Monday, 7 February 2011

Ask a stupid question... Census2011

and you get a stupid answer. Or so the phrase goes.

On the 27th March Office of National Statistics (ONS) will ask almost the entire population of England and Wales 61 questions. This is the 2011 Census. Censuses do invade privacy, however they create very useful and interesting data; but only when they ask the right questions, in the right way.

There has been a census every 10 years since 1801 (excluding 1941). Before that famous censuses include the Domesday Book, of 1085, and the Roman habit of carrying out censuses for tax purposes.

Some of the 61 questions are not actually used, for example;
  • In England question 17 is intentionally left blank, while in Wales it asks "Can you understand, speak, read or write Welsh?"
  • Question 43 reads "There are no more questions for Person 1."
But this is not what I am interested in.

As all listeners, readers and watchers of Douglas Adams' HHG2TG will know you need to know what question you want answered before you ask it. 42 doesn't quite cut it, especially when the question turns out to be "What do you get if you multiply six by nine"?

Asking for the Answer to the Ultimate Question of Life, The Universe, and Everything will not get you far.

It is this problem which the census is presented with. In order to ask the right questions they need to know what sort of answer they want; but then they restrict the potential range of answers they could get by asking those questions.

Some questions are sensible such as;
  • Question 25 - "Which of these qualifications do you have?" Which asks respondents to tick all the boxes which apply, not just their highest or most recent qualification, collecting all the relevant information.
Of course this may not seem sensible if you are currently working towards a qualification, but the census is a measure of a time point, not a prediction.

But then comes Question 41;
  • "How do you usually travel to work?"
This question asks respondents to "tick one box only... for the longest part, by distance, of your usual journey to work." Personally I use 4 modes of transport (Bicycle, Bus, Walking and Underground), and work at 2 different locations. My transport varies day to day, depending upon weather, illness and flat-tyres. I cannot answer this question with one tick.

We must also consider, does this provide the information the government want to know. In my case they will only know that I use the Underground, not that I cycle to the station at one end, and catch a bus at the other. These are all equally important, time wise each portion is approximately equal (15 minutes), but this gets me far further by tube. Those who fall through the gaps always present problems, but the ONS has completely failed to consider them here, and will therefore miss some info it would wish to know.

I fall through the gaps in other places. Question 9 asks;
  • "What is your country of birth?"
Respondents who were born within the UK are then asked to skip to question 13. This ignores the questions;
  • "...when did you most recently arrive to live here?"
  • "Including the time you have already spent here, how long do you intend to stay in the United Kingdom?"
  • (Question 11 is to direct respondents whether to answer question 12 or not)
Unfortunately, as an expat, I was born in the UK, left, returned again, and will probably leave again. I am sure that the ONS, let alone people 100 years in the future, will be interested in emigration of Brits as well as immigration of other nationals. This may well impact upon the analysis of the responses, showing that Brits also emigrate (and return) may help dilute the predictable headlines about immigration levels when the response data are released.

Then, of course, there are ethnicity and religion. I think that registering ethnicity is somewhat important, but religion? And then there is how the question is being asked. The following tick boxes are available;
  • No religion
  • Christian (including Church of England, Catholic, Protestant and all
    other Christian denominations)
  • Buddhist
  • Hindu
  • Jewish
  • Muslim
  • Sikh
  • Any other religion, write in (followed by 17 boxes)
The choice of religions seems reasonable, at the last census the religions listed were the 6 most observed. However there were almost 3 times as many people who observed "No Religion" than those observing all non-Christian religions. No religion encompassed stated beliefs including "Jedi Knight, Agnostic, Atheist and Heathen", some more serious than others, however there is a significant difference between just not having a religion, and actually being atheist. The British Humanist Association take issue with the wording, which assumes that respondents have a religion.

There is also the issue that, in 2001, respondents choosing "'Other, but not [writing] in any religion" were lumped as not religious. This is not necessarily a problem with the question, but with the use of the answers, but I for one would not like to be misrepresented in this way, especially when 17 characters is not enough space to state my nuanced view of religion.

Of course people will always argue that censuses are intrusive, and damaging to privacy, and I agree up to a point, however the data is going to be collected like it or not; why can it not be collected with more thought-through questions, less misrepresentation of data and fewer ways to fall through the gaps.

You can find the questions here and their final implementation here.

No comments:

Post a Comment