Contact SARs Support |
CCSR These pages are maintained by the SARs support team.
Send us comments on this web page.
ONS have begun work on the detailed specification for the 2001 SARs that was provided by CCSR in September 2001. A copy of the specification is available at http://www.ccsr.ac.uk/sars/2001/. As part of this process, ONS are considering confidentiality implications and re-working some of the research published in Dale and Elliot (2001). Given the current climate of concern over confidentiality and disclosure, ONS wish to release less detail on some variables in the 2001 SARs than in the 1991 SARs. We had argued that the 1991 SARs were, in fact, rather safer than had been assumed at the time of their release and thus more detail could be released for a similar level of risk. Despite the possibility of reduced detail on some variables, the 2001 SARs promise to be a valuable data resource, particularly with the addition of new Census questions on religion and health.
The ONS
position is being distributed with this newsletter as a separate
information sheet. This explains that ONS will conduct a consultation
exercise on the SARs which will begin in September and run through
October. We are holding a meeting on Friday 11 October at the Royal
Statistical Society, London to debate the proposals for the 2001 SARs.
We are keen that users have the opportunity to present their views and
explain to ONS how particular proposals will affect their research
plans. Programme 10.30 Coffee and registration 11.00 Introduction to
SARs 2001 - background on 1991 SARs and the developments and rationale
that led to the proposals for 2001 SARs (SARs team) 11.30 ONS
proposals on detailed specification (ONS Census Division) 12.00
Discussion: contributions from users* 12.45 Lunch 13.30 Costs,
registration and dissemination of 2001 SARs Demonstration of
registration Demonstration of NESSTAR, NSDstat 14.00 Discussion 15.00
Timetable for delivery of 2001 SARs Programme of user support and
training 16.00 End *We are happy to allocate 5-minute slots to allow
people to make a specific point. Please email and let us know if you
would like a slot booked in the programme.
For further information on any aspect of the SARs not covered in this
newsletter please contact Ruth Durrell on 0161 275 4721
The registration procedures for the 2001 SARs will be simple, quick and efficient. A Census Registration Service (CRS) has been developed by the Data Archive. This is available via the SARs website (www.cccsr.ac.uk/sars) or directly from the CRS (http://www.census-registration.ac.uk). SARs users in the academic sector will register using an Athens user ID and password. Lucy Bell provides full details of the registration service on page 5 of this newsletter.
We need all 1991 SARs users to re-register if they wish to access the SARs over the web. However, for people wishing to continue to analyse the 1991 SARs on their PC, there is no immediate requirement to re-register.
Registration for business users of the SARs will cost £1000 for each SAR. Local authorities, charities, health authorities and central government departments will be entitled to a discounted non-commercial rate of £500 per file. This will entitle the organisation to receive the data which may then be accessed by 10 named individuals. One person will be nominated as representative and be asked to supply the list of users. All ten users in each organisation will receive support, backup and a free place on a training course. Further copies of the CD and the registration of additional users will be available for an additional fee upon request.
As promised in SARs Newsletter No 17, the new and improved SARs website can be found at the new address of http://www.ccsr.ac.uk/sars/ . The website contains background information on the SARs, how to access and use the data, and details of support services. Registered users will also be able to download the data directly from the website. Any comments on the SARS website would be welcome (s.smith@man.ac.uk).
Whilst users will still be able to receive and analyse the SARs in
statistical packages SPSS, SAS and STATA, we are introducing new
web-based analysis and extraction software for the SARs, making access
much easier for both experienced and occasional users.
The Nesstar Light interface via a web browser, and the desktop package
NSDstat, are both new, and coming on-line shortly. Nesstar will be
made available by the end of
September 2002, and NSDstat by the end of October. Datasets that will
initially be available in both Nesstar and NSDstat will be the 1991 GB
and Northern Ireland Household and Individual SARs, and a teaching
dataset based on the Individual SAR
for the City of Birmingham. SARs for 2001 will be
added as they become available. To be able to download and use any of
these packages, academic users must have registered with the Census
Registration Service, and non-academic users must have registered with
CCSR.
User guides for both Nesstar and NSDstat will be on our website
as soon as when the 1991 data
is available in these formats in September and October
2002 respectively.
The full service now registers users for the Samples of Anonymised
Records (SARs) as well as for the other Census Data Support Units:
Current users of the four census services will need to re-register with the new Census Registration Service, in compliance with the Data Protection Act 1998. This is a quick and painless process, requiring the user to fill in a single online form.
Over the last nine months the Census Registration Service has consulted with more than 400 census data users in UK higher and further education in order to understand the registration concerns held by the current service users. The results of these surveys have been incorporated into the system as far as is possible. The new system includes the following functionality:
For more information about the development of the service please visit the Census Registration Service web site (http://www.census-registration.ac.uk) or contact its Enquiry Service via email (census-registration@essex.ac.uk).
