Scottish Island Regions (2023): Data Zones
This is a brief guide to best-fitting data zones to the Scottish Island Regions (2023) – a new geography for grouping Scotland’s islands into nine different island regions.
Data zones have been best-fitted to the Scottish Island Regions geography. This makes it possible to analyse existing data based on data zones, and allows island data to be compared between island groups and with the rest of Scotland.
This document is organised into the following sections:
1. What is a data zone?
2. Best-fitting data zones to Scottish Island Regions
3. Quantifying the quality of the best-fit of data zones
5. Other resources
What is a data zone?
Data zones are the key geography for the dissemination of small area statistics in Scotland. Composed of aggregates of Census Output Areas (OA), data zones are large enough that statistics can be presented accurately without fear of disclosure and yet small enough that they can be used to represent communities. They are designed to have roughly standard populations of 500 to 1,000 household residents, nest within local authorities (at the time of the Census), have compact shapes that respect physical boundaries where possible, and to contain households with similar social characteristics.
Aggregations of data zones are often used to approximate a larger area of interest or a higher level geography that statistics wouldn’t normally be available for. Data zones also represent a relatively stable geography that can be used to analyse change over time, with changes only occurring after a Census.
A problem for the Scottish Island Regions geography is that some of the current data zones cover both islands and mainland Scotland, or straddle more than one Scottish Island Region, meaning that it is not possible to get an exact fit to islands. This means that it has been difficult to use existing data to compare Scottish islands to the Scottish mainland. This report explains the process of best-fitting data zones to Scottish Island Regions, and attempting to quantify the quality of the best-fit based on Census 2011 population data (available at postcode level in the Postcode to Output Area (OA) 2011 Census Index).
Best-fitting data zones to Scottish Island Regions
The standard way to best-fit data zones to higher geographies is by following Pillar 4 of the GSS Geography Policy, which states that statistical building blocks (e.g. data zones) should be best-fitted by allocating whole building blocks based on the locations of their population weighted centroids.
However, for Scottish Island Regions, it is recommended that this policy is not followed in one instance. The centroid of data zone S01007320 falls on Eilean Loain, which is allocated to the Argyll Islands Scottish Island Region. However, based on Census 2011 data, it is evident that almost the entire population of this data zone lives on the mainland. Therefore, to reduce the difference associated with the best-fitting, it is advised not to follow the population weighted centroid method of allocation, and instead to remove this data zone from the lookup.
A data zone to Scottish Island Regions lookup is published with the geography boundaries on the spatial data metadata portal.
Quantifying the quality of the best-fit of data zones
In some cases, the entire data zone will fit within a Scottish Island Region. However, in other cases, the data zones straddle either more than one Scottish Island Region, or they straddle a Scottish Island Region and the mainland of Scotland (as described on the previous page).
It is possible to quantify the quality of fit of data zones by comparing the population enumerated within each Scottish Island Region on Census Day 2011 (using postcode-level population estimates) with the population which is allocated to the Scottish Island Regions when built from data zones.
The Census 2011 postcodes fitted exactly within the islands which were inhabited at the time of the Census, and therefore can be used to show the ‘true’ population of the Scottish Island Regions on Census Day 2011.
The postcode-level population estimates are available from the Postcode to Output Area (OA) table within the Census 2011 Indexes published by National Records of Scotland (NRS).
The 2011 Frozen Postcode Grid References are one of the Census 2011 Boundaries published by National Records of Scotland (NRS) and can be used to allocate postcodes exactly to the inhabited islands (in this case, the definition of inhabited is all islands which contained a Royal Mail Unique Delivery Point Reference Number in the January 2011 version of the Postcode Address File - which was used as the basis for the 2011 Census ‘frozen’ postcodes).
For eight of the nine Scottish Island Regions, the quality of fit of data zones is considered to be reasonable. However, it should be noted that there is a difference of around 15% for the reported population of the Shetland - Outer Islands Region.
This is caused by allocating data zone S01012397 to the Shetland - Mainland and connected Region (which is where its population weighted centroid is located). 44% of the data zone’s population falls within Bressay, which is in the Shetland - Outer islands Region. This population is therefore ‘missing’ from the Shetland – Outer islands and equates to a 15% difference when expressed as a percentage of the actual population for this Region (see calculations below).
Calculating the difference for the Shetland Outer islands
Actual Pop of Shetland Outer Islands (3,285) - Reported Pop of Shetland Outer Islands (2,794) = ‘Missing’ population (491)
‘Missing’ population (491) / Actual Pop of Shetland Outer Islands (3,285) * 100 = Difference in Reported Population (15%)
The new Scottish Island Regions geography can help to reframe existing data and collect new data. Having more data on islands will help our understanding of the issues faced by the population of Scotland’s islands. This new geography allows comparability in data between the Scottish islands and the Scottish mainland.
The Data Zones to Scottish Island Regions (2023) look-up is available via the spatial data metadata portal. It can be used with any dataset which has a Data Zone geographic identifier.
The Geographic Information Science & Analysis Team (GI-SAT) at the Scottish Government have published boundaries and assigned S-codes to the new Scottish Island Regions geography, enabling it to be used by a wider audience. These are available on the spatial data metadata portal and via the Scotland Register of GSS Codes.
Interactive charts are available, showing population change by age and sex, for the Scottish Island Regions. Mid-year population estimates for Scottish Island Regions are also now available from National Records of Scotland.
The Scottish Island Regions are based on the islands region framework initially created by the James Hutton Institute for the National Islands Plan Survey (2020). For details, see: National Islands Plan Survey - Final Report.
The researchers would like to thank Jonathan Hopkins, James Hutton Institute, for developing the original framework while conducting the Scottish National Islands Plan Survey (2020) on behalf of the Scottish Government. This survey was led by Ruth Wilson and Jonathan Hopkins. We are also grateful to the National Islands Plan Survey Research Advisory Group for their feedback on the initial framework, and for the work of Q-Step interns Joanna Mroczka, Andriani Klefti and Sean Choon in matching the framework to Data Zones. Q-Step is funded by the Nuffield Foundation and Economic and Social Research Council (ESRC).
There is a problem
Thanks for your feedback