What Works for Underrepresented Groups? Identifying Effective Measures for Enhancing Access to Higher Education in Europe

Over the last three decades, policy-makers have developed numerous measures, policies, projects and programmes with the intention to increase the enrolment and participation of underrepresented groups, however, little is known about the ways in which such initiatives shape opportunities for potential students. Knowing which of these initiatives work and whether they are achieving their intended goals is of utmost importance for policy-makers across Europe. This paper aims to collect, document, scrutinize and critically analyze the current research literature which assesses the effectiveness of different public initiatives for higher education institutions to widen access for underrepresented groups. At the same time, the aim is to identify gaps and make recommendations for potential further research. Seventeen studies have been identified in this respect and based on the access measures they analyze, they can be categorized as follows: (1) outreach, counselling and mentoring of prospective students; (2) financial aid measures, and (3) preparatory courses and programmes. The findings show that there is little research or information about the actual outcomes of most measures and whether they increase access to higher education. We found a lack of adequate, reliable and consistently collected data about the policy instruments already put in practice. Since there is no excuse for the lack of effective action towards more equitable educational systems, more evidence-based approaches will be necessary to learn from these specific access measures and move forward towards more efficient equity policies. 1 Torotcoi, Simona (Central European University, Vienna, Austria, Torotcoi_Simona@phd.ceu.edu); 2 Gologan, Delia (Center for Educational Policy (CPEdu), Bucharest, Romania, delia.gologan@cpedu.ro); 3 Kurysheva, Anastasia (Utrecht University and University Medical Center Utrecht, The Netherlands, a.kurysheva@umcutrecht.nl); Working Papers on University Reform no. 32 Torotcoi, Gologan and Kurysheva: What Works for Underrepresented Groups? 2

Abstract: Over the last three decades, policy-makers have developed numerous measures, policies, projects and programmes with the intention to increase the enrolment and participation of underrepresented groups, however, little is known about the ways in which such initiatives shape opportunities for potential students. Knowing which of these initiatives work and whether they are achieving their intended goals is of utmost importance for policy-makers across Europe. This paper aims to collect, document, scrutinize and critically analyze the current research literature which assesses the effectiveness of different public initiatives for higher education institutions to widen access for underrepresented groups. At the same time, the aim is to identify gaps and make recommendations for potential further research. Seventeen studies have been identified in this respect and based on the access measures they analyze, they can be categorized as follows: (1) outreach, counselling and mentoring of prospective students; (2) financial aid measures, and (3) preparatory courses and programmes. The findings show that there is little research or information about the actual outcomes of most measures and whether they increase access to higher education. We found a lack of adequate, reliable and consistently collected data about the policy instruments already put in practice. Since there is no excuse for the lack of effective action towards more equitable educational systems, more evidence-based approaches will be necessary to learn from these specific access measures and move forward towards more efficient equity policies.

Equity considerations within the Bologna Process
The Bologna Process (and the subsequent European Higher Education Area -EHEA) represents the most significant and ambitious higher education agenda in Europe with an equity dimension (also known as the social dimension). The Sorbonne Declaration referred to the fact that 'students should be able to enter the academic world at any time in their professional life and from diverse backgrounds ' (1998, 2), and this was the beginning of the sequence of Despite increasing access to tertiary education, higher education systems remain highly stratified (Marginson 2016), gender imbalances still exist between different fields of study, and students with an immigrant background or whose parents do not have an academic background, have lower chances to achieve a tertiary education (European Commission/EACEA/Eurydice 2018).

