Access and Widening Participation in College HE

Access and Widening Participation in College HE Briefing Paper 1: Widening Participation Measures and Indicators February 2015 Author: Dr Viv Wylie, MBE, Action on Access 1

Contents Introduction... 3 National and Institutional Measures for Widening Participation... 4 Life Cycle approach... 5 Target Groups... 6 Measurement... 7 POLAR3 and Low Participation Neighbourhoods (LPNs)... 8 Indices of Multiple Deprivation (IMD)... 9 Income Deprivation affecting Children Index (IDACI)... 10 National Statistics Socio-Economic Classification (NS-SEC) Four to Seven 11 Free School Meals... 12 First in Family... 12 Using the Measures: Implications for Colleges... 13 Reflective Questions... 14 2

Introduction This is the first in a series of briefings on aspects of college higher education (HE) sponsored by Association of Colleges (AoC) and Action on Access. The briefings describe key concepts and models of practice and realistic ways to implement widening participation policies and practices. The briefings will be useful for all colleges that receive HEFCE the student opportunity grant. The other guidance notes in this series of briefings will examine: Student success and retention; Effective and collaborative outreach; Evaluation. Measures and indicators for targeting students and evaluating performance in respect of widening participation are challenging for most institutions. The recent AoC survey of 63 further education (FE) colleges indicated that a number of institutions, while employing a diversity of measures, were generally more comfortable with monitoring than targeting and evaluation. Widening participation has a long history during which it has shifted criteria and changed practice. In the last few years monitoring and evaluation have become more central as governments have asked whether and how participation patterns are responding to widening participation investment. For example, since 2005 the Office for Fair Access (OFFA) has required an access statement, or evidence based plan for supporting and improving widening participation strategies across the student life cycle, where college higher education (HE) is directly funded and attracts fee income above the basic level. As a result measures have become smarter and the collection and analysis of data more critical. Widening participation students are those students who are under-represented in higher education because they are subject to various forms of disadvantage. National measures have been developed which aim to identify these students by proxy. This paper attempts to contextualise six key indicators currently in use and clarify what they measure and how they can be useful. 3

National and Institutional Measures for Widening Participation Widening participation to HE is a key strategy for the social equality agenda and a considerable amount of public finance has been invested in it. As a result governments and institutions alike wish to know how successful the strategy is and how effectively the money has been spent. National measures to define and identify the widening participation cohort have become a key toolkit in this endeavour. At the macro level Higher Education Statistics Agency (HESA) provides national widening participation performance indicators on which higher education institutions (HEIs) rely quite heavily to determine and benchmark their performance. FE colleges which have directly funded HE provision, however, are not represented in HESA statistics and so until recently have not been able to benchmark themselves against this range of national criteria. Since 2012 the Higher Education Funding Council for England (HEFCE), recognising the information gap in HE provision in FE colleges, has undertaken two annual exercises to return statistics for HE students registered at FE colleges in respect of widening participation and noncontinuation, 1 and in the second publication (2013) has also included employment indicators for students completing in 2010-11. This work currently does not include part time or mature students. However, all the indicators and the sector adjusted averages are as consistent as possible with HESA performance indicators (PIs). This means that the HEFCE published data can sit alongside the relevant HESA indicators and make available some valuable sector level comparisons. The 2013 publication shows that the sector adjusted average for widening participation in respect of all full-time undergraduates registered in FECs in 2010/11 was 16% and in HEIs was 10.6%, for non-continuation it was 21% (HEIs 16.1%) and for employment it was 92.8% (HEIs 94.2%). However, HESA data needs to be considered in context. Firstly the data applies to the previous year and arrives too late for prompt intervention. Secondly national statistics capture enrolments but not pre-entry students on outreach activities who may or may not be recruited to the providing institution 2 and who should be identified as belonging to the target group in order to evaluate the effectiveness of the college s outreach strategy. 1 Higher Education Indicators for Further Education Colleges, 2012 and 2013. 2 The tracking of outreach students is a national problem in recognition of which HEFCE has supported the extension of HEAT, the HE Access Tracker System. 4

