We’ve launched a new research project, based at the UCL Institute of Education, which looks at the real life impact of social media algorithms on the everyday lives of young people aged 13-22. We are working with colleagues from the University of Passau in Germany, the University of East Anglia, and the University of Nottingham, as well as a range of industry partners.
We are starting to realise that social inequalities are being exaggerated by invisible commercial algorithms that include or exclude individuals from particular resources depending on what hardware they have purchased, where they live, and what resources they might have used in the past. Without equal access to adequate technological infrastructure, including high speed broadband at a regional level and local 3G/4G (and soon 5G) data networks, developing the right learning patterns and an appropriate sense of time management is going to be significantly easier for some pupils than others. Something as simple as having very few large businesses in the immediate neighbourhood can have a real impact on learning, as broadband and 3G/4G provision locally is more likely to be comparatively poor. This is where an unfavourable algorithm can kick in.
We are investigating whether such algorithms are steering young people to spend time on particular forms of technology consumption that represent mainstream, cheaply provided entertainment rather than more challenging or niche products that encourage higher order thinking (and our early evidence suggests that this is already happening, as some are only shown advertising promoting things like shopping, whereas others are shown educationally-based advertisements, for example for university courses or open days). This is important because personal algorithms become increasingly refined as they are used, consolidating any associated disadvantage. As we move to a world dominated by artificial intelligence, getting it right becomes even more important.
The aim is that this research will inform more effective, and inclusive, national infrastructure policies, recognising the dangers of biased algorithms. This is something that is a stated priority for many governments as well as some of the larger social media companies, as we moved to increased use of artificial intelligence systems. We are also developing a simple digital tool that allows young people to train their own algorithms in a more helpful way, supporting their educational and personal development generally.
There are three aspects to the research:
- Mapping the impact of technology infrastructure deficits on adolescent social deprivation in some of the UK’s poorest areas, as well as in certain regions of Upper Bavaria, working with colleagues from the University of Passau in Germany. We do this by analysing national telecommunications infrastructure data. The locations selected have a known problem with the provision of technology infrastructure, a phenomenon known as digital differentiation, and this is something that can happen in poor areas within towns and cities as well as in rural areas (surprisingly we have found that these cold spots even exist in places like Berlin and London).
- Exploring the real life experience of technology infrastructure use, in schools and at home. We are analysing the digital literacy amongst young people, and how they are affected by infrastructure provision, both in a positive and a negative sense. This is done via an online survey of young people between the ages of 13-22 at school and college, as well as group interviews. If you fit this age profile and you would like to join in, the English language version is here: https://www.surveymonkey.co.uk/r/KTG522R and the German language version is here: https://www.surveymonkey.de/r/YCMNBTW
- Using contemporary psychographic tools, similar to those used by marketing companies, to investigate the way young people prioritise different aspects of their lives online.
We are using our research findings to develop an evidence-based robust smartphone app that teenagers can load onto their phones. This will run in the background, nudging their social media algorithms to provide them with a more varied and healthy social media diet than they might otherwise manage to experience.
Contact Sandra Leaton Gray, Project Director, at s.leaton-gray AT ucl.ac.uk, if you would like more information, or to discuss collaboration or participation.
Sandra Leaton Gray 