The research question and problems of defining access, equity and effectiveness
This paper focuses specifically on measures and interventions that higher education institutions take to widen access for equity groups. 4 For this purpose, the paper will collect, document, In its narrowest sense, access can be defined as entry/admission to higher education (Prodan et al. 2015) while more generally, it can be defined as the ability of people from all backgrounds to access higher education on a reasonably equal basis (Usher and Medow 2010;Wang 2011). This definition is comprehensive in scope and implies that students of all backgrounds must not only be 'reasonably' able to take advantage of educational opportunities, but also must be adequately prepared and equipped to do so in order for the system to be considered 'accessible'. While in both cases, it is merely the starting point, the final goal of access policies is successful participation (Tonks and Farr 2003); for the purpose of this paper, access is defined in its narrowest sense.
There is no one single definition of equity in higher education policy, but several are prominent both in the literature and among practitioners. Salmi and Bassett (2014) understand it as equality, providing equal opportunities for access to and success in higher education in order to 'improve the chances of success of under-privileged youths' and even out the circumstances that are beyond their control (e.g. the financial resources of the family or educational attainment of the parents). Geven (2012) also associates equity with evening out (previous or existing) inequalities through the special allocation of resources that are translated into higher education policies, and policy instruments. Equity is also sometimes considered to be synonymous with 'widening participation' through utilizing tools for ensuring diversity (e.g. affirmative action). Interventions aimed at higher education equity address one or a combination of access, retention and persistence and successful transition to further studies or professional career. Holistic approaches tackle all potential sources of inequity such as socioeconomic, ethnic, gender-and disability-based, both at an individual and a system-level, through policy instruments that equalize economic, cultural and social capital within the education system (Geven, 2012).
When it comes to the effectiveness of various approaches to increase access to higher education, the authors opted to consider the extent to which a programme has reached the goal(s) that has been set initially, or whether it achieves the set expectations or the goal(s) that were intended or desired by stakeholders. Similarly, Cowan (1985) refers to effectiveness as the ratio of the actual outcome to the possible or the ideal outcome. The three most often used indicators for measuring the impact of higher education institutions activities on diversity, equity and inclusion refer to the number/share of students enrolled from less represented/disadvantaged backgrounds, the success stories of the people targeted through the measures, and the graduation rate of students from underrepresented/disadvantaged backgrounds (Claeys-Kulik, Jørgensen, and Stöber, 2019). For the purpose of this paper, we will be looking at the first set of indicators but keeping in mind the initial goals and intentions of the measure under consideration.

Literature gap and methodology
A significant number of countries and higher education institutions have started investing resources and taking on board initiatives aimed at widening access for disadvantaged or underrepresented groups -that we will refer to from now on as 'equity groups'. These initiatives include quota systems and preparatory programmes. However, little is known about whether such initiatives actually shape opportunities for potential students. Knowing which of these initiatives work and whether they are achieving the intended goals is of utmost importance for policy-makers. Given the fact that national-level programmes might provide different responses/reactions from higher education institutions (which have a certain level of autonomy in this matter) this paper addresses the relative effectiveness of access initiatives implemented by higher education institutions. , we searched publication titles and abstracts in these databases using the keyword 'underrepresented groups' in combination with in/or/and 'higher education', 'tertiary education' or 'universities'. In total, a number of 17 studies written in English have been included in the sample used for this paper, including two that are non-academic (i.e., conducted by organizations with policy-oriented and not research aims).

Access measures and their effects: What works for underrepresented groups?
Claeys-Kulik, Jørgensen, and Stöber (2019) put forward 12 most frequently used access measures used by universities, among which are: guidance, counselling and mentoring, accessible building and activities, assurance about non-discrimination, part-time study options and flexible courses, financial support, preparatory courses, recognition of prior learning, childcare on campus, positive action, housing support, quotas for students from certain groups/backgrounds, and general positive discrimination measures. Usher (2015) identifies several categories of measures universities can use to enhance access to higher education, including early intervention strategies designed to eliminate barriers in the educational pipeline prior to tertiary education. Our literature review indicated that access measures can be categorized in three main ways based on the problem they tackle and the phase they intervene in (i.e., during secondary education, during the transition to higher education or after the enrolment in the higher education system).
The first kind of measures is financial aid. It might primarily target students with low socio-economic backgrounds, but other equity groups as well. However, it is widely known that the principal dimensions of inequality overlap in many ways, for example, ethnic minorities are more likely to live in rural areas or peripheral neighbourhoods and, therefore, are more likely to be affected by poverty.  argues that nowadays financial aid policies are the most commonly used, often in combination with non-monetary aid policies.
These measures include tuition-free or partially subsidized higher education, needs-based grants, scholarships and bursaries, student loans, and a variety of funding formulas.
The other two measures are non-financial. The most widespread non-financial policy practices relate to reformed selection procedures, different forms of positive discrimination, and/or preferential admission programmes. Our literature study showed that, among this array of non-financial measures, most studies addressed outreach, counselling and mentoring, and preparatory courses for students to access higher education. its data and method, and the main finding with regards to effectiveness ('+' -effective, '-'negative, or '0' -no effects). To assess the impact on enrolment and retention of non-traditional students attending the summer school Institutional data +/-As Table 1 above shows, with some exceptions (i.e., the studies focusing on Aimhigher), the identified studies look at different measures targeting different equity groups of students, in different countries and higher education institutions. Therefore, the findings do not allow for a cross-comparison of the results. Thus, we abstain from making absolute conclusions and cause-effect inferences. Rather, the following section provides a more detailed analysis of the above studies as follows. The section starts by providing (1) a general description of the type of measures under consideration. It then provides (2) several examples of such measures by specifying the university accommodating the measure, the type of measure and its components, its target group and the intended goals of that specific measure. Last but not least, if offers (3) a synthesis of the evidence collected on the effectiveness and impact of the outreach, financial aid and preparatory programmes covered in these studies.

