How should universities select students? #covid19 challenges have shone a spotlight on this issue both last academic year and this. In a new @CEPEO_UCL briefing note, I unpack key aspects of what goes into an application: https://repec-cepeo.ucl.ac.uk/cepeob/cepeobn8.pdf

All measures that British universities use to select students have the potential to bring bias into the process. But some have greater risks in this regard than others. All should be used with appropriate contextual data to minimise such distortion.

Prior academic attainment is predictive of performance once at university, but it’s even more predictive if you take into account students’ educational experiences e.g. private school applicants with the same academic profile on average do less in their undergraduate degree.

Aptitude tests should not be seen as a silver bullet to fix such bias. They are likely to be coachable. In the US, analyses have highlighted their biases. Meanwhile, pilots of their value in the UK have not generally found they add much over and above information in GCSE grades.

Personal statements and interviews also bring risks of inequality e.g. through ‘homophily’ between admissions tutors and interviewees perpetuating existing over-representation of higher SES individuals within universities.