Methodology

Participation measures in higher education

Published

Methodology and coverage

The Cohort-based Higher Education Participation (CHEP) measure identifies first-time participation in Higher Education by age, gender, region, and level and mode of study. 

CHEP replaces the previous Higher Education Initial Participation (HEIP) measure that had been published by DfE (and former Departments) since 2004.  

Coverage

The denominator used for the CHEP measure is pupils who attended state-funded schools and special schools in England at age 15 as recorded in the School Census. The numerator is calculated by identifying pupils who progressed to UK Higher Education (Levels 4 and above) in the years following their schooling.

The measures are calculated using matched data. This matches the National Pupil Database (NPD) to the Education and Skills Funding Agency (ESFA) Individualised Learner Record (ILR) and the Higher Education Statistics Agency (HESA) Student Record and Alternative Provider Student Record. This allows pupils to be tracked from English schools at age 15 to higher education (HE) at older ages.

The matched data relates to students who were aged 15 in the period 2001/02 to 2018/19, which allows us to track HE entry up to age 34 in the latest year (2020/21). 

The measures cover HE courses at UK Higher Education Institutions, Alternative Providers and English Further Education Colleges. 

It must be noted, due to the matching procedures deployed, all figures in this publication should be treated as estimates, and may include a small amount of double counting where pupils attend more than one school. Further details of the matching procedure can be found in the Matching process section below.

Independent schools

Pupils who attended independent schools are not currently included in the measure. The Department collects less detailed information on pupils who attend independent schools. This means that the figures are more sensitive to changes in the matching methodology over time than for state-funded schools which then affects the comparability of the figures.  We will keep monitoring the stability of the matched independent schools data in the future with a view to including them.

Higher Education

Pupils are deemed to have participated in Higher Education where they are:

  • Recorded in HESA's Standard Registration Population, which requires a minimum of 2 weeks participation; or
  • Recorded in the ILR college data as starting a learning aim at level 4 or higher that was either:
    1. Completed successfully
    2. Continuing into the following year
    3. Lasting over 6 months

Participation in HE at Alternative Providers has been included for students who started courses in 2014/15 or later.

Work-based learning, including apprenticeships, is not included in the measures.

16 year olds in Higher Education

Due to the small numbers involved, figures for students who entered HE at age 16 have been included in the figures but categorised as 17-year-olds.

Breakdowns

Statistics are provided for the following breakdowns:

Age

Age relates to the age at the start of the academic year. Figures are provided for individual ages and are also grouped to show HE entry by age 20, 25 and 30. The main focus of the publication is HE entry by age 25 as this provides a balance between the timeliness of the measure and the peak participation in HE.

Gender

Gender is recorded as male or female in the School Census. Gender is self-declared and recorded according to the wishes of the parent and/or pupil.

Region

Region is based on the location of the school that the pupil attended at age 15. The location of the school is determined by the Local Authority (LA) it reports to. In some cases, pupils may have lived in a different area to the one they have been reported in, including some pupils attending schools in England who lived in Wales or Scotland.

Mode

Mode of study refers to whether a student has undertaken full-time or part-time higher education.

Level

Level of study refers to the qualification that the individual was aiming to achieve when they set out in their studies.

Matching process

The National Pupil Database (NPD) contains administrative data on all pupils in schools in England, collected by the Department for Education. Pupil Level Annual School Census (PLASC) records were matched to the Higher Education Statistics Agency’s (HESA) Student Record and Alternative Provider Student Record, and the Skills Funding Agency (SFA) Individualised Learner Record (ILR).

The matching process allows for school pupils to be tracked through to Higher Education. The match achieved is called a “fuzzy match” where we rely on names, postcodes, dates of birth, etc and there is some potential for minor errors in the matching process. For this reason, the figures are deemed estimates due to the reliability of the matching procedure used.

Limitations

The measures provide a generally consistent time series, however there are some factors that will impact on this.

The publication relies on “fuzzy matching” which matches data based on the personal details collected in each dataset. Match quality and match rates will change over time as new data is added and improved methodologies applied.

These changes will impact the figures for the percentage of pupils entering HE presented in this publication. Calculations involve a very long run of data dating back to 2001/02. If the match rate increases due to methodological changes this could result in an increased CHEP rate even if there has been no behavioural change in terms of participating in HE. 

