The irony of the press criticising Michael Gove's dodgy surveys

Most PR-commissioned surveys are bunk – but it's not just Michael Gove who cites them.

Education secretary Michael Gove was roundly – and correctly – ridiculed yesterday after it emerged his criticisms of shoolchildren’s English history knowledge was based on PR research carried out by Premier Inn and UKTV Gold.

While the irony of a man looking to improve educational standards in history being himself unable to spot a compromised source is one which hasn’t escaped the attention of commentators, oddly little spotlight has been shone on the source of the dodgy data itself. Ludicrous as it is that a senior politician in a position of power could be so embarrassingly unfamiliar with acceptable standards of research, it seems remiss not to question why marketing-led data ended up in a position to be taken seriously in the first place.

The reason no questions have been raised might conceivably be due to a degree of embarrassment, given that the primary conduit for PR puff-pieces is the mainstream media itself. Take the Telegraph, who pulled no punches in deconstructing Gove’s embarrassing error - despite themselves having recently published "research" from B&Q highlighting the annoyance of neighbours mowing their lawns in the morning.

Similarly, the BBC, who also ran an account of Gove’s cock-up, only last week published a press release from pen-manufacturers Bic declaring – in the run up to exam season, naturally – that students should rely less on their lucky underwear, and instead follow advice to revise diligently… and "pack a spare pen".

Perhaps most red-faced of all ought to be the Independent – who reported merely a week ago that we lazy Brits don’t do enough walking (from research commissioned by the Ramblers Association). The paper reported the hapless Education Secretary’s reliance on PR surveys... despite having themselves published one of the very PR surveys they criticise Gove for citing. Perhaps someone at the Independent needs to be reminded what they say about having your cake and eating it.

The Daily Mail, wisely, have yet to criticise Gove – smart, given that they previously published each of the PR surveys he cited, and especially considering that in the last week alone they’ve published more than twenty two articles derived from the same questionable opinion poll mould. While the error from the Education Secretary was cringeworthy and troubling, there’s a degree of hypocrisy in the news media criticising Gove for believing what they presumably, at the time, felt was worthwhile enough to print.

The ease with which PR surveys pass into the mainstream media – and into the eager hands of Education Secretaries, it seems – would be less concerning if the research they presented was of a high standard. Unfortunately, for the majority of such opinion polls, commercial interest and crippling methodological flaws often render the results worthless.

Take, for example, three of the surveys cited by Gove, whose original press releases can be easily located: all three polls were conducted by polling company OnePoll – a prolific operator in the online opinion poll industry. It was OnePoll who recently informed us that women hate their bodies more than ever (in a survey sponsored by a dieting company), that women can’t manage money as well as men (according to a cashback website) and that men should be less lazy when it comes to doing DIY (says an entirely-impartial DIY retailer). There’s no shortage of OnePoll surveys in the news each week, yet, on examination, serious questions can be raised over their gathering of data.

OnePoll claim to have a community of over 100,000 users, with users paid around ten pence for each survey completed - with surveys regularly consisting of over a dozen questions. As moneyspinners go, it’s hardly lucrative. What’s more, users can only collect their earnings after accruing £40 - equating to roughly three hundred surveys and potentially hundreds of hours.

This low return on investment becomes an incentive to undertake as many polls as possible – and with limits to the number of users able to complete each survey, users are soon tempted to spend no time at all thinking honestly about their answers. If only the first 2,000 people earn 10p for their ten minutes of work, why be the person who lingers too long and finds they’ve wasted their time for no reward? This, combined with the ease of second-guessing the "desired" answer to screening questions (why yes, I DO have children under the age of seven…), almost certainly results in some polls being completed by users with no relationship to the subject matter and no concern for the answers they provide.

The problems don’t stop there - as any psychology student will tell you, any answer you receive can depend entirely on the question you ask and the way you ask it. Take, for instance, a question posed by OnePoll to new mothers:

Do you find you don’t actually care as much about your appearance now that you have had a baby?

A) Yes absolutely

B) No, I care, I just don’t have the time to do anything about it

C) Neither

(OnePoll, 13 May 2013, survey NH SDG 3004 VBM)

Once you pick through the positives and negatives, it’s clear that both of the first two options presume new mothers don’t take care of their appearance. While this may or may not be true, it’s not a finding which the question solicits – yet it would be easy for a PR firm (perhaps acting on behalf of a cosmetics or clothing retailer) to spin an article just as well from either response.

Similarly, multiple-choice questions can often force users into making fairly clumsy sweeping generalisations - such as the question posed to male users earlier this year:

Please be brutally honest - whose cooking do you think is better, your mum’s or your wife’s/partner’s?

(OnePoll, 1 March 2013, survey title unknown)

 

Heaven help the men faced with that choice. Yet, with a shiny 10p on the line, and no other option available, a choice had to be made… and thus, as the Daily Mail declared a fortnight later, “Over half of British men think their mums are better cooks than their partners”. Somewhat lost in the write-up is that the men in question were forced into the choice, and only marginally more than half opted for the first option on the list.

The lack of transparency of survey method and availability of actual data in the final write-up is a further issue with PR polls, and brings us neatly back to those cited by Gove. Take the finding – highlighted as the very headline of both the original press release and the Daily Mail article – that Delia Smith was married to Henry VIII: nowhere is it reported how many of the 2,000 children to take the poll actually gave this as a response. It’s perfectly plausible – if not highly likely – that the main line of the news coverage of this story was generated by only a handful of respondents choosing this rather silly option. It’s equally likely, too, that the survey wasn’t of 2,000 11-16 year olds, but of 2,000 people who said they’d had their children present, some of which may not have been telling the truth. Those screening questions are, after all, not rocket science to figure out.

