You are in :

POLAR: A short guide

What is POLAR?

POLAR - Participation Of Local Areas - is a series of maps showing the participation of young people in higher education (HE) for geographical areas ranging from regions to wards. Some maps also show the size of the young population for each area or the nature of local secondary schools. These results are also provided as tables and data sets.

These young participation rates are calculated using new methods developed by the Analytical Services Group within HEFCE during the course of our research into participation in HE. The findings of this work are described in full in the HEFCE report 'Young participation in higher education' (2005/03). The report sets out the national patterns of young participation and uses groups of small areas to look at trends in the participation of young people living in advantaged and disadvantaged areas. POLAR complements the national patterns in the report by showing how the country is divided up into high and low participation areas.

POLAR history

POLAR was developed from the early results of the work reported in 'Young participation in higher education'. It was made available in September 2002, through a user account system on the HEFCE web-site, to people involved in work relating to young participation in higher education. At that time, significant resources were being committed to projects aiming to reduce participation inequalities. However, there was a conspicuous absence of reliable data on young participation rates, especially for areas smaller than regions. The early release of these results was an effort to provide a set of nationally consistent estimates of young participation for those who had an urgent need of such information to guide widening participation activities.

The information available through POLAR has subsequently been widely used by those working on young participation projects, and has also been used in some resource allocations for widening participation activities. Comments from users indicate that the POLAR results have been helpful in their work, and are consistent with local knowledge.

With the local results verified, and the method and context of the local participation rates set out in the 'Young participation in higher education' report, it is no longer necessary to restrict access to the POLAR resource. Accordingly, POLAR has been made available to all, so that it can act as a local information supplement to the participation patterns in the report.

Young participation

The statistic used by POLAR is young participation. Put simply, this is the proportion of young people in an area (the 'cohort') who go on to enter higher education aged 18 or 19. For these maps, POLAR defines higher education as full-time first degree, or Higher National Diploma/Certificate (HND or HNC) courses at any UK higher education institution or GB further education institution.

The true extent of inequalities in participation only becomes evident for small areas such as wards. However, these areas have small numbers of young people - a typical annual ward cohort is around 50. This means that the random variation in the observed annual participation rates is proportionally large. To increase the reliability of the participation rates, especially for small areas, the participation rates used in POLAR are formed by using three consecutive cohorts. These are children who had their final year of compulsory secondary education in 1994-95, 1995-96 and 1996-97 and then went on to enter higher education 3 or 4 years later aged 18 or 19 respectively. The cohorts are usually referred to by the year in which they would be 18 years of age: POLAR uses the 1997, 1998 and 1999 cohorts.

The count of young entrants is obtained from administrative records covering students studying full-time higher education courses in either higher education institutions or further education institutions. The records of individual students are linked through time and across different data sources to give a comprehensive and robust count that eliminates double counting. More details of this method are given in the 'Young participation in higher education' report, particularly Annex C.

The estimates of the size of the cohort in each area are produced using a specially developed combination of child benefit data and 1991 Census small area statistics. This method is described in detail in Annex A to the 'Young participation in higher education' report.

The entrants and cohort estimates are then combined geographically, using the All Fields Postcode Directory maintained by the Office of National Statistics (ONS). The resulting participation rates are mapped using boundary data from the UKBORDERS service (http://edina.ac.uk/ukborders/).

School information

The ward-level maps also show information about secondary schools (with the exception of special schools and some very small schools). Each school is shown by a coloured symbol. The symbol shape indicates the type of school; the size (area) of the symbol is proportional to the size of the GCSE cohort; and the colour indicates the proportion of these children who obtain 5 or more A-C grades at GCSE.

The school statistics are drawn from the Government's School and College Achievement and Attainment Tables (http://www.dcsf.gov.uk/performancetables/) and only cover England. The school results shown are based on the three GCSE result years 1995, 1996 and 1997: they therefore reflect the GCSE results of the same cohorts for which POLAR measures young participation. This allows an assessment of whether there is any spatial association between school GCSE results and area participation rates. However, this does mean that the school data used do not reflect any subsequent improvement or deterioration in GCSE results. Similarly, results may be shown for schools that have since closed, and newly opened schools will be omitted. If more recent school performance information is required then this should be obtained from the school performance tables directly.

Geography used

POLAR shows the broad pattern of young participation using four large geographical units: Government Office Regions, Learning and Skills Councils (LSC) areas, local education authorities (LEAs), and parliamentary constituencies. Regions, LSC areas and LEAs are used as they often reflect the operational structure of the organisations involved in widening participation. Constituencies can be particularly useful for planning participation work as they show a wide range of young participation rates but, unlike other units, are all of similar sizes in terms of their young population. The range of cohort size and participation rates for these units is shown in Figures 25 and 26 in the 'Young participation in higher education' report.