Outreach, counselling and mentoring of prospective students
Early interventions for eliminating barriers prior to access to higher education include outreach and bridging programmes or services like personal and professional counselling, mentoring and tutoring systems or general academic support. Counselling is applied in a wide range of areas such as education problems, psychological issues, career guidance, or disability guidance, and it can be used as a tool for reducing dropout (Wulz, Gasteriger and Ruland 2018).
Counselling can be provided by universities, private associations, NGOs, etc. and can support the increase in demand to access higher education (Wulz, Gasteriger and Ruland 2018). The counselling of prospective students can serve as a source of social capital for first-generation students (Pham and Keenan 2011) helping them to overcome a lack of social capital, assistance, and advice from their families. Career counselling and personal development programmes can also contribute in improving retention rates and results (outcomes). While in some countries, universities have an obligation to provide counselling, in othersespecially those with high demands to access higher education -many private associations or NGOs provide it. Specific counselling is offered to different equity groups based on their needs. Table 2 below exemplifies the different measures within this category implemented by universities. The identified studies focusing on this category of measures show that they have a relatively positive effect on access. However, one should consider that outreach initiatives aimed at increasing interest in science and technology are evaluated either based on whether participants liked the activities or not, or based on changes in the enrolment of a degree programme (Gumaelius et al. 2016). For example, the Stockholm University summer school and the Praktikum UPV (at Universitat Politècnica de València) provide activities for prospective students closely related to universities' everyday activities, including the opportunity for school students to perform small research projects with PhD students.
Praktikum UPV includes a one-week stay for secondary school students within university research groups for fostering engineering and scientific vocations. Both initiatives measured the number of participants who chose to enrol in a STEM programme at their university after the activities are completed. At Stockholm University, 12-14% of them chose to study a STEM field at Stockholm University, and 70% chose to do so at UPV (however, participants might choose to attend a STEM programme at another university, which is not reflected in the percentage but could be considered a success).
Pekkala-Kerr et al. (2015) examined the impact of an information intervention offered by student guidance counsellors to randomly chosen high schools in Finland about the financial returns of higher education. The information included labour market prospects associated with post-secondary programmes. The results show that on average, the information intervention did not affect the likelihood of being enrolled in a post-secondary programme or the type of programmes where students enrolled. Furthermore, the study shows that the application patterns among students graduating from the treatment and control school were indistinguishable from each other, nonetheless a third of the students reported that the intervention led them to update their beliefs regarding their returns to higher education.
In Germany, Ehlert et al. (2017) conducted a field experiment among high-school students from Berlin who had higher education intentions to find out whether information deficits lower the likelihood of college-eligible students from less-privileged families pursuing their college intentions. The findings show an increase in the application rates overall, including for students whose parents did not have an academic background or who had one college-educated parent, though there was no significant effect when both parents had an academic degree.
A large-scale clustered randomized experiment (Abbiati et al. 2018) involved over 9,000 high school seniors from 62 Italian schools, and provided students with personalized information on the costs, benefits and chances of success in higher education through three meetings. The results showed students enrolled less often in less remunerative fields of study in favour of postsecondary vocational programmes. This was especially the case among children of low-educated parents. The study showed that children of higher education graduates increased their participation in more rewarding university fields.
Looking at existing practices and needs in terms of guidance for inclusion in European universities, Cullen (2013) suggests that 'institutions that adopt peer and mentoring support programmes have lower rates of dropout' (cited in Wulz, Gasteiger and Ruland 2018). More specifically, they are successful in preventing dropouts. Wulz, Gasteiger and Ruland (2018) consider that counselling is an effective measure to widen participation in higher education, together with the provision of student facilities (e.g. housing, medical support, childcare). 74% of the beneficiaries of student union counselling perceived it as useful (study referred to by Wulz, Gasteiger and Ruland, 2018).
In England, the impact of Aimhigher (2004-2011) on widening participation in higher education for young people from underrepresented groups (pupils aged 12 to 16 including firstgeneration students) was studied by Doyle and Griffin (2012). They found that pre-entry mentoring (information advice and guidance) had positive effects on students' aspirationraising and access to higher education when combined with other measures, such as campus To conclude, the seven identified studies show that these types of outreach, counselling and mentoring do not have a strong potential by themselves but work best when combined with others. Personal and professional pre-entry counselling, mentoring tutoring systems, and academic support reach maximum results when complemented by a 'school culture that values and promotes going into tertiary education, that sets high expectations for participation in higher education and offers a curriculum that attracts and supports students in their postsecondary and career development' (Salmi and Bassett 2014).