Most intriguing of all, then, is the possibility that the Education Secretary criticised the ignorance of children in this country based on number of newspaper reports of a minority of errant responses given to a PR opinion poll by people who weren’t children at all. Until there’s real transparency on the data, and better control over survey methodology, there’s no real way of knowing. This survey isn’t unique in that respect – there are dozens of similar cases featured in the news each week, often with significant potential flaws.

As for Michael Gove, he ought to be more cautious about believing what he reads in the newspapers, and should learn to be more skeptical of his sources. Those who don’t learn history are doomed to ignorantly repeat things.

Michael Marshall, Vice President of the Merseyside Skeptics Society, regularly writes and lectures on the role of PR in the tabloid news. He tweets as @MrMMarsh.

Photo: Getty
Show Hide image

After Richmond Park, Labour MPs are haunted by a familiar ghost

Labour MPs in big cities fear the Liberal Democrats, while in the north, they fear Ukip. 

The Liberal Democrats’ victory in Richmond Park has Conservatives nervous, and rightly so. Not only did Sarah Olney take the votes of soft Conservatives who backed a Remain vote on 23 June, she also benefited from tactical voting from Labour voters.

Although Richmond Park is the fifth most pro-Remain constituency won by a Conservative at the 2015 election, the more significant number – for the Liberal Democrats at least – is 15: that’s the number of Tory-held seats they could win if they reduced the Labour vote by the same amount they managed in Richmond Park.

The Tories have two Brexit headaches, electorally speaking. The first is the direct loss of voters who backed David Cameron in 2015 and a Remain vote in 2016 to the Liberal Democrats. The second is that Brexit appears to have made Liberal Democrat candidates palatable to Labour voters who backed the party as the anti-Conservative option in seats where Labour is generally weak from 1992 to 2010, but stayed at home or voted Labour in 2015.

Although local council by-elections are not as dramatic as parliamentary ones, they offer clues as to how national elections may play out, and it’s worth noting that Richmond Park wasn’t the only place where the Liberal Democrats saw a dramatic surge in the party’s fortunes. They also made a dramatic gain in Chichester, which voted to leave.

(That’s the other factor to remember in the “Leave/Remain” divide. In Liberal-Conservative battlegrounds where the majority of voters opted to leave, the third-placed Labour and Green vote tends to be heavily pro-Remain.)

But it’s not just Conservatives with the Liberal Democrats in second who have cause to be nervous.  Labour MPs outside of England's big cities have long been nervous that Ukip will do to them what the SNP did to their Scottish colleagues in 2015. That Ukip is now in second place in many seats that Labour once considered safe only adds to the sense of unease.

In a lot of seats, the closeness of Ukip is overstated. As one MP, who has the Conservatives in second place observed, “All that’s happened is you used to have five or six no-hopers, and all of that vote has gone to Ukip, so colleagues are nervous”. That’s true, to an extent. But it’s worth noting that the same thing could be said for the Liberal Democrats in Conservative seats in 1992. All they had done was to coagulate most of the “anyone but the Conservative” vote under their banner. In 1997, they took Conservative votes – and with it, picked up 28 formerly Tory seats.

Also nervous are the party’s London MPs, albeit for different reasons. They fear that Remain voters will desert them for the Liberal Democrats. (It’s worth noting that Catherine West, who sits for the most pro-Remain seat in the country, has already told constituents that she will vote against Article 50, as has David Lammy, another North London MP.)

A particular cause for alarm is that most of the party’s high command – Jeremy Corbyn, Emily Thornberry, Diane Abbott, and Keir Starmer – all sit for seats that were heavily pro-Remain. Thornberry, in particular, has the particularly dangerous combination of a seat that voted Remain in June but has flirted with the Liberal Democrats in the past, with the shadow foreign secretary finishing just 484 votes ahead of Bridget Fox, the Liberal Democrat candidate, in 2005.

Are they right to be worried? That the referendum allowed the Liberal Democrats to reconfigure the politics of Richmond Park adds credence to a YouGov poll that showed a pro-Brexit Labour party finishing third behind a pro-second referendum Liberal Democrat party, should Labour go into the next election backing Brexit and the Liberal Democrats opt to oppose it.

The difficulty for Labour is the calculation for the Liberal Democrats is easy. They are an unabashedly pro-European party, from their activists to their MPs, and the 22 per cent of voters who back a referendum re-run are a significantly larger group than the eight per cent of the vote that Nick Clegg’s Liberal Democrats got in 2015.

The calculus is more fraught for Labour. In terms of the straight Conservative battle, their best hope is to put the referendum question to bed and focus on issues which don’t divide their coalition in two, as immigration does. But for separate reasons, neither Ukip nor the Liberal Democrats will be keen to let them.

At every point, the referendum question poses difficulties for Labour. Even when neither Ukip nor the Liberal Democrats take seats from them directly, they can hurt them badly, allowing the Conservatives to come through the middle.

The big problem is that the stance that makes sense in terms of maintaining party unity is to try to run on a ticket of moving past the referendum and focussing on the party’s core issues of social justice, better public services and redistribution.

But the trouble with that approach is that it’s alarmingly similar to the one favoured by Kezia Dugdale and Scottish Labour in 2016, who tried to make the election about public services, not the constitution. They came third, behind a Conservative party that ran on an explicitly pro-Union platform. The possibility of an English sequel should not be ruled out.  

Stephen Bush is special correspondent at the New Statesman. His daily briefing, Morning Call, provides a quick and essential guide to British politics.