“58% of the sample disagreed, compared to 36% who agreed”

The Sinn Féin spokesman on education John O’Dowd’s response was “Opinion polls vary..”, but I’m not sure there have been that many other polls on this.. Anyway, the Belfast Telegraph reports that a poll of 876 adults in Northern Ireland, commissioned by the Conservative Party, carried out by YouGov, on post-primary education – the issue which “could define the executive” – produced the following results.

The main question was: “The Minister of Education Caitriona Ruane would like to see the current 11-plus system scrapped. To what extent do you agree or disagree with this proposal?” A total of 58% of the sample disagreed, compared to 36% who agreed.

Interestingly there’s also a break-down of those figures by political party

The poll — commissioned by the Conservative Party — shows that almost 80% of unionist voters, 36% of SDLP supporters and 34% of Sinn Fein followers are against the Minister of Education’s proposals, while Alliance voters have chalked up 75% opposition.

, , , ,

  • elvis parker

    Alliance voters are completely at odds with their voters! At least SDLP and SF can be said to have the majority of their own supporters on side.

  • Mark McGregor

    Seems a bit pointless reporting the Alliance breakdown, with a sample size of 876 and an expected Alliance representation of 5% you’d be looking at a margin of error of around +/- 15%. Even the other parties are going to be running between a 7-10% margin of error. The sample size just isn’t large enough for decent subgroup analysis.

  • willis


    “Alliance voters are completely at odds with their voters! ”

    I think I know what you mean.

  • joeCanuck


    I agree with your overall analysis regarding the poll size but, still, the total percentages are significant, I think.

  • nineteensixtyseven

    Interesting findings but I’d like to see a poll with a larger sample size before reading too much into things.

  • Continental Drifter

    Hold on, against which proposals?

    I’m an Alliance supporter, fairly content with Alliance policy, but would consider myself against the proposals on the basis that what happens in 2011 remains completely unclear.

  • Mark McGregor


    A bigger sample size isn’t going to get you much better than +/- 3% on the overall sample though it would make the subgroup data more relevant. Biggest problem is that polling in the 6 counties rarely reflects outcomes you can check against like elections.

  • Regarding the Alliance data the unweighted sample was 111 and the weighted sample 72. Therefore the Alliance voter data were more robust than other Parties.

    Jeffrey Peel
    Conservatives in Northern Ireland

  • Mark McGregor


    Thanks for the figures but regardless of weighting the subgroup sample sizing is so small it is always going to return a +/- 10-13% margin of error in the case of APNI and slightly less bad but still pretty poor levels for other subgroups.

  • Mark McGregor


    Did you have to carry out similar levels of weighting to the other subgroups? – having to weight any subgrouping as you’ve done at a level of around 35% should indicate your sampling method wasn’t particularly accurate and might call the entire survey into doubt as an accurate or reflective tool.

  • The survey was carried out by YouGov and, with a sample size of 876, is very robust. YouGov has a great reputation for producing very accurate polls. It was the only polling organisation to predict the result of the London mayoral elections, for example.

    Alliance voters were slightly over-sampled but DUP, SDLP and UUP samples were pretty spot-on. Sinn Fein voters were slightly under-sampled based on votes cast at the Assembly elections.

    We will be releasing further results from the survey in the coming weeks. Owen Paterson also quoted some other results from the survey in his article that appeared in the Belfast Telegraph last week:


  • spiritof07

    Big John wheeled out again to defend Ruane. If Martina Anderson really is being primed to take over when will she get a piece of the PR action on this one?

  • Reader

    YouGov: A total of 58% of the sample disagreed, compared to 36% who agreed.
    Amazing. It used to be that there was a majority in favour of selection but against the 11+. How did Ruane manage to make the 11+ into the least worst option?

  • billie-Joe Remarkable

    The “decent” people have spoken. Grammar schools for their kids (and an over-sized people carrier or 4×4 for the school-run). The lifestyle of Northern Ireland’s aspirant middle classes is “not up for negotiation”, to borrow a phrase.

  • Mark McGregor


    If YouGov is so respected, while having no record of polling in the notoriously impossible to poll accurately 6 counties, it shouldn’t be a problem to give the sampling method, raw data, weighting and subgroup stats including the demographics.

    Unless of course we have another poll that is showing a bias towards Unionism and against SF that is never actually translated into the mass sampling of an election/electorate.

    I’d refer you to the oft used NILT polls that never, ever get close to reflecting the reality of the electorate.

  • Mark,

    The survey was conducted using YouGov’s UK panel. The panel has some 200,000 members in every part of the UK, including Northern Ireland. Our sample was drawn from this panel.

    Furthern information about the YouGov panel can be found at:


    Benchmark questions in our survey indicate that the survey findings are very accurate e.g. voting at the last Assembly elections.

    Jeff Peel

  • Dev

    I used to work for a company that supplies YouGov with their sample (the respondents to their surveys). From recollection, there was at best about 450 Northern Irish people signed up to partake in such surveys, other sample providers had smaller numbers for NI than ours. The response rate in the UK for online surveys is about 5% (occasionally gets up to 10%), so of that 450 only a tiny fraction wld actually complete any survey they were invited to take. I would be amazed if YouGov managed to get 800+ respondents even if pooling all the resources of several sample providers, and if they did the sample certainly would not be in any way representative. In short, ths survey is a complete load of balls.

  • Dewi

    An interesting question is why did the Tories commission this poll?

  • Mark McGregor


    Thanks, so its a pretty worthless poll being self-selecting rather than random sampling. I see by a quick google that their methodology has been questioned numerous times.

    File under b for bin.

    They even pay people for responding – pffttt.

  • joeCanuck

    They even pay people for responding

    Who are they, Mark? I’ve got time on my hands.

  • Mark McGregor


    YouGov, its in Jeffrey’s link, they pay preregistered people to complete large numbers of surveys – you only get money after carrying out at least 50 polls.