**This information has been copied across from the SCROL website and some of it may be out-of-date. This will be updated in due course**
Results from the 2001 Census are available for a wide range of area types, from the smallest, Output Area, to the largest, Scotland. The principal area types described below for which results will be made available have been decided in consultation with users.
The main building bricks for Census areas are output areas. All higher geographies or area types are built from output areas. The diagram below depicts how the different areas nest into higher areas, all derived from output areas. Any area for which Census output is produced is the aggregation of Output Areas (OAs) that approximate best to the area. OAs will aggregate exactly to a council area but not necessarily to any other type of area.
|Click on the image to enlarge|
All Census geography is based on the set of postcodes and their boundaries which were frozen in December 2000. Any postcode collected in enumeration that does not belong to this set was replaced during processing by the most appropriate frozen postcode.
Counts of the number of households with residents and the number of residents in each postcode are generated during processing. These counts are used to create Output Areas and are also published in their own right as a Census product on the postcode index.
The postcode to Output Area Index is a look-up table that indicates to which Output Area a postcode belongs. Download from the link above or contact Statistics Customer Services to obtain a copy of the Index.
Output Areas (OAs) for 2001 are created as groups of postcodes nesting as well as possible into the following areas: Council Area, 2001 locality, 1991 OA, postcode sector and 2001 electoral ward in descending order of preference (which has to be applied when not all postcodes in the OA belong to a single combination of these area types). The main aim governing this order of areas is to give continuity with the 1991 OAs while ensuring, as far as possible, that 2001 OAs fit into the locality or urban area which is seen as an increasingly important area type.
Each Output Area covers a sufficiently small area for user defined, or ad-hoc, areas to be created while maintaining a sufficient level of quality.
GROS creates only one set of OAs and defines all other output geographies using the OA as the building brick. Each OA will be assigned to an area in a 'higher' geography by first selecting one of the postcodes in the OA as a 'master' postcode. The OA inherits all of the characteristics of the master postcode including its assignments to higher areas and its centroid grid reference. Not all the postcodes in an OA may belong to the same set of higher areas as the master postcode. So an area aggregated from the postcodes assigned to it will not necessarily be the same when aggregated from OAs. It is also possible that the OAs assigned to a higher area do not form a single continuous area and it is divided by another area of the same type. Nevertheless the OA approximation is good enough for almost all purposes.
The OA is also the building brick for geographies based on either of the two 'remote' postcodes collected on the Census form: address one year ago and destination of travel to work or study.
An Output Area to Higher Area Index, and look-up table is available which provides a link between the OA and the 'higher' areas that the OA belongs to, enabling users to aggregate OA level Census results to 'higher' areas, such as Council areas or user defined areas. Download from the link above or contact Statistics Customer Services to obtain a copy of the Index.
Maps of OAs are not provided on the Census Data Explorer . For example, in the Standard Outputs section, at the point of selecting an area, the user may request a search by postcode and then click on a link that provides a map of the OA containing that postcode.
The Council Area is the main area for 2001 output. Council Areas were created on 1 April 1996 following a review of the local government structure in Scotland. The 32 Council Areas provide a single tier of local government covering the whole of Scotland. Each Council Area is divided into contiguous electoral wards. Council Area boundaries are contiguous with the country boundary between Scotland and England.
Parliamentary Constituencies are defined in terms of wards existing at the time of their definition. 2001 wards do not nest exactly into any of the current parliamentary areas: Scottish Parliamentary Regions, Scottish Parliamentary Constituencies, or Westminster Parliamentary Constituencies.
Westminster Parliamentary Constituencies are the same as Scottish Parliamentary Constituencies except for Orkney and Shetland which are combined into a single Westminster constituency.
Health Board Areas
In 1974, 15 Health boards were set up to administer the Scottish Health Service. Health Board areas have remained unchanged since their inception.
A postcode sector is the set of unit postcodes that are the same apart from the last two characters.
The postcode sector has been used in Census output since 1981. Special postcode sectors are created for Census output to ensure that sectors conform to a minimum threshold and do not cross Council Area boundaries. Because the confidentiality thresholds (method B) disclosure control differ for Census Area Statistics (CAS) and Standard tables (ST), there are two types of postcode sectors in Census output: Standard Table (ST) and Census Area Statistic (CAS).
Census Area Statistics (CAS) Sector - First, postcode sectors that cross council areas are split and each portion treated as a postcode sector in its own right. Then, as described above (using master postcodes), OAs area assigned to postcode sectors. The resulting 1,010 aggregations are denoted CAS Postcode Sectors (or CAS Sectors) and will meet the minimum threshold for CAS (20 households and 50 persons). CAS sector names that include '(part)' indicate that the original sector had to be split to nest within Council Areas.
Standard Table (ST) Sector Where a CAS sector fails to meet the minimum threshold for Standard tables (400 households and 1,000 persons) it is merged with one or more neighbouring CAS sectors within the same Council Area so that these thresholds are met.
The resulting 859 areas are called Standard Table (ST) Postcode Sectors (or ST Sectors or Standard Sectors). ST sectors that are mergers of CAS sectors are labelled as, say, 'DD1 1; DD1 3' with a semi-colon to indicate the merger.
The 32 Council Areas are divided by satute into 1222 electoral wards.
For Census output there are two types of wards, CAS and Standard Table (ST), based on the electoral ward. They are both created by aggregating output areas and do not necessarily fit electoral wards exactly. No census statistics are available for exact electoral wards.
CAS As described above (using master postcodes) OAs are assigned to electoral wards. Tthe resulting 1,222 aggregations are denoted CAS Wards and will fall within a Council Area boundary and meet a threshold of 20 households and 50 persons.