The match quality and rate are likely to be more volatile for pupils where fewer personal details are known such as for pupils who attended independent schools and pupils who have lived outside England or changed address between compulsory schooling and prior to entering HE (which is more likely to be the case for older students). We therefore chose to exclude independent school pupils from the coverage of this release.

Our analysis suggests that there are specific years that are more affected by changes in the matching than others:

  • The match to HESA data did not include 16 and 17-year-olds prior to 2014/15
  • The filters used to remove under 18s from the HESA data in 2008/09 were incorrectly applied, which led to a small number of 17-year-olds being matched and some 18-year-olds not being included in the match in that year. For 18-year-olds this is likely to have reduced the number being matched by up to around 10%, although the vast majority will have participated in HE at age 19 in 2009/10 and so the impact on cohort measures will be limited. For 17-year-olds the numbers are very small and so the key measures used are not affected, however it has caused a large proportional increase in the number entering HE at age 17 in 2008/09.  
  • Changes in the matching methodology in 2018 have led to an increase in the number of HE entrants that matched to a school census record for HE entry from 2017/18 onwards where they were categorised in their HE records as not being domiciled in England immediately prior to study. We believe the numbers involved are relatively small but will have some impact on comparisons between figures looking at HE entry in 2016/17 and later years. Overall, we are confident these changes have minimal impact on the robustness of the CHEP-25 rates.

Although the school census provides a largely stable population base, it is not an estimate of the population of 15-year-olds. Changes in the number of pupils who are home-schooled or who attended independent schools will have an impact on the comparability of the figures over time. The impact is minimal for the cohorts in this release, but will become more apparent in future as home-schooling has increased since the Covid-19 pandemic.

These figures do not include participation in HE in colleges in Scotland, Wales and Northern Ireland or any HE outside the UK and so will not provide a complete picture of participation in HE.

Difference from previous measure

The new ‘cohort’ measure

This year we have transitioned to reporting higher education participation using a Cohort-based Higher Education Participation (CHEP) measure, which tracks a 15-year-old school cohort over time until they participate in higher education.

The previous ‘projected’ measure

The previous measure summed together age-specific initial entry rates recorded in the latest year to estimate a Higher Education Initial Participation (HEIP) rate. This estimate was described as a projection of the likelihood that a 17-year-old would participate in higher education by age 30 if current participation levels persisted into the future.

The denominator for the previous measure used mid-year population estimates published by the Office for National Statistics (ONS) which had been adjusted to an academic year basis, rather than a 15-year-old school cohort.

Rationale for change

HEIP produced a timely measure but with some known limitations:

  • When there is steady growth in entry rates for younger age groups (as has been observed over many years for English 18-year-olds), the HEIP method of summing current participation rates will estimate a higher participation rate than the real rate for a particular entry cohort. For example, today’s 30-year-olds will have had a lower initial participation rate when they were aged 18 compared to today’s 18-year-olds 
  • It was not possible to create reliable figures broken down by region and key demographics using the projection methodology 

We now have a long enough series of longitudinal data to justify switching to the new cohort-based measure (CHEP).  As these participation rates are based on tracking a clearly defined school cohort, it lessens the impact of both inward and outward migration flows over time resulting in a more stable base population. It is also not affected by revisions to ONS population estimates that occur following each Census.

Using the new cohort-based approach means we can analyse participation by pupil characteristics taken from the school census, including breakdowns by specific ages, gender, and region of school attended. Under the previous method it was not possible to create reliable figures by region.

The focus of the breakdowns in this publication is participation by age 25. CHEP is less timely than HEIP as it requires waiting for each 15-year-old school cohort to have reached age 25 before inclusion in the CHEP-25 measure. However, the figures reflect actual participation rather than projected participation, so they provide a more accurate picture of entry to higher education.

Relevance to users

Why publish a participation rate?

Participation rates have been published by the department in different forms over many years. They help us to understand how the proportion of the population participating in higher education is impacted by changes in the economic, social and policy environment. Up until publishing statistics for the 2019/20 academic year, the Department measured participation by age 30 using the Higher Education Initial Participation measure (HEIP). As we publish statistics for the 2020/21 academic year, we have transitioned to measuring participation using the Cohort-based Higher Education Participation measure (CHEP).

