Show me which groups you take part in on social media or which pages you like, and I’ll tell you how good you’ll do in school. Well, that could well be the interpretation of this study, but a couple of things:

• correlation doesn’t mean causal relation (!)
• being part of some social media groups means higher risk on performing less, but not necessarily.
• Neither vice versa!
• And no: it’s not a good idea to check the social media profiles of your students – privacy!!!

Now you know this, do read on. From the press release:

High school students’ membership in certain social media groups can be used to predict their academic performance, as demonstrated by Ivan Smirnov, junior research fellow at HSE’s Institute of Education. The analysis of school students’ membership in groups and communities was used to detect low-performing and high-performing students. https://www.aaai.org/ocs/index.php/ICWSM/ICWSM18/paper/viewFile/17798/17027 Teenagers’ performance at school can be evaluated based on their digital footprint and their activities in social media, according to the researcher. High achieving children usually subscribe to pages with scientific and cultural content. Lower achieving students are more interested in online humour and horoscopes. The researcher found the students’ academic performance to correlate with their social media interests on the basis of the Russian national longitudinal study ‘Trajectories in Education and Careers’. The sample included 4 400 students. Their knowledge was measured using PISA, a widely recognized international study. Social media activity was studied using VK, the most popular social network among young people in Russia. About 4,500 groups were selected. The study participants had 54 subscriptions on average. The PISA study, which is conducted by the Organisation for Economic Co-operation and Development (OECD), allows researchers to assess the level of reading and mathematical and natural science skills and knowledge among 15-year-olds. The list of social media subscriptions reflects the real interests of the school children. Obviously, it does not reflect the complete range of their interests since it is possible to enter the groups as an ‘invisible’ user, without leaving any digital footprint. Nevertheless, social media profiles are a good indicator. This source of information has great potential for educational studies. Personal pages, posts, photos and comments made by social media users provide a lot of information. Researchers can analyse a person’s lifestyle and psychological profile by using this digital footprint. For example, researchers found out that a person’s demographics (ethnic identity, gender and income level) can be forecasted following the analysis of their tweets, visual information from their profile, posts, photos of the neighbourhood etc. According to research, behavior on social media says a lot about an individual’s character and intelligence. Social media can also be used to ‘test’ academic performance. ‘We have created a simple model that predicts PISA results based on students’ subscriptions to certain groups,’ Ivan Smirnov explained. The results show that there is a strong academic, knowledge-related aspect of teenagers’ online preferences. Ivan Smirnov’s study revealed segregation among teenagers in terms of their interests and in correlation with academic performance. Lower-achieving students are usually interested in horoscopes and jokes. Higher-achieving students visit pages on science, technology, books, and films more often. One could assume that this influences their performance at school and general knowledge. But there is no such data. ‘We do not know whether subscriptions impact school performance, but I tend to think that they do not’, the researcher commented, ‘There are much stronger factors, including predisposition, family resources, level of the school, and so on.’ The author mentioned two effects. Firstly, ‘those who perform better tend to choose something educational, rather than entertaining, online.’ Sadly, the low achievers, who particularly ‘need support, are surrounded by horoscopes and jokes’. Secondly, if we take a closer look at the groups, it’s clear that ‘their content is not educational at all’. This means that the difference is due to ‘self-identity, and that it probably serves as a signal, since the groups in question are public,’ the researcher emphasized. The researcher compared the real PISA results and the grades calculated on the basis of digital footprint. The model was found to predict the teenagers’ achievements very accurately. ‘Its ability to correctly identify stronger and weaker students is 90% for mathematics, 92% for natural sciences, and 94% for reading,’ the author clarified. Two groups of students were investigated: those who don’t even have a ‘basic second level, which is necessary for survival in the contemporary world, according to the OECD, and those who achieve one of the two highest levels (fifth or sixth)’.

Abstract of the study:

The Programme for International Student Assessment (PISA) is an influential worldwide study that tests the skills and knowledge in mathematics, reading, and science of 15-year- old students. In this paper, we show that PISA scores of indi- vidual students can be predicted from their digital traces. We use data from the nationwide Russian panel study that tracks 4,400 participants of PISA and includes information about their activity on a popular social networking site. We build a simple model that predicts PISA scores based on students’ subscriptions to various public pages on the social network. The resulting model can successfully discriminate between low- and high-performing students (AUC = 0.9). We find that top-performing students are interested in pages related to sci- ence and art, while pages preferred by low-performing stu- dents typically concern humor and horoscopes. The differ- ence in academic performance between subscribers to such public pages could be equivalent to several years of formal schooling, indicating the presence of a strong digital divide. The ability to predict academic outcomes of students from their digital traces might unlock the potential of social media data for large-scale education research.