‘Big data’ makes an experiment out of education

Published by Dr Ben Williamson

Teachers in Scotland will be familiar with the demands of data collection. Recording and reporting information is part of the everyday work of schools.

The demands of data are set to grow dramatically in coming years. The Scottish Government’s improvement framework has made collecting ‘performance information’ a key priority. New standardized tests will generate national attainment data. Schools will also be expected to gather data about academic progress and well-being of pupils, teacher performance, the impact of parents on school improvement, and much more.

The demand for performance information in Scotland, however, is just the crest of a great wave of data collection washing across education worldwide.

‘Big data’ is a term usually applied to databases of information so large and complex they can only be analysed with advanced computing power and sophisticated algorithms. Companies like Facebook, Google and Amazon collect and ‘data mine’ masses of user data in order to make inferences about what products, information or content to display. Big data has become big business.

‘Stacks’ – photo credit beachmobjellies

Education is the next target of the big data industry.

A few years ago the chief executive of the educational big data company Knewton claimed education was the world’s most data-mineable sector. According to its marketing, Knewton would track every mouseclick and keystroke on its technology platform. The data would be mined and analysed using algorithms to infer and predict each individual’s probable future progress. Knewton would then ‘adapt’ and ‘personalise’ the learning experience around the individual.

The promise of ‘personalised learning’ through ‘adaptive learning platforms’ is driving the big data industry to invest in education with massive financial enthusiasm.

Facebook’s Mark Zuckerberg recently announced a huge charitable donation to personalised learning projects and technologies in the US.  These personalised learning technologies have the potential to transform schooling. At Silicon Valley’s AltSchool, a private school chain set up by an ex-Google executive with Zuckerberg funding, a team of in-house engineers is developing a showcase model of a school designed around personalised learning technologies.

AltSchool is the model twenty-first century school as imagined by Silicon Valley—and it aims to sell its model to other schools.

The high-tech culture of the tech sector is a long way from Scotland’s schools. However, early signs of big data use are already visible.

‘NetWork’ – photo credit cea+

With national standardized tests coming, commercial testing companies are selling products that allow primary schools to benchmark progress. Some tests taken digitally have adaptive capacity. As students take the tests they adapt to their progress. Computer-adaptive testing is prototypical of personalised learning.

ClassDojo will be familiar to many Scottish primary school teachers as well. Another Silicon Valley product, ClassDojo allows teachers to use their smartphone to award points for good behaviour in the classroom and facilitate positive classroom management. Its founders have considered selling behavioural trend reports back to school leaders and local authorities based on these classroom data.

ClassDojo has also expanded dramatically to become a social media site for schools. As well as ‘liking’ student behaviours, ClassDojo users can share classroom-created content, photos, and videos. It has also become the main communication platform between teachers and parents in many schools.

Just as Facebook is central to many people’s friendship networks, ClassDojo is becoming integral to school communities.

Global textbook companies are focusing on digital big data technologies too. The biggest publisher, Pearson, has restructured as a ‘digital-first’ company producing digital textbooks and courseware that can ‘learn’ from their own use, predict outcomes and personalise the learner’s experience.

Pearson has even partnered with IBM to develop ‘artificial intelligence tutors’.

During a lecture at this year’s British Science Festival, the headteacher of Wellington College claimed this kind of educational AI would replace teachers within a decade. Robot teachers, he said, will be able to listen to learners’ voices, read their faces and process their writing in order to provide an optimal level of challenge and inspiration.

Despite the hype, big data in education raises big concerns.

A key concern is privacy and data protection. The EdModo learning platform was recently hacked, with details about 77 million users posted on the ‘dark web’ for sale. Student privacy remains a critical issue.

Another concern is whether personalised learning platforms might automate the teaching profession. Though robots replacing teachers seems unlikely, automated systems that can personalise lessons may well feature in future classrooms. The role of the teacher will be changed by the demands of working alongside powerful algorithms.

Critics of AI note that data processing algorithms may contain inaccuracies. Personalised learning platforms have the potential to make significant decisions about learners’ progress through school. What if the algorithm is wrong? Teachers would never know since algorithms are protected by intellectual property rights.

‘Silicon Valley’ – photo credit JD Hancock

Efforts to bring big data into education reflect the Silicon Valley ideal that all complex problems can be solved with software. But as educators know, schools are complex places, located in unique contexts, which may be hard to process computationally.

Silicon Valley also tends to disrupt existing industries while seeking new sources of profit. Its business is to generate financial rewards from data.

Education is the next big thing for big data to disrupt. The value of big educational data for Knewton, ClassDojo, AltSchool, Pearson and the rest is not in advertising. It’s to create optimized products that can be sold to schools, and to reshape schools to fit the ideals of the tech sector. Students and teachers are guinea pigs in this new experiment.

With increasing demands for data in Scottish education, the conditions may be right for big data companies to close in. From Silicon Valley to Scotland and beyond, the tech sector is trying to capture public education for a profit.

About the author

This post was written by Dr Ben Williamson, Lecturer of Education within the Faculty of Social Sciences and member of the Faculty’s Education Studies research group. Ben’s research focuses on education policy and educational technology from a critical, sociological perspective. Ben also recently authored the acclaimed book Big Data in Education: The digital future of learning, policy and practice, which you can find out more about here.


Originally published in TES Magazine, 10 November 2017: https://www.tes.com/news/tes-magazine/tes-magazine/big-data-makes-experiment-out-education