international business investment into NI and will create tens of thousands of jobs to [sic] NI over the next two decades. Would you accept a 2% reduction per annum in NI spending to pay for this?
This is the classic leading question. Firstly, there are of course other reports that suggest it would make no difference; secondly, the “2% per annum” is not defined (is it year-on-year or just once) and nor is “NI spending”; thirdly, it is not clarified as to whether this move would apply only to NI, or what the difference would be if it were at the same time applied, say, to Wales. Also, we do not in fact know what the actual cost would be, nor do we know what the political ramifications will be; and, importantly, we do not know if the move in NI would be matched by a similar move, say, in Wales, and what difference that would make.
So, rephrased, the question could read: “Several highly respected academic reports cast doubt on the assumption that reducing Corporation Tax from 24% to 12.5% would create any jobs at all and that it is, at best, a risk. Would you accept cutting public spending by the amount needed to pay for at least 8,000 nurses in order to take this risk?” – an equally legitimate question, but one which obviously would get a different answer!
Do not trust the answers to leading questions.
Another question asks: Do you think that the current way Northern Ireland’s First Minister and deputy [sic] determined [sic] should be carried out as per the original Belfast agreement in 1998, were [sic] the First Minister would come from the largest community (i.e Unionist or Nationalist), and not from the largest political party?
What on earth does this even mean?!
It is factually incorrect anyway. The “original agreement” had the First and deputy First Minister elected by the Assembly via cross-community vote.
It was in fact the St Andrews Agreement in 2006, not the “original agreement”, which determined that the largest party in the largest community designation would get the First Minister. A minor amendment to the subsequent legislation, however, essentially removed “in the largest community designation” and automatically gives it to the largest party (provided, effectively, that it fall within one of the two largest designations).
So, do not trust polls which are factually incorrect (far less still if they are grammatically incoherent).
The poll also claims: 68% of Catholics, if they lived in Great Britain, would vote Conservative.
Remind me, which party set the questions for the poll?!
Beware outliers – and ignore outright nonsense! Most of all, if you ask daft questions, don’t be surprised if you get daft answers.