Why was the HEIP measure introduced?

In the National Statistics Quality Review of 2003, Professor Brian Ramsden surmised that:

"…after the 1997 general election, there was increased policy focus on the idea of "lifelong learning". At that time, the main Departmental measure of HE participation was the Age Participation Index (API), but its restricted coverage (full-time, aged under 21) led the (then) DfEE Analytical Services Branch to start developing new measures, which would for example include the take up of HE by mature students, part-time study, postgraduate study and by students returning to HE in a context of lifelong learning."

The HEIP measure rose in prominence and the public interest as it was used to measure the former Labour Government's Public Service Agreement (PSA) target to progress towards 50% participation in higher education.

Limitations with the HEIP methodology 

When there is steady growth in entry rates for younger age groups (as has been observed over many years for English 18 year-olds), the HEIP method of summing current participation rates will show a higher participation rate than the participation rate for a particular entry cohort. For example, today’s 30-year-olds, will have had a lower initial participation rate when they were aged 18 compared to today’s 18-year-olds. 

A rough estimate of rates for particular entry cohorts can be constructed by summing the initial entry rates across academic years. For example, the 17 year-old entry rate in 2006/07, would be added to the 18-year-old rate in 2007/08, and to the 19-year-old rate in 2008/09, and so on.  

Why transition to the CHEP methodology?

Since passing this milestone of 50% participation, in the 2018/19 publication DfE asked users for feedback on the future of the series.

Feedback was invited on whether the HEIP measure remained relevant and was still necessary to users. This included opportunity for comment on how users would be impacted if methodological changes were made to the series. 

The decision was taken to retain a higher education participation statistical series, but for the 2020/21 academic year the headline measures would be developed from tracking cohorts of school pupils rather than relying on population estimates. 

The cohort measure aims to be: 

  • More accurate. The new method measures participation by a clearly defined school cohort, rather than current population estimates that are subject to inward and outward migration flows.
  • More intuitive to users. Cohort tracking is easier to explain than the existing projection methodology.

Cohort measures also benefit from having a reliable estimate of the population across a wide range of personal characteristics and geographic breakdowns such as those published in Widening participation in higher education, 2020/21. Following the change in methodology, from the 2020/21 academic year we have included a regional breakdown in the statistics for the first time.

As illustrated in the Comparison with the previous measure section of the publication, cohort measures can be used to estimate participation by age 30 based on ‘current’ participation levels on a similar basis to the HEIP measure. Projection estimates can be derived by considering the sum of changes in entry rates by age between the two latest academic years.

Feedback 

DfE continues to welcome feedback on how the statistics can be developed to become even more relevant to users at he.statistics@education.gov.uk.  We are open to making further characteristic breakdowns available where there is a strong business case.

Timeline of methodological changes

In December 2006, the National Audit Office published a review of the data systems underpinning target measurement in a number of Government Departments. The HEIP measure methodology was reviewed as part of this process. The report is available on the National Audit Office website, at:

Third Validation Compendium Report - National Audit Office (NAO) Report

The methodology for calculating the HEIP measure was revised in 2007, in line with recommendations set out in the National Statistics Quality Review and the recent National Audit Office review. The details of the revisions and their impact on the time series were reported in Statistical First Release 03/2007, “Methodological Revisions to the Higher Education Initial Participation Rate (HEIPR)”, which is available at:

[ARCHIVED CONTENT] Methodological Revisions to the Higher Education Initial Participation Rate (HEIPR) - Data, research and statistics (nationalarchives.gov.uk)

The methodology for calculating the HEIP measure was further revised in 2009 due to a change in the underlying data. Details of the revisions were reported in the Statistical First Release 03/2009, “Participation Rates in Higher Education: Academic Years 1999/00-2007/08 (Provisional)” which is available at:

[ARCHIVED CONTENT] DIUS: Participation Rates in Higher Education: Academic Years 1999/2000-2007/08 (Provisional) (nationalarchives.gov.uk)

Details of the impact of the revisions on the HEIP time series were detailed in an appendix to the main Statistical publication which is available at:

[ARCHIVED CONTENT] DIUS: Participation Rates in Higher Education: Academic Years 1999/2000-2007/08 (Provisional) (nationalarchives.gov.uk)

In November 2020 the HEIP measure had been boosted by the inclusion of data from Alternative Providers (APs) of Higher Education.  The Higher Educations Statistics Agency (HESA) Alternative Provider Student record was first collected in 2014/15. These data were incorporated into the HEIP measure in 2018/19.