*Lucy Bell is Service co-ordinator at the Census Registration Service at the University of Essex.
A number of pilot units are now ready for testing, and available from our project website: www.ccsr.ac.uk/rschproj/chcc/home.htm .* Methodological units currently available include the following:
In September, Wendy Olsen will be joining CCSR and taking responsibility for the development of further substantive units over the following year (see www.ccsr.ac.uk/rschproj/chcc/issues.htm for a provisional list of topics).
All materials are being developed using 1991 SAR data, but will be revised to incorporate 2001 SARs as they become available.
CHCC team at CCSR: Angela Dale; Mark Brown; Jo Wathan; Sam Smith
Other project partners are: MIMAS and the Census Dissemination Unit,
the University of Manchester; The School of Geography, University of
Leeds; The History Data Service, The Data Archive, University of
Essex; The LTSN Centre for History, Archaeology and Classical Studies
and University of Glasgow.
The detailed occupational information in the SARs allows a large number of social status measures to be attached. The 1991 SARs contained the Registrar General's social class and Socio-economic group - coded by ONS as part of the Census. However, on the Household SAR, CCSR added the Goldthorpe class scheme as well as various international measures of occupation and social status - the International Standard Classification of Occupations (ISCO), International Socio-Economic Index of Occupational Status (ISEI) and Standard International Occupational Prestige Scores (SIOPS).[1]
For the 2001 Census ONS have replaced RG Class - widely criticized as lacking a theoretical base - with its new schema, National Statistics Socio-economic class (NS-SEC), which is closely related to the Goldthorpe class scheme (Rose et al 1998). However, in order to provide comparability between 1991 and 2001, CCSR will derive RG and SEG classes for the 2001 SARs and a 5-category version of NS-SEC for the 1991 SARs. The international class measures will, again, be added to the Household SAR for 2001. In addition, the Cambridge Social Interaction and Stratification Scale (CAMSIS) will be added to the Individual and Household SAR, replacing the Cambridge Occupational Scores, which are available on the 1991 SARs.[1] Table 1 summarises the class measures for each SAR.
Notes: 1 Variables marked by * are supplied by the ONS and the rest are derived by CCSR and added to the files. 2 All variables, regardless of the mark, are available in the 1% SAR at the household, family and individual levels. For instance, soclass is available for the individual, and the family and the household reference persons. 3 Variables marked by /- are not available on the 2% SARs.
The structure of the 2001 SARs will be similar to that of the 1991 SARs. There are, however, some notable differences in the questions of the 2001 census, and hence in the variables of the new SARs. Key developments include the new class schema adopted (see note on `Class Measures in the SARs', this issue); greater differentiation and availability of educational and professional qualifications; religious affiliation (for the whole of UK) and religious background (for Scotland and Northern Ireland); `general health' conditions; weekly hours of care for people with long-term physical/mental ill-health, disability or problems related to old age; the household membership matrix; and the changing definition of unemployment from self-reported unemployment in the 1991 census to the ILO (International Labour Organisation) definition in the 2001 census. There are, however, many other possible research applications including ethnic differences, labour markets, and migration. A detailed discussion of how the 1991 SARS have been used is available at http://www.ccsr.ac.uk/sars/use/findings/
It will also be possible to use the 2001 SARs to test competing sociological theories. For instance, the SC and SEG schemas are said to lack a clear conceptual basis, but one can make comparisons between theoretically-grounded class schemas, such as that based on the employment relationship theory and that based on the assets theory. The two theories debate the nature of social class, especially concerning the class character of professional and managerial employees. The former predicts a growing convergence in professional and managerial class formation whereas the latter foresees the opposite. Using class measures now standardised for the 1991 and the 2001 SARs, one might test the construct validities of the schemas against a range of issues of crucial importance to people's lives, such as long-term unemployment, long-term limiting illness, and young people's drop-out from educational participation. Such an analysis would ascertain the relative explanatory powers of competing theoretical paradigms.
These questions are not only of academic interest but have policy implications as well. Whether an increased, similar or decreased proportion of people refuse to acknowledge their religious affiliations, a study of such allegiance would contribute to the development of benchmark data for use by employers and public bodies, particularly the North Ireland Equality Commission, in fulfilling their statutory obligations against discrimination. The kind of research would also inform debates on the changing nature of a deeply divided society.
These are just some examples where research can be conducted using the 2001 SARs. Interested readers can contact yaojun@man.ac.uk for a more extensive paper.
The role of Director will be taken by Ed Fieldhouse, who brings to this role a wealth of research experience in the academic and local authority sectors. Ed has been Deputy Director of CCSR for the past three years.
Early in 2003, we shall be advertising a new chair in social research at CCSR. Informal enquiries prior to the formal announcement are welcome and should be made to Ed Fieldhouse (e.fieldhouse@man.ac.uk).