Financial aid measures
Financial aid measures aim at easing the financial pressure for students already considered eligible for higher education. Generally speaking, they are either reimbursable or nonreimbursable. Reimbursable financial aid schemes (i.e., student loans) are sustainable forms of financial support requiring a lean administration setup, low subsidies, and an effective recovery system. They differ in terms of the source of capital, the type of expenses they cover, the eligibility rules, the guarantees they require, and the repayment scheme (e.g. direct loans; loans guaranteed by the Government that are shared-risk loans; income-contingent loans). Nonreimbursable schemes take the form of needs-based grants and scholarships that target students coming from families with lower income, certain ethnic minority groups or rural areas, women or students with disabilities. The alternative is often tuition fee waivers or subsidies for the traditionally underrepresented groups. All of them aim at eliminating the possibility that the low family income acts as a deterrent to access and success in higher education. Table 3 provides several examples of the shape and dimensions of financial aid measures embraced by several universities. Existing studies focusing on Europe show that the amount of aid had a direct effect on higher education enrolment and access. Fack and Grenet (2015) show that a fee-waiver (which amounted to 174 euros) in France had small positive effects on enrolment in the first year of undergraduate programmes, whereas the provision of 1,500 euros cash allowances to prospective undergraduate or graduate students increases their college enrolment rates by 5 to 7%. Baumgartner and Steiner (2006) evaluated the effectiveness of a student aid reform in Germany that substantially increased the amount received by eligible students to raise enrolment rates into tertiary education. The study found that the reform had a small positive, but statistically insignificant effect on enrolment rates, despite the 10% increase in the federal students' financial assistance scheme. Hatt et al. (2005) evaluated the Opportunity Bursary scheme (introduced in 2001), for students from low-income backgrounds. Here, higher education institutions were granted considerable discretion over the allocation of these awards. The research reports differences in the way two institutions -in the South-west of England -administered their bursary schemes, and the generated effects: bursary students from low-income backgrounds were more likely to continue beyond the year of entry than those students from low-income backgrounds who did not access the award. Moreover, it also revealed that the award of a bursary strengthened students' motivation to succeed, and it played an active role in underpinning student persistence and success. Hatt et al. (2005) argue that there are two possible explanations for this effect upon higher education continuation: the money might be useful and might strengthen the student's commitment to study. Lannert and Garaz (2014)  The evaluated evidence shows that financial aid can have positive effects on enrolment, but, depending on the target groups and the field of study, financial aid measures need to be complemented by other measures in order to foster enrolment.

Preparatory courses and programmes
Preparatory courses and programmes aim to even out previous or existing inequalities with regards to prospective students' previous education. This can refer the quality of previous education, a switch of the field of study, the language of instruction, or academic ability.
Preparatory courses target first-generation and non-traditional students, disadvantaged people and students who do not have any experience of academia and higher education. The general purpose of these measures is to enable the students to prepare efficiently to continue their studies towards higher education. The format for these courses and programmes differs from university to university (see Table 4). They range from intensive academic courses in areas students would like to pursue higher education studies to general academic preparation (academic writing, critical thinking and study skills), auditing courses, introductory semesters, language courses enabling students to pursue studies in English (or other) language, and general application process support and information. Completing the programme enables students to apply for university studies in various fields of study but also to gain first-hand experience and insights into a higher education programme. Finally, such measures could also contribute to enhancing students' familiarity with a higher education environment and help them overcome (academic and social) integration barriers while in universities. In Germany, prospective refugee students are treated like all international students, but during their application and enrolment, they receive special support in order to deal with their specific situation. Since the entrance criteria for the preparatory colleges include advanced knowledge of the German language, special classes prepare them for the entrance test in order to enrol in the preparatory courses. According to Berg (2018), these preparatory colleges and courses are important institutions for the internationalization of German higher education and the support of prospective refugee students.
The Roma Graduate Preparation Programme (formerly known as the Roma Access Programmes) at Central European University (CEU) is an initiative providing preparatory courses for Roma students. The programme aims to prepare Roma university graduates across Europethrough academic English, academic writing and tutoring in a field of choice -to compete for master's programmes either at CEU or abroad. Since 2004, when it was founded, the programme has enrolled 218 Roma students from all over Europe. Out of them, 215 graduated, and 141 (nearly 65%) were accepted into a master's programme at the end of their studies (Rostas 2017).