Wards are used as the geographical unit for the maps of local participation as our work has suggested they have a number of useful properties. They show a wide range of young participation rates that captures much of the local variation in participation. At the same time, by combining cohorts together they have sufficient numbers of young people so that the observed participation rates are not rendered meaningless by random fluctuations. Results in the participation report suggest that wards that are internally very mixed in terms of young participation rates do occur - but only rarely. In particular, pockets of low participation in high participation wards are unusual, and the overwhelming majority of low participation micro-areas are found within low participation wards. Finally, wards are contiguous areas that are large enough to typically refer to a recognisable named neighbourhood, which makes them easy to use.

Two types of wards are used in the maps, 1991 Census wards and 1998 electoral wards (note that the participation rates shown are the same, those of the 1997, 1998 and 1999 18 year-old cohorts, regardless of the boundaries used). The 1998 electoral boundaries have been created as best-fit aggregations of 1991 Census enumeration districts. Since 1991 enumeration districts do not match exactly to 1998 wards, the boundaries can be quite eccentric and suffer from anomalies such as small detached areas. Because of this, we would recommend using the 1991 boundary maps. The 1998 ward boundaries may be of use if there has been substantial local reorganisation, or if the results are to be used in conjunction with the 2000 Index of Multiple Deprivation (which uses 1998 ward boundaries). Both the 1991 and the 1998 boundaries have been simplified for clearer presentation and to reduce file sizes.

Cartograms

Cartograms are provided as an alternative to the geographical maps. In these, each area is replaced by a circle, the area of which is proportional to the number of children in the cohort. The circles are then re-arranged by an algorithm that tries to account for the new sizes of each area whilst maintaining the original geographical relationship between each area as much as possible. Finally, the circle is colour-coded to show the level of participation as in the standard geographical maps.

The advantage of this presentation is that the visual prominence of a region is related to the size of its young population rather than its geographical size. This can be helpful for planning, since it emphasises the number of children in each area. This is particularly important when looking at the participation of young people: physically small but densely-populated urban wards often have substantial numbers of children, and some of the lowest HE participation rates, yet can be nearly invisible on a standard geographic map.

Using POLAR

The home page of the POLAR web-site contains links to index pages for the national and regional maps and tables, as well as to information sources such as this guide. At the top of each page is a link back to the POLAR home page.

The national index pages have direct links to the maps, tables and data sets. The regional index provides links to maps of the large units by region and also links to further regional index pages of local participation maps. These regional index pages give links to the ward-level maps for each LSC area (or LEA area in Scotland and Wales).

All the POLAR maps are presented as portable document format (PDF) files. This gives small file sizes and, when used in conjunction with a viewer such as Adobe Acrobat Reader allows the files to be viewed without the need for specialised software. In particular, users may find the zoom, hand tool (for panning across the map) and text search functions helpful in navigating the maps. All the maps have titles and footnotes which briefly summarise what is shown.

The maps themselves use colours to indicate the different levels of participation. There are always five levels, progressing from blue (high participation) through green (average participation) to red (low participation). The key (shown in the upper right hand corner for the geographic maps and the lower right for the cartograms) links these colours to ranges of young participation. These ranges are set so that approximately equal shares (by young population) of areas are assigned to each band. Thus the five levels represent approximate quintiles of the young population (exactly analogous to the quintile groups used in the participation report). The variation in young participation between large areas (such as LEAs) is much less than between local areas (such as wards): therefore the participation bands are specific to the geography shown, and much narrower for the maps of large areas.

A similar colour scheme is used to indicate bands of GCSE results in the ward-level maps that show schools. In this case, the bands are set so that approximately equal numbers of the cohort are in schools assigned to each band. Circles are used for non-selective state schools, diamonds for selective state schools, and squares for independent schools. The size (area) of the school symbol is proportional to the number of GCSE-year children.

The cartograms share the same colour scheme and bands as the geographical ward maps. There is an additional key which relates the size of the circles used to the young population in each area.

Tables and data sets

Tabular information, as both web tables (not for wards) and downloadable data files, are provided for those who want the information shown in the maps in a form more suitable for calculations. These tables give the mean annual cohort size (over the three cohorts used in POLAR) and the young participation rate for each area. For the regions the cohort size is rounded to the nearest thousand; for LSC areas, LEAs and parliamentary constituencies to the nearest hundred; and for wards to the nearest ten. For regions, LSC areas, LEAs and parliamentary constituencies the young participation rate is reported to the nearest whole number per cent. For the wards the quintile bands from the maps are used.

The web tables can be viewed online or copied into spreadsheet applications. The data files are presented in comma separated variable (.csv) format, with quoting of character fields, and can be read directly by most spreadsheet and data base applications. The first three lines of each file form a header giving information about the data set, the fourth line describes the variables. Each data record contains the name of the area, the name of any higher areas (for example, the region name is provided along with the LSC area name in the LSC files), and the rounded values for the cohort and for young participation.

Help

The website contains a list of frequently asked questions (FAQs). In addition, as already mentioned, the 'Young participation in higher education' report provides more information. For example, the report covers issues such as why we have focused on young entrants and why we have used areas, as well as putting the POLAR quintiles in the context of national results and giving a detailed account of the method. For other questions, comments or suggestions please e-mail m.corver@hefce.ac.uk.

Last updated 16 July 2007