ESRC Research Methods Programme Angela Dale formally takes up the post of director of this programme on 1 October 2002. Ruth Durrell, who has been CCSR's administrator and SARs administrator since 1992 will become the programme administrator. She will, however, still retain her role as SARs administrator.
ESRC/JISC Data Archiving and Dissemination Service CCSR and MIMAS will be working with the UK Data Archive to provide a national data archiving and dissemination service from January 2003. The UK Data Archive has overall responsibility for the management of the service, the core archiving and dissemination function, and a value-added service for longitudinal and qualitative datasets. However, CCSR will collaborate with the Data Archive in providing a value-added service for large scale government datasets, and MIMAS will provide a value-added service with respect to international data. We look forward to an exciting and fruitful collaboration with colleagues at the UK Data Archive.
New posts at CCSR
We shall shortly be advertising a number of posts at CCSR. Details
will be available on the CCSR web-site and on the University of
Manchester curent vacancies page.
In brief, the posts are:
1. Research associate to work on a Leverhulme-funded project, directed
by Angela Dale and Shirley Dex, to use the Labour Force Survey and the 2001
Samples of Anonymised Records to examine ethnic differences in patterns of
employment and unemployment with particular respect to gender differences
and generational change. This will be a half-time post for 3 years.
2. Research associate at 50% time to join the SARs team. This will
replace Yaojun Li who is moving to a research post as part of a
successful ESRC-funded project to analysis social capital and
consumption. The appointment will be for 3.5 years starting in
January 2003.
3. Two posts comprising 1.5 FTE to provide the CCSR part of the
value-added service for the large-scale government datasets under the ESRC
Data Archiving and Dissemination Service for just under 5 years.
Any of the posts can be combined to make a full-time job or held
separately as a half-time post. All start in January 2003. Informal
enquiries can be made to Angela Dale (angela.dale@man.ac.uk).
What are the SARs? The Samples of Anonymised Records are datasets drawn from samples of 1991 and 2001 Census responses which have had identifying information removed to protect confidentiality. They are microdata files, similar to the sort of data one would obtain from a sample survey. However the files are much larger than survey datasets permitting analysis of small groups and sub-national areas.
How much do the SARs cost? The data are free to Academic Users. The 2001 SARs will be available to public sector bodies for a charge of £500 per file and to the business sector for £1000 per file.
How do I register to use the data? Academic users can register through the Census Registration Service at www.census-registration.ac.uk. Non-academic users should contact us at the address below.
When will the 2001 SARs be available? We anticipate that the 2001 SARs will be available autumn 2003. · Can I use the SARs for teaching? We are keen to actively promote the use of the SARs for teaching. To support this, teaching and learning materials have been produced. Pilot materials are available at http://www.chcc.ac.uk
How do I get regular updates about the SARs? If you do not receive this newsletter and would like to be added to our mailing list please contact Ruth Durrell. To join the SARs email list go to http://www.jiscmail.ac.uk/lists/sars.html
How will the 2001 SARs differ from the 1991 SARs? Registration and Access will be easier. The 2001 SARs will also contain new Census topics (e.g. caring and religion). The precise content of the SARs, however, is currently under consideration by ONS. A decision is expected Autumn 2002.
How do I use the SARs? The SARs can be explored online or downloaded either in their entirety or as subsets and analysed with your own statistics package by registered users.
What topics do the SARs cover? The SARs cover the full range of Census topics including, housing, education, health, transport, employment and ethnicity.
What other 1991/2001 Census products are there and how do I access them? Census results are also available in aggregate (tabular) form for small areas, which are also suited to mapping. Travel to work and migration data are also available. Academic users can seek additional details from http://www.census.ac.uk. Non academic users should contact the Office for National Statistics http://www.statistics.gov.uk
For more information, see http://www.ccsr.ac.uk/sars/. Queries should be directed to sars-helpdesk@man.ac.uk.
SHORT COURSE PROGRAMME 2002 - 2003 A new, enhanced short course programme has been introduced for 2002 - 2003. New courses include An Introduction to Sampling Theory, Data Management with STATA and Statistical Modelling with STATA. A full list of courses is enclosed with this newsletter. Further details, including course outlines, cost and on-line booking form are available at http://www.ccsr.ac.uk/courses/short2002.htm, or contact nasira.asghar@man.ac.uk (0161 275 4736). Early booking is recommended.
Website: http://www.ccsr.ac.uk/sars/
Email: sars-helpdesk@man.ac.uk
SARs Helpline: (0161) 275 4735
Fax: (0161) 275 4722
Postal Address:
CCSR, FSSL,
Dover St. Building,
University of Manchester,
Oxford Road,
Manchester,
M13 9PL
CCSR is a multi-disciplinary research centre in the Faculty
of Social Sciences and Law at the University of Manchester.
Website: http://www.ccsr.ac.uk/