![[personal profile]](http://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
The UK Census
Mar. 27th, 2011 11:54 amI've just been reminded that today is census day, as apparently hundreds of people are expected to forget to fill it in so the census organisers are issuing a reminder. I am not one of those forgetful people - I filled mine in ages ago when it first arrived in the post - but I'd just like to say that, as questionnaires go, it was a pretty bad one.
Firstly, this was actually on the census:
Q17) Question 17 is not a question. Please proceed to question 18.
WTF?! It's like they knew they needed 40 questions but couldn't think of enough, so decided to make question 17 a non-question instead. I can't think what practical purpose it serves other than to fill up space on the page.
Also, I can't remember the question number but they asked this:
How would you rate your health?
Based on what criteria, exactly? It's a bit of a woolly questions. The options for answering were this:
Very Good, Good, Fair, Bad, Very Bad
Ok, I don't consider myself an unhealthy person, but since I was just recovering from a rather nasty virus when I filled it in should I have put 'bad'? Or should I have put 'good' (which is what I actually did) as that's what I expected to be on census day (as it is I still have a really horrible cough that I need to see the doctor about). But whatever I should have put, it's still a question where it's impossible to get an objective answer. One person's idea of 'good' might be someone else's idea of 'fair'. I'm pretty sure my statistics teacher would have facepalmed when he got to that question, as he's always teaching us to avoid vague and subjective things like that if we want our results to be statistically significant.
There was also one of the questions asking whether or not, last week, was I unemployed, at work, or off sick. It was incredibly vague as to whether it meant long term sick leave or if you'd just happened to be ill that day (or if you'd gone in for half of the week and missed the other half due to illness). At the time I filled this is, that particular week in question I was off ill. But if I'd been filling it in on census day I would have been at work the previous week (so that's the answer I put as it was the only day I've had off sick all year). That was another one of those questions where it needed to make it more clear what it meant. Like the question where they asked if you'd had no permanent address in the past twelve months where were you living last year? Er, at what point in the last year? If they've had no permanent address they've clearly been living all over the place. I think it meant where were you living precisely 365 days before census day, but once again it wasn't very clear.
Most of the questions were pretty standard but there were some really daft ones too. If the question could be interpreted differently by different people it makes the results of the census pretty meaningless. I really hope they've got some better statisticians/census devisers working for them by 2021.
Firstly, this was actually on the census:
Q17) Question 17 is not a question. Please proceed to question 18.
WTF?! It's like they knew they needed 40 questions but couldn't think of enough, so decided to make question 17 a non-question instead. I can't think what practical purpose it serves other than to fill up space on the page.
Also, I can't remember the question number but they asked this:
How would you rate your health?
Based on what criteria, exactly? It's a bit of a woolly questions. The options for answering were this:
Very Good, Good, Fair, Bad, Very Bad
Ok, I don't consider myself an unhealthy person, but since I was just recovering from a rather nasty virus when I filled it in should I have put 'bad'? Or should I have put 'good' (which is what I actually did) as that's what I expected to be on census day (as it is I still have a really horrible cough that I need to see the doctor about). But whatever I should have put, it's still a question where it's impossible to get an objective answer. One person's idea of 'good' might be someone else's idea of 'fair'. I'm pretty sure my statistics teacher would have facepalmed when he got to that question, as he's always teaching us to avoid vague and subjective things like that if we want our results to be statistically significant.
There was also one of the questions asking whether or not, last week, was I unemployed, at work, or off sick. It was incredibly vague as to whether it meant long term sick leave or if you'd just happened to be ill that day (or if you'd gone in for half of the week and missed the other half due to illness). At the time I filled this is, that particular week in question I was off ill. But if I'd been filling it in on census day I would have been at work the previous week (so that's the answer I put as it was the only day I've had off sick all year). That was another one of those questions where it needed to make it more clear what it meant. Like the question where they asked if you'd had no permanent address in the past twelve months where were you living last year? Er, at what point in the last year? If they've had no permanent address they've clearly been living all over the place. I think it meant where were you living precisely 365 days before census day, but once again it wasn't very clear.
Most of the questions were pretty standard but there were some really daft ones too. If the question could be interpreted differently by different people it makes the results of the census pretty meaningless. I really hope they've got some better statisticians/census devisers working for them by 2021.
