The Campaign Company specialises in social research and behaviour change. This is your guide to what we’ve been reading. Here’s what’s coming up this week:
This week we get ultra-methodological in our Polls, Politics and Policy section, mounting a spirited defence of focus groups – a medium that’s close to our hearts. Psephology nerds, enjoy!
And of course, there’s the chattering Joe Bloggs that is Charlie’s Attic, this week featuring the definitive explanation of why you can’t tickle yourself.
Focus on focus groups
Our eye was caught this week by a Twitter thread from polling expert and TCC friend Chris Curtis. Within it, Curtis points out a methodological issue with the increasingly popular medium of focus groups.
Chris Curtis‘ point is that it’s a mistake to read discussion groups as a barometer for the balance of public opinion, because of their tight focus on swing voters. No matter what the wider state of polling, he writes, there will always be a raft of swing voters. Hence, the sentiments heard in focus groups will invariably imply that things are on a knife edge – even if a landslide election is in the offing. (He uses the two diagrams below to illustrate this).
This is a really important argument, and one which is not emphasised nearly enough. It’s especially vital in light of the growing appetite among broadcasters for showing such groups, as a way to get under the skin of British politics. It’s important that media organisations don’t make the mistake of implying that the ‘swing voter’ and the ‘average voter’ are the same thing.
Yet it’s also an error to do as some of those replying to Curtis did and imply that this undermines the whole practice of focus grouping. Cynicism about focus groups – as superficial or shallow – doesn’t tend to stack up.
At TCC we deploy qualitative research a lot. We believe that discussion groups tell you something wholly different to quantitative research (which we also use extensively). They reveal the texture of public opinion and crucially the reasoning and deeper motivations behind the opinions and behaviours which people hold.
In the field of public health for example – an area where we do a lot of work – the decisions which people make are often extremely hard to tease out through polling. Yes/ No questions or rating scales, of the sort you might get in a typical survey, will often fail to explain why someone might struggle to quit smoking, put off an important visit to their GP, or be frightened to have the Covid vaccine. The same goes for any field where people’s behaviour does not correspond to an entirely rational cost-benefit process – including politics.
The best set up is often to use quantitative work like polling to understand ‘what’, and then to use qualitative approaches like focus groups to understand ‘why’. And the biggest risk is that you mix these things up, presuming to understand the internal logic behind a decision on the basis of how people answer a polling question, or to present the shape of public opinion using a focus group of eight people.
And finally this week Charlie’s Attic, the perennial vox-pop to our peer-reviewed study:
