Atomik Research

The Creative Market Research People

Read All About It

Archive

Closer Than You’d Think…

Atomik's Latest Labour Leadership Poll


Atomik Research recently commissioned a quick insight into who the British public think will win the upcoming Labour leadership race.

For those of you who haven’t been within reasonable distance of a TV, newspaper or smartphone lately, Labour will soon choose the person that will lead them into the next election in 2020.

Look at the graph in its most basic context and you see a simple, basic opinion that suggests (when the poll was commissioned) Jeremy Corbyn is 13% ahead of his nearest opponent. That’s that. Call it all off and give Corbyn the badge, or hat, or ceremonial staff or whatever it is they use to signify a political victory nowadays.

But, obviously, there’s more to it than that. You only have to look at the differences in the pre-election polls and the exit polls to know, trusting one and only one slant on the numbers is a wild and daft thing to do (Paddy Ashdown and Alistair Campbell – we’re looking at you).

Consider this, when you’ve asked someone ‘WHAT DID YOU JUST DO?’ – it’s a cut and dry question. No real alternative, unless you’re just straight-up lying about it.

“I voted for this person.” “I preferred sample A.” “It was Number Three who mugged me, officer.”

However, ask someone an opinionated question a week before they’re doing something and the results, invariably, will be different.

“I’m torn between these people, I’ll tell you the safe option.” “I can see the merits of both samples.” “I hope not to be mugged in the near future, now who let you into my house…?”

That kind of thing.

Add extra elements to the questions, extra layers to the options that you lay in front of people and while it may seem you’re starting to tie yourself in knots, you can use these seemingly scrambled strands of dialogue to develop a more detailed picture of actual opinion.

Specifically, the Labour vote will ask people to choose their favourite of the four candidates and apportion a lower ranking to those they deem less worthy. You can nominate as many of them as you’d like, as long as you rank them in order.

Using the Atomik numbers, 47% have placed Corbyn as their ‘1’, 34% have picked Andy Burnham – so on and so forth. But where do their second choices lie? When there are three choices left, does the one who stands on their own become the better or worse option?

Adding layers into lines of questioning, when done correctly, can lead an audience to re-think their initial selection; when it comes to information – the more, the merrier.

Let’s look at it this way – say someone asked you if you’d prefer a glass of milk or Bovril. The vast majority have already made up their mind, ticking ‘Milk’ or ‘Bovril’ without even thinking about it. Prime Minister Milk is duly elected – cue: ‘Got Milk?’ headlines nationwide.

Now, asked to order those two black-and-white (brown-and-white, whatever…) choices in amongst a selection of others (we’ll use water and cola) and the result isn’t as clear as it was.

Invariably, Prime Minister Milk’s majority starts to wane – given more options, the people who chose ‘milk’ because they just do NOT like Bovril might choose water or cola as their first choice – the same is, of course, true for those who chose Bovril.

However, now we’re asking people to put their options in order from first to fourth (or as many as they deem appropriate), so someone who chose ‘milk’ because it was the closest option to ‘water’ will now pick ‘water’ first with ‘milk’ potentially in second, third or fourth place. The full landscape of opinions becomes much more intricate.

The votes are tallied, milk still has a lead, with water now in second place, Bovril in third and cola in fourth. Cola is eliminated from the vote – where does that share of the vote go now? How many cola-folk will jump over to milk, water or Bovril?

It’s virtually impossible to work out exactly what happens next. Milk may still win, but by what appears to be a much slimmer margin than it was before.

Apply this to the current Labour leadership race and noticing that Corbyn, who is (in everything but name) practically part of a different party and the 13% lead he has seems increasingly-slim. Say Yvette Cooper finishes fourth in the first round of voting, where does her 8% go?

Realistically, you’d think it wouldn’t go to Corbyn. Are we, effectively, looking at a three versus one scenario, where the three are winning by 3%?

Even though Corbyn has become the most searched of all the leaders, with the highest search-hits in 539 of 650 British constituencies, who are the people searching for him? Are people that support Corbyn more inclined to use the Internet to find their facts? Of course, not every search will be positive; just because you’ve searched ‘Adolf Hitler’ doesn’t make you a Nazi.

Which proves, more than anything, that the best kinds of research and development of insight are all about the questions you ask and who, specifically, you ask them to. We probably won’t be any closer to the answer of who’ll be the Labour leader until after a couple of rounds of voting – what we can do, is make our best, most informed guess.

By Stuart Buchanan

 


  1. No comments yet