Standard Table (ST) Where CAS Wards fall below the ST thresholds (400 households and 1,000 persons) they are merged with neighbouring CAS Wards within the same council areas to exceed the threshold. It is also necessary to make a few adjustments to ST Wards so as to remove any 'slivers below ST threshold created by differencing ST Wards and ST Sectors. There are 1176 ST Wards.
ST Wards that are mergers of CAS wards are labelled South Ronaldsay; Holm and Burray with a semicolon to indicate the merger. For slivers, ST wards containing part, or sliver, of a CAS ward are labelled ending in part Innerleithen and Walkerburn; Peebles and District South (part).
A simple definition of a settlement is a collection of contiguous high density postcodes bounded by low density postcodes (or water). The first stage is to identify high density postcodes and the second is to group these into settlements retaining those groups expected to have at least 500 residents. For each postcode in Scotland (e.g. EH12 7TB) GROS have created and maintained a boundary containing its addresses. The number of addresses, both residential and non-residential, and the area in hectares is known for each postcode. Densities can be expressed as addresses per hectare; densities for both residential and non-residential addresses are calculated. A postcode is high density if either of the following is true
the density of residential addresses per hectare exceeds 2.1
the density of non-residential addresses per hectare exceeds 0.1
The second condition is included so that non-residential parts (e.g. industrial estates) of built-up areas can be identified. The threshold densities of 2.1 and 0.1 were found to give a good approximation to the built-up areas identified in previous Censuses using more traditional methods. These threshold densities were adjusted in two council areas (Eilean Siar and Shetland) where, because of crofting and other factors, settlement patterns vary considerably from the Scottish norm. Any holes of low density postcodes within a group of neighbouring high density postcodes were added to the group. Estimates of the population in each group were made and any groups considered to have fewer than 500 residents discarded. The estimates of population were consistent with GROS small area population estimates for mid-2000, which were themselves consistent with the Departments estimates of population for Council areas for 30 June 2000. More information about the creation of settlements is given in Scottish Settlements Urban and Rural areas in Scotland (GROS 2001, ISBN 1-874451-60-5), also available on the GROS website (www.nrscotland.gov.uk). This paper contains a description of the method used for the previous year (2000) using mid-1999 population estimates.
While settlements can go a long way in defining the towns and cities in Scotland, some are very extensive and group together very large populations. For example, the settlement of Glasgow in the 2000 publication mentioned above was estimated to contain 1,090,530 residents but no breakdown was given of the settlement into any constituent towns or cities such as Airdrie or Paisley. Accordingly, for 2001 GROS divided the larger settlements into localities using as a basis the areas so designated in the 1991 Census report Key Statistics for Localities in Scotland (GROS 1995, ISBN 0-11-495736-3) (Note that this publication is now out of print. It should be available in main reference libraries).
Census results are also available for Scotland outside settlements. This area is denoted Rest of Scotland. The term rural is not used for Rest of Scotland. To do so would cause confusion given various other definitions of rural Scotland that exist, some of which are based on settlements created by GROS (e.g. areas outside settlements containing at least 10,000 residents).
Although Civil Parishes lost their former administrative function in 1929, they have changed very little since the 20th Century and Census data has been produced for these areas for all Censuses from 1891. There are 871 Civil Parishes.
For the 2001 Census, Civil Parishes are a best-fit aggregation of 2001 output areas (OAs). OAs are each given a centroid a geographical point within the boundary of the OA in order to assign the OA to a civil parish or other such area. When the centroid of an OA falls within the boundary of a civil parish, the OA is assigned to that parish. The statistics for the parish are produced by aggregating the statistics of the OAs assigned to it. A few small parishes have no OAs assigned to them so their statistics are combined with neighbouring parishes and they do not appear separately in output. These parishes are:
|388||Hoy and Graemsay|
Some figures for civil parishes from the 1991 Census have been revised in producing Key Statistics. Tables KS16 and KS18 contain comparisons with the 1991 Census. The 1991 figures used are produced in the same way as those for 2001 i.e. by using centroids of 1991 OAs to assign them to parishes. The 1991 figures used for KS16 and KS18 do not always agree exactly with those published in, say, the 1991 Census Monitors for Wards and Civil Parishes. The reason is that the process of allocating 1991 OAs to civil parishes for the 1991 Census Monitor was clerical (and subject to error) whereas that for 2001 output was automated.
An island is a mass of land surrounded by water separate from the Scottish mainland. An island connected by a bridge or causeway or ford to other land masses is treated as a separate island i.e. man-made connections are ignored for the purpose of the Census.
At the time of planning the Census, there were 117 islands known to receive mail of which 95 were recorded as having at least one resident during enumeration. The processing of the collected data revealed a further inhabited island (Holy Island) making a total of 118 individual islands, with 96 of these inhabited by at least one resident.
The Process of creating Output Areas resulted in the smaller islands (in terms of population) becoming merged with neighbouring islands (or with the mainland where this was not feasible). Census output is produced for 54 islands or groups of islands that meet the confidentiality thresholds for Census Area Statistics. Click here [link] to see how islands have been grouped.
Census results are also available for the residual part of Scotland which is denoted "Mainland of Scotland".
GROS has published a paper on the inhabited Islands, in its series of Occasional Papers. This paper gives, among other things, details of residents and households on the 54 island groups as well as counts of residents and households for all inhabited islands, regardless of size.
The area of each Output (OA) in hectares forms part of the OA to higher areas index information and are derived from the OA digital boundaries. The area in hectares of higher areas is aggregated from OAs, except for the area of Scotland, the council areas and the health boards. Ordnance Survey supplied the hectarage of Scotland, the council areas and consistent figures for the health boards.