University’s unpublished 2017 report shows ‘structural or implicit biases in the interview process’, says campaigner
An unpublished University of Oxford internal report from 2017 has revealed “unexplained” bias against black, Asian and minority ethnic applicants in its undergraduate admissions process, some of which—but not all—the university says it has since addressed, Research Professional News can reveal.
The report based on data from 2013-2015 found that BAME applicants were 3 per cent less likely than their white counterparts to obtain an offer after taking into account background characteristics, prior attainment and test and interview scores.
In particular, the report noted a “significant unexplained gap in favour of white applicants” within the medical sciences, as well as mathematical, physical and life sciences, where BAME applicants were 7.1 per cent and 5.8 per less likely to receive an offer, respectively, than white applicants.
“This implies that within these divisions, shortlisted BAME applicants getting the same prior attainment, test and interview scores do not have the same chance of being offered a place,” said co-authors Samina Khan, head of undergraduate admissions, and her deputy at the time, Alison Matthews.
The revelation comes amid increased scrutiny on the underrepresentation of BAME people in academia as part of worldwide protests sparked by the murder of George Floyd, a black man who died following brutal restraint by a white police officer in Minneapolis. This week, Black Lives Matter protesters gathered in Oxford city centre to demand the removal of a statue of imperialist Cecil Rhodes.
“I am not sure how this data can be explained without invoking some sort of structural or implicit biases in the interview process, which is disadvantaging BAME candidates compared to their white counterparts,” said Ben Fernando, a PhD student at the university who obtained the report through a freedom of information.
In the note attached to the FOI response and dated 3 February, the university’s Information Compliance Team said that the “purpose of the report was to inform internal thinking and further work” on issues including improving the undergraduate admissions processes and the University’s Access and participation plan 2020-21 to 2024-25.
“Since the report was produced, Oxford’s proportion of UK undergraduate students from ethnic minority heritage has increased,” the team said, noting that the findings “do not necessarily represent the performance of the most recent BAME candidates applying to Oxford”.
When the report’s data were broken down by ethnic background, Chinese applicants and “other Asian background” applicants were found to have a “significantly lower probability of obtaining an offer of a place, after accounting for all available background information, relative to white applicants”.
Khan wrote: “In all cases where significant differences persist this implies that even BAME applicants getting the same prior attainment, test and interview scores do not have the same chance of being offered a place.”
The findings were based on admissions data from between 2013 and 2015 and drew on information from the university, UCAS and the Higher Education Statistics Agency.
The report took a detailed look at the progression of UK-based white and BAME students through the Oxford admissions process—including the proportion shortlisted for interview and offered a place at the university.
The data show that of the 6,568 BAME applicants to Oxford between 2013 and 2015, 52 per cent were shortlisted. Of those, 35 per cent were offered a place, with 89 per cent of those students going on to be accepted at the university.
Meanwhile, of the 26,081 white students who applied during that period, 66 per cent were shortlisted, with 42 per cent of those then offered a place. Some 92 per cent of those offered a place were eventually accepted to the university.
The paper found that “at each stage the white applicants are more successful—converting higher proportions of this cohort from shortlisted to offer and from offer to accepted—having a cumulative advantage, a larger proportion of a larger proportion progressing forward”.
This was true across all subject divisions, the report said, but the difference was “greater in some divisions than others, with medical sciences showing the greatest difference percentage-wise at all stages”.
A total of 1,445 BAME people applied to study medical sciences during that period, compared with 2,951 white people. Of the BAME applicants, 40 per cent were shortlisted, compared with 57 per cent of white applicants. At the offer stage, just 12 per cent of BAME applicants received an offer, compared with 28 per cent of white applicants. And while a quarter of white applicants were ultimately accepted, the figure was just 11 per cent for BAME applicants.
The report also looked at the proportion of BAME students with a widening participation flag who were “lost” during the admissions process—that is, they exceeded the requirements for a place at Oxford but did not get an offer. It found that the number was “often surprisingly high, when compared to the general level of BAME representation among applicants and offer holders”.
“For example, of the 68 widening participation applicants to medicine who were not offered a place, despite attaining well at A level, 50 per cent—34—were BAME.” Widening participation status is given to applicants who have “at least one socioeconomic and one school performance flag or a care flag”.
A university spokesperson said they have “acknowledged the gaps highlighted in the report”.
“To correct these gaps we implemented a number of initiatives including bespoke outreach targeted at students from ethnic minority backgrounds. The initiatives implemented are working and we don’t have an offer gap for UK student with African and Caribbean heritage but we still have a gap for UK Asian students.”
“We are working towards closing the Asian student gap and have set target to close it by 2021/22. We are fully committed to reviewing our progress and attracting and supporting students who are under-represented at Oxford and continue to do this in an evidenced-based approach.”