Corrections made to census data for statistical purposes
Zoe O'Connell made this Freedom of Information request to Office for National Statistics This request has been closed to new correspondence. Contact us if you think it should be reopened.
Dear Office for National Statistics,
This is a follow up request to FOI01148/O'connell/QE1 and FOI01193/O'Connell/QE1 and you may like to refer to those requests for context.
Please can you supply:
1) A complete list of adjustments (including the underlying algorithm used for this adjustment if possible) to be made to data used for statistical purposes for the 2011 Census.
For example, this list might consist of entries such as "If P1 and P2 both have Gender Male and indicate "Married" as a relationship, the relationship is instead treated as "Civil Partnership"
I do not require English-language output. Extracts from computer programs, SQL queries or spreadsheet formula are also useful.
2) In FOI01148/O'connell/QE1, it is stated that "It
is recognised that the majority of respondents recording themselves as being in a polygamous relationship in a UK census do so erroneously". In FOI01193/O'Connell/QE1 it is also reported that the rate of such responses is not recorded.
Please can you provide any statistical or other material (E.g. analysis of follow up surveys) used to come to the conclusion that the majority of such responses are erroneous.
3) Please provide the same data for question 2 for any other adjustments identified in response to question 1, if this would not exceed the allowable time for this request.
4) Please supply the same data as for question 1 for the 2001 census, if this would not exceed the allowable tiem for this request.
Yours faithfully,
Zoe O'Connell
Our Reference: FOI01217/O'Connell/QE1
Dear Ms O'Connell
I am writing regarding your request for information relating to
Corrections made to census data for statistical purposes.
Firstly let me apologise for the time taken to respond to you. After
speaking with the business area concerned with this matter, we have
concluded that we will be unable to proceed with your request without
clarification of the information you wish to receive, as your request as
it stands appears to be very wide ranging and would exceed the appropriate
limit as imposed by FOI.
As such we would appreciate the opportunity to discuss this request with
you and try and narrow down your requirements so that we can provide the
information that best suits your needs, therefore please could you contact
myself or my colleague Paul Wearn on 01329 444465 to discuss this case,
alternately if you would like us to call you please can you provide us
with a contact number.
Please note that if I do not hear from you within three months from the
date of this letter, then I will consider your request closed.
If you wish to discuss any of the above, please contact me. Please
remember to quote the reference number above in any future communications.
Yours Sincerely,
Nikki Shearman
For the latest data on the economy and society consult National Statistics
at http://www.ons.gov.uk
Legal Disclaimer : Any views expressed by the sender of this message are
not necessarily those of the Office for National Statistics
*********************************************************************************
Communications via the GSi may be automatically logged, monitored and/or
recorded for legal purposes.
Dear Nicola Shearman,
Apologies for the delay in replying. Since I submitted the original request, I have learnt that you are now using the CANCEIS system for the 2011 Census Data.
Question one can thus best be summarised by:
"Please can you provide the CANCEIS edit code/DLTs being used for analysing 2011 census data."
If it helps to limit the scope of the request, I am predominantly interested in anything that relates to gender and relationship status. Please feel free to ignore questions 3 and 4 as they are no longer relevant.
If my assumptions are incorrect or unclear or would still be problematic to release data on, please let me know and I shall give Paul a call.
Yours sincerely,
Zoe O'Connell
Our Reference: FOI01217/O'Connell/QE3
Dear Zoe O’Connell
In response to your revised request under the Freedom of Information Act,
the requested Design Logic Tables (DLTs) and edit rules to be implemented
for the 2011 Census relating to relationship and sex are attached.
With regard to the ‘sex’ question, where someone has ticked both options
or left the question unanswered, a single response will be created. This
is not in any way intended to reflect the true gender identity of any
individual, it is simply done to ensure the completeness of the final
outputs as for every other census question (except the question on
religion which is voluntary). Note that the scanned image of the original
census record is microfilmed and retained for 100 years when they are made
publicly available. These images will, of course, retain the original
response.
The census processing system does not accept instances of multi-ticking
both ‘male’ and ‘female’. In these cases the probabilistic statistical
system will impute a sex of either ‘male’ or ‘female’, in the same way as
for those who have left the question unanswered.
The system identifies a "donor" record (someone who has answered the
question with a single tick, and has other similar characteristics) and
copies their response.
Attachments
You have the right to have this response to your freedom of information
request reviewed internally by an internal review process and, if you
remain unhappy with the decision, by the Information Commissioner. If you
would like to have the decision reviewed please write to Dennis Roberts,
Office for National Statistics, Room 1214, Government Buildings, Cardiff
Road, Newport, Gwent, NP10 8XG.
If you have any queries about this email, please contact me. Please
remember to quote the reference number above in any future communications.
Kind regards,
Paul Wearn LLB (Hons)
Legal Services
Office for National Statistics
For the latest data on the economy and society consult National Statistics
at http://www.ons.gov.uk
Legal Disclaimer : Any views expressed by the sender of this message are
not necessarily those of the Office for National Statistics
*********************************************************************************
Communications via the GSi may be automatically logged, monitored and/or
recorded for legal purposes.
Dear Paul Wearn,
Thanks for this - unfortunately, the font used in "Relationship and Gender DLT s V0.1.pdf" is very small and has been rendered as graphics rather than text, so it's unreadable. Would it be possible to resend with a larger font or, ideally, as a text (.txt) file?
Yours sincerely,
Zoe O'Connell
We work to defend the right to FOI for everyone
Help us protect your right to hold public authorities to account. Donate and support our work.
Donate Now
Zoe O'Connell left an annotation ()
Folk that end up here may also find this paper from Southampton University on the CANCEIS system being used for the 2011 Census data interesting. www.s3ri.soton.ac.uk/isi2007/papers/Pape...