Nevertheless FE colleges, like HEIs, are still accountable for their individual performance on widening participation, whether or not they submit access agreements to OFFA. Consequently FE colleges need to be confident that they have engaged the target cohort and are therefore able to support widening participation colleges successfully at all stages of their journey. The use of national measures which enable clarity and comparability is a key requisite in the demonstration of a robust institutional approach to widening participation. Life Cycle Approach Emerging concerns about under performance by students from poor backgrounds and ethnic minorities have made it imperative that widening participation strategies include support at all stages of the student s journey. OFFA s Director of Fair Access, Professor Les Ebdon, makes the point: An effective approach to access should not stop at the front door when a person enters higher education. Disadvantage can follow you like a shadow down the years, affecting the degree you end up with and your ensuing postgraduate study or search for a job. For access to be meaningful, there must be appropriate support for students as they progress through their studies and continue to employment or postgraduate study. 3 Widening participation is now regarded as encompassing the cycle from pre-entry support through to successful completion and entry into the next phase of work or education. 3 OFFA Press Release, July 7 2014. 5

Target Groups The primary challenge for any institution is to be able to identify the widening participation students both in terms of outreach groups and those enrolled. Before national measures were extensively used by practitioners widening participation students were identified as a range of groups under-represented in HE nationally Black, Asian and Minority Ethnic groups, people with disabilities, first in family to go to HE, people from low income backgrounds, travellers, vocational students, adult students, white working class males, Afro Caribbean males - among whom there were many individuals who did not meet the criterion of disadvantage. By 2006 a HEFCE review of Aimhigher reduced the criteria for widening participation to disability and belonging to National Statistics Socio-Economic Classification (NSSEC) categories four to seven. Care leavers were later added as a separate category. The current OFFA advice on who is defined as a widening participation students, that is, belonging to under-represented groups or OFFA countable groups, is as follows: groups that are currently under-represented in higher education at the national level rather than at a particular institution or course including (but not limited to) people from low income backgrounds people from lower socio-economic groups people from low participation neighbourhoods some ethnic groups or sub-groups people who have been in care disabled people 4 In view of this definition it is important to use indicators which will clearly identify these groups. 4 OFFA 2014, How to produce an Access Agreement for 2015/16, Questions and Answers on Drawing up you r2015/16 Access Agreement (Students covered by access agreements). 6

Measurement Broadly speaking measures of disadvantage are based on geographical areas (e.g. Participation of Local Areas (POLAR), Indices of Multiple Deprivation (IMD), Income Deprivation Affecting Children Index (IDACI) and on individual level characteristics (e.g. NSSEC, free school meals (FSM), HE Heritage) which identify socio-economic status. 5 These measures mostly support each other but they do frequently overlap and even when there is good correlation there are inevitably individuals who do not conform to statistical norms. Some indicators are more relevant to widening participation than others which measure different constituents of disadvantage. Some measures are more difficult to use in terms of the accessibility or validity of the relevant data and some are more appropriate to targeting possible cohorts than to monitoring individuals. The Supporting Professionalism in Admissions (SPA) report on Contextualised Admissions usefully summarises the problem of Performance Indicators as follows: data collected from applicants (such as the socio-economic status of the household) can be hard to verify; data relating to the circumstances of an individual s school/college experience can be hard to match to applications in a consistent way; and postcode data relating to an applicants neighbourhood/community may not reflect individual circumstances. 6 It is always good practice, therefore, to use a number of indicators in relation to each other as there is no one completely accurate indicator. The following looks at some key indicators for widening participation which could be used together to produce individual student profiles and institutional evaluations of performance. 5 See Aimhigher: A brief guide to data sources on HEA websitefor more detail. This is useful despite the fact that it only relates to targeting young students who would not otherwise progress to HE. 6 Contextualised Admissions: Examining the Evidence, Joanne Moore, Anna Mountford-Zimdars, Jo Wiggans 2013, p2. 7