In November 2022 DfE wrote to the Office for Statistics Regulation (OSR) seeking to suspend the National Statistics designation of Participation Measures in Higher Education due to the planned substantial re-development of the publication. The statistics published in January 2023, which include higher education participation up to the 2020/21 academic year, are derived by tracking 15-year-old school cohorts over time.

Previous HEIP measure - Methodology

Previous Methodology

This methodology relates to the previous HEIP measure that was published up to the 2019/20 academic year

Sum of age-specific participation rates

The HEIP measure was a sum of the participation rates for each age from 17 to 30 inclusive. For each age from 17 to 30, the initial participation rate was calculated as the fraction of the academic year population that are initial entrants. These rates were added to create the total HEIP measure. The HEIP measure was not a sum of the total number of initial entrants divided by the total academic year population. Doing this would have made the false assumption of an equal likelihood of participation across all ages. Further information about the previous methodology can be found in the National Statistics Quality Review of Higher Education participation statistics (2003) at: 

[ARCHIVED CONTENT] National Statistics Quality Review of the Initial Entry Rate into Higher Education - The Department for Education (nationalarchives.gov.uk)

Numerator

The numerator for each age-specific rate included English domiciled entrants to higher education who participated for a minimum period of six months for the first time.  Minimum participation of six months was the recommendation of Professor Brian Ramsden in the National Statistics Quality Review (2003); he stated that there ought to be an: ‘“actual minimum length of study” of, say, six months, since this is a better measure of “participation” in the sense of engagement with the Higher Education experience.’

Available data allowed participation for a minimum of six months to be measured for participants at UK higher education providers (HEPs), and at English, Welsh and Scottish Further Education Providers (FEPs). Students at FEPs were included if they were registered on courses designated as National Vocational Qualification (NVQ) level 4 or above, or were listed as Higher Education courses. 

Data referring to Higher Education Providers (HEPs) in the United Kingdom was collected by the Higher Education Statistics Agency (HESA). For Further Education Providers (FEPs) in England, data was collected by the Education and Skills Funding Agency (ESFA), in Wales by the Welsh Government, and in Scotland by the Scottish Funding Council (SFC). 

Since 1994/95 HESA has collected data on HE participation from publicly funded UK Higher Education Institutions (HEIs) in the annual ‘Student' record. Since 2014/15 information was also collected about undergraduate students at Alternative Providers (APs) of Higher Education in England in the Student Alternative record. Since 2017/18 the AP record has also included postgraduate students at all APs. Once DfE was assured of the robustness of the newer data from APs, they were added into the HEIP measure. 

There is less of a distinction between HEIs and APs under the new regulatory regime in England, so student counts for these providers were combined into a ‘Higher Education Providers’ (HEPs) category in the previous publication. 

In order to ensure that only initial participants are counted in the HEIP in a given year, algorithms were applied to match back over 12 years for HE data and FE data . This was in addition to filtering based on qualifications declared at entry. This matching looked back over the previous years’ HE and FE datasets and checked whether a student appeared in these datasets, recorded as studying at HE level for a minimum period of six months. If so, they were eliminated from the calculations, as they were not initial participants in the current year. Matching the 2017/18 figures back over 12 years ensured that any 30-year-old in the dataset, who had at least six months’ prior HE experience, could be detected all the way back to when they were 18-years-old. The matching algorithms were provided to DfE by the Office for Students (OfS). 

The academic year 2007/08 was the first year for which this approach was possible (as 12 years’ worth of HE data was available up to that point). In order to provide some time series comparison, and to demonstrate the impact of this change, the 2006/07 figure was also calculated using this methodology – although this used one less year of matching due to data availability. 

Denominator

The denominator for each age-specific rate used population estimates for England, which were acquired from the Office for National Statistics (ONS) at: 

Analysis of population estimates tool - Office for National Statistics (ons.gov.uk)

Population estimates were adjusted to an academic year basis using their data on the month of birth for each cohort.