Pinheiro-Torres and Davies (2008) evaluate Brunel University's Urban Scholars
Programme, a 3-4year intervention aimed at increasing achievement and higher education aspirations among talented students from lower socio-economic backgrounds and disadvantaged areas, aged 12-16. The paper discusses the emerging findings after the first 2 years of the programme and suggests the biggest change occurs within scholars' confidence.
Updated, self-reported data shows that programme leavers had a higher education orientation of 88% after 3-years attendance, and almost half of them received offers of places in universities, and 83% of them started higher education studies. Looking at the same programme, Casey, Smith, and Koshy (2011) found that it 'had some success in steering students toward greater ambition and an awareness of the rewards of higher education' (p. 43).
They showed that 90% of the students who participated and completed the programme either met or exceeded the school targets compared with 22%, of the rest, of the gifted and talented group who met or exceeded their school targets. Walker (2010) investigates the academic performance of adults who entered the University of Glasgow via the Scottish Wider Access Programme (SWAP) between 1988-1993, including a pre-university summer school where non-traditional students (i.e., adults, with a socioeconomic disadvantage) received preparation and independent advice about opportunities to access higher education. The findings show that there were few differences between the students who attended the summer school and those who did not, however, both dropped-out at largely the same rate. Recent self-reported data shows that, since the first pilot of SWAP in 1987, more than 32,000 adults have taken the SWAP route to return to study. The programme helped many adults realize their potential and gave them the confidence to succeed in college or university. In general, preparatory courses and programmes prove promising in terms of efficiency, however, this depends on their specific components.

Conclusions
Overall, this research found few publicly available studies and little information about the actual outcomes of most measures. First of all, the identified studies cover a limited number of access measures available in Europe (i.e., none of the studies we identified looked at the effectiveness of the widespread online platformsmost of them targeting refugees). Secondly, the existing studies do not provide a comprehensive geographical overview across Europe.
With few exceptions, most of the identified research explores the context of the UK, Germany, France, Italy or Finland. This suggests that more evidence-based approaches will be necessary to effectively learn from these specific access measures.
The measurement of impact is hampered by the impossibility of isolating the effects of such policies in order to attribute cause and effect, as well as by the difficulty of generalizing particular results. The current promising but limited amount of research in the European context shows that the most effective way to tackle unequal access to higher education is through measures that combine financial assistance with help to overcome non-financial obstacles. As highlighted in the previous sections, many of the described policies benefit from governmental support, and success is also greater if there is cooperation between governments, higher education institutions or other education providers, NGOs, public authorities (in fields like health or welfare that complement the interventions in education), families and/or private companies. Programmes with a positive track record in terms of improving equity seem to be those combining financial support with non-financial aid offered to students (Salmi and Basset 2014) as well as those empowering students, setting high academic expectations and helping students and parents believe in themselves and in their educational success (Usher 2015). These latter ones tend to be more intrusive and require frequent contact with the targeted individuals e.g. academic support, mentoring programmes. Usher (2015) contends that making definitive statements about 'what works' is hindered by the impossibility to generalize particular results (i.e., issues of transferability in different institutional settings) and the tendency to re-define the term 'equity' when results become inconvenient (i.e., politically unwelcomed results). Similarly, Claeys-Kulik, Jørgensen and Stöber (2019) argue that the collection and use of data on equity are often subject to controversial discussions, and perspectives vary according to cultural, political and legal contexts across Europe.
Lack of adequate, reliable and consistently collected data is often used as an excuse for the lack of action towards more equitable systems, but it also hinders the option of evidencebased policy-making or of measuring the impact of the initiatives already put into practice.
Referring to specific measures targeting refugees, Streitwieser et al. (2019) argue that while sponsors described their plans for supporting refugees, they often did not share information on the amount of funding, the number of beneficiaries impacted, and other key data.
To conclude, this paper addressed a question about the relative effectiveness of various university-level access measures for underrepresented groups. The inference that can be made from this literature review is that all measures have a limited effect when implemented solely.
More profound effects can be achieved when the access measures are implemented in combination with each other, accounting also for the field of study and underrepresented group in focus.
The recommendations to policy-makers include ensuring prospective students have access to a combination of financial aid and non-financial measures. The measures that have already been developed and validated at other, but similar contexts could be put in practice first. Development of new measures, their constant evaluation and extensive research on their effectiveness should be encouraged both by the higher education institutions themselves and national governments.