POLAR3 and Low Participation Neighbourhoods Low participation neighbourhoods (LPN) is a key HESA PI 7 for young full-time students. From 2011/12 it has been based on POLAR3, a UK wide measure which identifies the relative rates of progression of young students to HE. It does not seek to link low participation with any other deprivation factor such as poor backgrounds. POLAR3 is based on 18 to 19-year-olds entering HE between 2005-6 and 2010-11. 8 It works by ranking the 2001 Census Area Statistic 9 wards according to young participation for these cohorts in five Quintiles, each Quintile containing 20% of the young population. The majority of those who fall into Quintile one (i.e. the lowest participation area) will constitute the widening participation cohort. This is a key measure since it is the only one which addresses the issue of young participation and correlates well with other measures of disadvantage. There is a small likelihood that not all young students living in Quintile one wards will be widening participation students as identified by economic disadvantage. 10 Conversely there is a fair chance that some widening participation students as defined by socio-economic disadvantage reside in other Quintiles. However, POLAR3/LPN is a robust national measure used by HESA and by HEFCE in respect of HE indicators for FE colleges. Indeed, OFFA lists those from LPNs as a target constituency per se and so can be used for targeting outreach on schools in Quintile one neighbourhoods and for evaluating institutional performance in terms of actual recruitment. 7 This measure is flexed slightly by HESA for mature and part-time students as some mature students have a degree and NSSEC/School information for part-time can be difficult to obtain. Here the widening participation cohort is defined as those who come from a low participating neighbourhood and do not have a first degree. 8 See HEFCE website for POLAR 3 data maps of young participation in the UK. http:www.hefce.ac.uk/whatwedo/wp/ourresearch/polar/mapofyoungparticipationareas/ 9 These wards, of which there are 8,850, were created for the 2001 census and are largely identical with political wards except that in some cases small wards have been merged. 10 Further Information on POLAR3: An Analysis of Geography, Disadvantage and Entrants to Higher Education, HEFCE, 2014. This helpful analysis demonstrates that most young students live in sub wards with a participation rate which is not very different from that of the ward. This is particularly the case with the most and least advantaged Quintiles. Overall it is estimated that only 1 in 14 young students lives in a sub ward with a participation rate markedly different from that of the ward. 8

Indices of Multiple Deprivation (IMD) This is an area based national measure of multiple deprivation which was introduced in 2000 and updated in 2007. Unlike POLAR3 which is a UK wide measure IMD relates only to England. It charts multiple deprivation at the level of the lower super output area (LSOA) which is smaller than a ward and contains an average of 1500 people and 650 households. 11 It is based on a combination of seven domains capturing different aspects of disadvantage: income deprivation; employment (ie those excluded from work by unemployment, ill health, family circumstances) ; health and disability; education, skills and training (this is divided into 2 sub-domains lack of attainment among children and young people and lack of skills qualifications among the adult population); barriers to housing and services; living environment (based on quality of housing stock and environmental quality); and crime. Altogether there are 38 indices across the seven domains plus a further two supplementary domains involving children (IDACI) and adults in relation to income deprivation. While it has a correlation with POLAR3 IMD is not such an accurate measure of educational disadvantage since it is based on a synthesis of measures rather than directly on progression to HE and it identifies the nature and severity of deprivation in an area rather than numbers of deprived people. It has a potential usefulness for targeting widening participation students in higher Quintile POLAR3 wards by identifying pockets of material disadvantage which could affect participation of some young students. However, because it addresses a range of disadvantage IMD is a blunter instrument than IDACI. 11 There are 32,844 LSOAs in England compared with 8,850 CAS wards. 9

