Really counts

Am I the only person in Britain who hasn’t been ‘surveyed’?

Dog owners, home owners, car owners. Shoppers, holidayers, workers, retirees. Bridge-players, football watchers. Family people, voters, non-voters… everyone is asked their opinion.

Except old-geezers. No one has ever asked me what I think… I’ve never been stopped or phoned.

Plenty are. The polling and survey business is worth £5.3bn. Mostly, connecting with people on-line or the phone…

… getting the right people is tricky. Pollsters claim to have all sorts of mathematical jiggery-pokery and can represent the views of us all, based on the upshot of a thousand, or so phone-calls.

They get it wrong. The Teresa May election, Brexit, the 2015 general election. In the US, polls suggested Hilary Clinton would win the 2016 election by 5%. They were Trumped!

Telephone polling, is quick and easy but we know not to answer the phone from a masked or a weird number… it’s either a scam or a time-waste and…

… as more of us don’t answer, we screen ourselves out of the poll and the pollsters lose ‘randomness’.

Gathering opinions is getting more difficult. Questions like; ‘what do you think’ are complex and answers have to be considered. Often difficult for people to articulate on the hoof.

And, we lie! People always say they’re going to vote, when the numbers tell us a lot of us don’t.

Can standard sampling and weighting techniques correct for sampling problems? I’m not persuaded. Entertain yourself with Meng and his paradoxes and see if you agree.

Can we trust them?

Pollsters like to pretend their work is terribly-complex-dahling… but it’s not.

If a poll finds a survey has 100 members of a particular demographic group, but should have 110 of them (in a poll of, say, 1,000), then it will ‘weight’ the answers of that group so that each of those 100 respondents counts as 1.1 people.

Happy with that?

Think about a chef tasting a saucepan full of soup, with a tea spoon. If the soup is stirred properly she’ll have a representative flavour. If not… well, it’s all about the stirring.

Statistical theory tells us; in a random poll of 1,000 people, with a 100% response rate, then 19 times out of 20, a poll will be accurate to within 3%. In other words, it will record at least 47%, and no more than 53%, for each view.

Happy with that?

Don’t forget the one in 20 chance that the poll will fall outside this range.

In my view, small sample sizes, the time window, geography, the day of the week, gender, age, experiences, events all conspire to make pollster-ing an extension of witch-craft.

If you want to know what’s what, ask an old-geezer… but pollsters never do!

That said, a poll that I thought was worth looking at didn’t get much of an airing. I think only the Indy ran with it?

In a poll of NHS people;

‘…some 47% have looked at work outside the NHS… 29% have actively inquired about non-NHS work.’

The study was done by the University of Bath and YouGov. Tracking opinion from 2020 for four years. Involving 1,500 NHS people.

Only 37% of staff questioned in spring 2023 said they would recommend working for the NHS… a drop from 61% in winter 2020/21.

Only one in four nurses said they would recommend working in the health service… ouch!

I know, you are going to tell me… ‘catch me on the wrong day and I’ve probably had a look at the jobs pages’, and you’re right.

Who hasn’t?

You’ll also tell me… we may not know the details but we all know the direction of travel and what’s happening ‘down-your-way’… and you are right. But…

… this is a longitudinal study, measured over time, ironing-out knee jerks after a bad day.

We might dismiss polling as flimflam but this is McCoy.

This says there are still serious structural problems with staffing, retention and attitude.

The Indy (towards the end) quotes Navina Evans, NHSE’s workforce honcho.

She makes perfect sense about what the centre is doing and it’s good stuff, but a workforce engaged on national conditions is one thing. It’s what’s happening when they are working locally that really counts…

… and why this poll, really counts.

Have the best weekend you can.

News and Comment from Roy Lilley
Contact Roy – please use this e-address
Reproduced at by kind permission of Roy Lilley.