The University of Oxford recently with its announcement of two new access initiatives aimed at significantly increasing the number of ¡°academically talented students from under-represented backgrounds¡± that it admits.

The first scheme, Opportunity Oxford, targets students from under-represented backgrounds who would have narrowly missed getting a place in previous years. It will offer them a bridging programme consisting of structured home-based study and two weeks of residential support. The second scheme, Foundation Oxford, builds on the innovative work of the , and provides a foundation programme of support for students from disadvantaged backgrounds with the academic potential to succeed.

For those of us involved in researching widening participation, Oxbridge admissions and social mobility, the day of the announcement saw a storm of excited Twitter activity. Many admissions tutors and commentators collectively patted each other on the back.

They were right about the huge significance of Oxford¡¯s to draw a quarter of its students from disadvantaged backgrounds by 2023. It is momentous that the university has finally formally acknowledged the importance of contextualising admissions and the fact that students from disadvantaged backgrounds have to achieve so much more to meet minimum entry requirement than peers from educationally advantaged backgrounds. This is a huge step forward towards a fairer admissions system.

É«ºÐÖ±²¥

However, amid the celebratory Twitterstorm there were also a few dissenting voices desperately attempting to bring attention to the enormous elephant in the room: when contextualising admissions, how will Oxford be defining disadvantage?

The university¡¯s press release emphasises the importance of the LMH Foundation Year in the decision to develop a similar campus-wide programme of support. Since 2016, LMH has offered 12 disadvantaged students a year ¡°the expert tuition necessary to enable them to progress on to undergraduate study either at Oxford or another top university¡±. One of the most innovative things about the programme is that it moved its admissions criteria away from a reliance on two group-level measures of disadvantage: (based on postcode) and (based on demographic classification). These have been roundly criticised for the huge number of false positive and negative indicators of disadvantage that they produce.

É«ºÐÖ±²¥

Instead, LMH focuses on individual-level contextual indicators, such as parental income, household income and receipt of free school meals. This is contextualisation at a level that is actually meaningful for the most disadvantaged potential students, fair at an individual level and, as LMH has shown, both effective and practicable in the Oxford context.

Unfortunately, Oxford has not explicitly outlined its position on the measures it will use to define the ¡°under-represented groups¡± that it is targeting. The university¡¯s official Twitter account receipt of free school meals as a potential indicator, but also indicated that POLAR and ACORN measures would form the core part of the decision-making process.

While the difference between group-level indicators and individual-level indicators seems like a technical detail, the evidence of the flaws in POLAR and ACORN is significant. Research has repeatedly shown the accuracy of these criteria is very low when linked with indicators such as free school meals. Widening participation practitioners repeatedly criticise them as being almost unworkably blunt tools, and my research with Oxford¡¯s admissions tutors revealed near complete disdain for them, with almost every tutor providing anecdotes about their failings.

Oxford¡¯s apparent decision to use them anyway for these two new initiatives, despite the criticism of them within the university, is undoubtedly related to the Office for Students¡¯ continued use of POLAR and ACORN in its own widening participation criteria. That means that, institutionally, British universities are stuck with them. However, developing these widening participation initiatives took institutional bravery and it will be a real lost opportunity if the university does not use its unique position to challenge the OfS¡¯ use of group-level indicators and drive change across the sector as a whole by pushing forward contextualisation at an individual level.

É«ºÐÖ±²¥

Once the excitement has died down, the real impact of these initiatives is likely to be shaped by how contextual indicators are defined. The devil will be in this detail.

James Robson is research fellow and lecturer in higher education at the University of Oxford.