Income Deprivation affecting Children Index (IDACI) This is one of the two supplementary measures within the Indices of Multiple Deprivation. It is useful because it focuses on low income in relation to children. IDACI assigns scores to LSOAs where children are living in poor households which are in receipt of benefits such as income-based job seekers allowance or income support. The most deprived LSOA has the highest score and a ranking of one. There is a strong correlation between IDACI scores (when aggregated to ward level) and POLAR3 across the regions but there can be overlap in all Quintiles. Some Quintile 5 wards, for example, have higher IDACI scores than in Quintile three which would suggest that low income is not inevitably a determinant of low participation in all cases. This is particularly so in London which, unlike the regions where the correlations are more consistent, has high levels of income deprivation affecting children but also high levels of participation. 12 Since OFFA identifies individuals from lower socio-economic groups as widening participation students IDACI can be used effectively with POLAR3 for targeting and for evaluation of performance. Both indicators are concerned with disadvantage as it affects children. The combination can help to confirm the coincidence of poor backgrounds and low participation, and to identify economic disadvantage in more advantaged areas with the caveats that low income may not inevitably be a deterrent to progression and that the combination will not improve the identification of mature widening participation students. 12 Further Information on POLAR3, p13. 10

National Statistics Socio-Economic Classification (NS-SEC) Four to Seven This indicator is based on parental occupation and is used by HESA as an indicator for young full time participation. It is based on seven categories of occupation, (an eighth category covers unemployment): 1. Higher managerial and professional occupations 2. Lower managerial and professional occupations 3. Intermediate occupations 4. Small employers and own account workers 5. Lower supervisory and technical occupations 6. Semi-routine occupations 7. Routine occupations 8. Long term Unemployed/Never worked NSSEC data can be related to areas by using information taken from the 2001 Census. This enables a correlation between POLAR3 and the distribution of NSSEC categories. The managerial and professional occupations are more prevalent in the more advantaged POLAR Quintiles and in general these show that the proportion of children participating in HE rises in NSSEC one to three households. The correlation between NSSEC information and POLAR3 holds across the regions, though London is somewhat out of kilter in that it has the highest participation rate nationally despite having a proportion of NSSEC one to three households only slightly above the national average. 13 It is worth remembering that there will always be families from NSSEC one to three in poor areas: Quintile one in the North East, the poorest region for participation, has 20% of households in the NSSEC one to three category. Although NSSEC has the benefit of relating to an individual s status it is a difficult indicator to use at the individual level as it is reliant on the way the question about parental occupation is asked and answered. UCAS found that a high proportion of applicants returned unknown in response to the question on parental occupation. 14 Using NSSEC data in relation to the Census information and in conjunction with POLAR3 can be helpful in targeting likely groups but it is probably easier to establish whether an individual is from a low income background by using Free School Meals as a proxy though this measure also has its problems. 13 Further Information on POLAR3, p16. 14 Over 25% applicants returned unknown to the question on parental occupation in 2007, Exploring the KnownUnknowns: A Study of UCAS Data, Sue Hatt and Neil Harrison, UWE Widening Participation Conference paper, 2009. 11

Free School Meals (FSM) In general there is a correlation between the proportion of students claiming FSM and rates of participation with 21% of all students claiming in Quintile one wards and 8% in Quintile five wards. In all regions except London, a high number of claimants are associated with lower participation rates. From 2012 this indicator was preferred by the Government as a proxy for socio-economic disadvantage. However, criticism has been levelled at FSM because it would appear to exclude some significantly deprived pupils and because over 300,000 of those eligible in the UK did not claim. 15 It also excludes mature students. However, for the purposes of determining whether a student is in the widening participation category FSM is a reliable indicator of low income and low potential for progression. If a student is not in receipt of FSM, however, they are not necessarily excluded from widening participation status and First in Family may be a useful alternative indicator. First in Family There is a close correlation between children having a graduate parent and young participation rates which makes it a useful indicator for widening participation. The average number of children with a graduate parent rises from 12% in Quintile one to 46% in Quintile five. The lack of overlap at ward level between Quintiles would suggest that this measure charts more accurately the areas of high and low participation than income based measures. 16 Used as an identifier for individuals it is, however, like all individual data, open to possible misunderstanding or abuse. 15 Take up of Free school Meals ISER Working Paper, 2012; the BIS publication on Performance Indicators 2013 acknowledges that there may be pupils who qualify but do not claim which would affect the measure. 16 Further Information on POLAR3, HEFCE, 2014, p18 para 58. 12