Measurement of initial participation by over 30s

Age-specific participation rates for 17 to 60 year-olds were not as robust as the figures for the HEIP measure for 17 to 30 year-olds itself, because it was not possible to check as thoroughly whether participants aged 31 to 60 were initial participants. The HEIP estimate for entrants up to age 60 was however included in the publication to provide additional context on lifelong learning and enable comparability with statistics published by other countries.

Previous HEIP measure - Coverage

Previous Methodology

This methodology relates to the previous HEIP measure that was published up to the 2019/20 academic year

Initial participants included 

The HEIP measure includes 17 to 30 year-old English domiciled first-time participants in higher education at UK higher education providers (HEPs), and at English Further Education Providers (FEPs). Students at FEPs are included if they were registered on courses designated as National Vocational Qualification (NVQ) level 4 or above, or are listed as Higher Education courses. 

The coverage of the HEIP measure had been boosted in last year's publication by the inclusion of initial participants registered at additional providers of higher education who have been returning data to the Higher Education Statistic Agency's (HESA) Student Alternative record. This impacts the HEIP measure from 2014/15 onwards, and the Postgraduate Initial Participation (PGIP) measure from 2017/18 onwards. 

The impact of the change in coverage is shown in the headline chart and table of the publication. In the remainder of the publication figures are presented with a break in the series to make clear that figures prior to 2014/15 are on a different basis.

The HEIP measure includes students studying via both Full-time (FT) and Part-time (PT) modes. FT study is defined by the Office for Students (OfS) as attendance at an institution for periods amounting to at least 24 weeks per year, with learning activity amounting to an average of at least 21 hours per week. The PT mode refers to study which is less intensive than this definition. There is no lower limit on the intensity of PT courses for inclusion in the HEIP. 

Initial participants not included 

The HEIP measure does not include English domiciled Higher Education students: 

  • at Welsh and Scottish FEPs
  • at FEPs in Northern Ireland;
  • at institutions outside the UK. 

Information on these students is not currently available to the Department in sufficient detail to be included in the calculations. 

Students taking courses outside the UK

Whilst sufficient detail to include initial participants studying wholly overseas in the HEIP measure is not available, we have investigated how much these groups could contribute to the HEIP estimate. 

The United Nations Educational, Scientific and Cultural Organization (UNESCO) produces figures on the
number of students studying wholly overseas by country of origin on their ‘Global Flow of Tertiary-Level
Students’ webpage. Latest data shows that there were an estimated 39,504 UK students studying wholly
overseas in 2018/19. This can be accessed at the following link: 

UIS Statistics (unesco.org) under side menu filters “EDUCATION”, “Other policy relevant indicators” and “Number and rates of international mobile students (inbound and outbound)”

It is estimated that approximately 6,870 of the 39,504 UK students studying wholly overseas in 2018/19 were English domiciled initial entrants to higher education. This could add an estimated 1.1 percentage points to the initial participation rate.

Annexes

National Statistics reviews of participation measures

In November 2002, the Department for Education and Skills commissioned a National Statistics Quality Review of Higher Education participation statistics. The review was carried out independently by Professor Ramsden in accordance with Office for National Statistics guidelines and was designed to ensure rigour and transparency. The report on the review was published in July 2003 and is available on the Office for National Statistics (ONS) website, at:

[ARCHIVED CONTENT] National Statistics Quality Review of the Initial Entry Rate into Higher Education - The Department for Education (nationalarchives.gov.uk)

The report recommended that the HEIP be a measure of initial participation in higher education. In August 2004, the Department for Education and Skills commissioned a further review to assess the feasibility of disaggregating the HEIP by ethnicity, disability, social class and region. This review was also carried out by Professor Ramsden. The report on this review was published in August 2005 and is available at:

[ARCHIVED CONTENT] Participation in Higher Education: A Study to Determine Whether the Higher Education Initial Participation Rate Should be Disaggregated : The Department for Education (nationalarchives.gov.uk)

In December 2010, the UK Statistics Authority (UKSA) published an assessment of compliance with the Code of Practice for Official Statistics for Higher Education statistics in England and the UK. The Participation Rates in Higher Education Statistical First Release was covered as part of this assessment. The full assessment report is available on the UKSA website at:

Assessment Report 77 - Higher Education in England and the UK (statisticsauthority.gov.uk)