Algorithmic bias may mean that universities that spend money on online advertising in the hope of diversifying their student recruitment are wasting their money, according to an academic.
Sandra Leaton Gray, associate professor in education at the UCL Institute of Education, said that the development of “personalised advertising” – for instance, the adverts presented to web users based on their Google search patterns, use of social media and geographic location – meant that some groups of teenagers were more likely to be exposed to information about university open days or degree courses than others.
In a pilot study, researchers from the IoE, the University of East Anglia and the University of Passau in Germany surveyed 150 young people aged between 13 and 19 from both a large UK city and suburban UK region “with rural characteristics”. This was matched against data on broadband and mobile telecommunications infrastructure.
Teenagers who lived in deprived areas were more likely to have poor broadband, something that Dr Leaton Gray said could have an impact on the likelihood of students using the internet for their studies.
In a separate study group, students were asked to show researchers the kind of adverts that they could see on their mobile phones. Participants who tended to use their phones for homework were shown adverts for degree courses, while others who used them for shopping or watching sport were more likely to be shown adverts for jewellery or betting shops.
This could be attributed to users being categorised by commercial databases, and directed towards different websites, Dr Leaton Gray said.
“A cocktail of algorithms privileges certain users and they get shown the good stuff. Everyone else is [classed as] a ‘waste user’,” she said. “The more you use your phone for educational purposes, the more likely it is your algorithm starts to show you positive resources that might help you along a productive life path.
“You will have taught your algorithm how to help you. If you use it for ‘lowest common denominator’ stuff, then you can expect it to hinder you.”
Dr Leaton Gray advised that, “if universities really want to target young people in deprived areas”, they were better off taking a creative approach, for instance linking adverts to “witty playlists” or creating “free, top quality A level revision course apps with links and dates for their open days, relevant transport search information and offers”.
“Advertisements for university courses and open days are particularly poorly targeted,” she said.
