Diary: Pollsters, parliament and pets

?7.05.10
I tweet therefore I am
The opinion pollsters faced stiff competition in the run up to the general election from some social media upstarts looking to muscle in on their territory. Tweetminster, which tracks discussion of UK political happenings on Twitter, ran an experiment in which it tried to predict the election result based on what people were tweeting. Its prediction of the share of vote (Tories 35%, Labour 30%, Lib Dem 27%) came out closer to the actual result than several of the traditional pollsters. When we raised this – on Twitter of course – with Ipsos Mori’s Ben Page, he seemed inclined to embrace the new buzz-tracking technique for public opinion research. “If we can prove and publish full methods – it’s reliable,” he wrote, although he also pointed out how Tweetminster wrongly predicted that TV personality Esther Rantzen would win in Luton (she came fourth). When asked whether Twitter analysis might be able to beat the traditional polls by the 2015 election, Ben replied: “Only by chance :)”

7.05.10
Representing research
Amid the fevered media coverage of the election result, there was an important detail that most of the national press overlooked: we now have two market researchers in parliament. Simon Danczuk, who co-founded social research agency Vision 21, is the new MP for Rochdale, winning the marginal seat back for Labour from the Lib Dems, while Nadhim Zahawi left his role as CEO of YouGov to stand in the safe Tory seat of Stratford-on-Avon, duly winning it with a thumping majority. Zahawi will be looking to make his mark on the government benches (not literally) but Danczuk has already had his first taste of the political limelight, as the man standing awkwardly in between Gordon Brown and Gillian Duffy during the infamous Bigotgate encounter. He still won the seat, despite losing her support thanks to his leader’s unguarded comments. Things can only get better, Simon.

11.05.10
Nazi-related story of the month
West Midlands Ambulance Service has been in hot water for commissioning a survey that asked staff to rate how ‘cool’ they thought Hitler was, compared to other famous leaders such as Winston Churchill, Fabio Capello and Richard Branson. The tabloid outrage pretty much wrote itself, especially when it emerged that the survey was part of a study that cost £10,000 (the words ‘was part of a study that’ being omitted from much of the coverage). It all goes to show how a single mention of the Nazis can turn a questionable but boring use of public funds into a full-blown PR catastrophe. The ambulance service said the survey was aimed at working out how to “make leadership cool” as part of a staff development programme. “With hindsight,” a spokesman said, “it would have been better to have used a different example.”

19.05.10
All pets are off
In order to better understand the world of market research, Diary occasionally takes part in surveys. We’ve been asked some weird things in the past, but we were bemused to be faced with the following question in an online questionnaire: “Please select hamster from the following options: hamster, cat, dog, mouse, rabbit.” Diary duly selected ‘hamster’ and clicked ‘next’. After that the questions went back to normal, and the hamster was not spoken of again. Presumably the question was designed to weed out bots, because only a human could deduce the correct answer. Diary suspects this may be an example of the unspoken rule highlighted Doron Meyassed and Tom Hoy in their article in this month’s issue of Research: Thou shalt not explain the logic behind the question. Surely in this instance it would be wiser just to say what’s going on rather than asking people mad questions out of the blue?