Using the Measures: Implications for Colleges The OFFA report on Outcomes of Access Agreements, Widening Participation Strategic Statement, and National Scholarship Programme Monitoring for 2012/13 17 draws attention to a range of measures which institutions develop for evaluation of the impact of their widening participation activities. These include identifying schools with low GCSE performance, collecting institutional data, using student tracking systems, commissioning research, and canvassing feedback from a range of stakeholders. However, the key measure is the number of widening participation students who are targeted and recruited and tracked through the institution. All evaluation, and indeed, all management of the institutional widening participation strategy, depends upon this. The choice of which measures to use therefore has significant implications for colleges. The HEFCE publication HE Outreach: Targeting Disadvantaged Students 18 although written in the days of Aimhigher has useful advice on targeting and evaluation. In targeting outreach constituencies a proportion of around 65% of students engaged in pre-entry activities who conform to widening participation criteria by area and economic status is acceptable. That said it is important to use a basket of indicators in order to ensure that as far as possible students are accurately assigned to the widening participation category. Which selection of indicators is, of course, a matter for each institution and may well depend upon local knowledge and institutional experience. Institutions in London may, for example, find that socio-economic factors are not as significant as elsewhere since participation rates are generally higher. The most indispensable indicator has to be POLAR3. It has the key advantage of linking wards to young participation (and to some extent to adult students) and so is superior to IMD. It also demonstrates compatibility with a range of other key indicators and is used by HESA to define LPNs. POLAR3 will indicate the catchment area of a school 19 for targeting purposes and will identify cohorts of Quintile one students who are defined by OFFA as belonging to the widening participation target groups. When it comes to individual students, however, whether attending outreach or entering the institution, additional indicators need to be brought into play in order to define and evaluate the college s widening participation performance across the student lifecycle. At this point when monitoring numbers of widening participation students recruited to the college it is important to use a range of measures in order to establish individual levels of disadvantage. 17 OFFA, 2014. Outcomes of Access Agreements, Widening Participation Strategic Statement, and National Scholarship Programme Monitoring for 2012/13, p26-7. 18 HEFCE, May 2007. 19 This can be refined through use of IMD and IDACI to locate schools in areas of high deprivation. 13

Not all widening participation qualifiers live in POLAR3 Quintile one areas. A students living in a POLAR3 Quintile two area but in a high scoring IDACI LSOA who is in receipt of free school meals, has no HE heritage or is in NSSEC categories four - seven would certainly count as a target student. Reflective Questions Measures and indicators often involve detailed work and so it is useful to step back from time to time and ask some questions about their use in the context of institutional strategy. These could include the following examples: 1. What indicators are we using and why? Are they the most effective indicators? 2. Do we have adequate access to information and resources for efficient data collection and analysis? 3. Do we distinguish between indicators for targeting potential widening participation cohorts and indicators for evaluating performance in terms of the recruitment and progress of individual students? 4. Do we have more problems identifying particular groups of widening participation students such as adults and/or part time? How do we tackle this? 5. Do we need to design any additional measures for evaluating particular activities strategies, cohorts? For example, might it be useful to factor in predicted GCSE grades of individual students? 6. Do we adequately factor widening participation statistics and their implications regarding institutional performance into the institutional strategic plan? 7. What policies and strategic groups drive forward monitoring and evaluation of widening participation within the institution, how are they reported and where is the accountability located? 14