Kalervo Gulson, Carlo Perrotta, Ben Williamson and Kevin Witzenberger reveal how Google Classroom works behind your screen…
As we write this article, our children are sitting in front of computers during yet another COVID-19 lockdown in Australia, doing their schoolwork through Google Classroom. The use of Classroom, an education platform, has been introduced as one of a range of education technology solutions to allow teaching and learning to continue when schools have been closed during the COVID-19 pandemic. Google Classroom is the 21st century update to the distance learning packs of the 20th century, or online radio schools in remote areas of the world.
While we all do our best, struggling at home to keep our children on track, and teachers struggling to create meaningful online interactions, behind the ubiquitous rectangles of the Google Classroom home page lies an infrastructure that connects our homes and our classrooms to a global technology company. Following mass school closures, by April 2020 Google reported 120 million users of G Suite across 250 countries and 54 languages; over 100 million active users of Classroom, doubling its reach from 50 million a month before; and a 60% share of the market in education computers in the US. A year later, Google added at least another 40 million Classroom users.[i]
While important work needs to be done on the experiences of teachers, students and parents who use Google Classroom, this article focuses on how Classroom works. Not what we see when we use it, but what is happening behind the screen, and how the use of platforms like Classroom might change what we understand as pedagogy. Our research looks at what are usually seen as very ‘boring technical things’, including Google’s privacy policy, its terms of service, and the documentation provided to hundreds of third-party developers who have built digital tools that can ‘integrate’ and share data with Classroom via Google’s application programming interface, or API. In a synopsis of this research, in what follows we cover three key areas: i) an introduction to platforms in education; ii) how Classroom shapes pedagogy; and, iii) a discussion of whether we should be worried about Google Classroom.
What are platforms in education?
If you attended university in the last 15 years or so, you will have experienced one kind of education platform, the Learning Management System, or LMS, like Moodle, Blackboard or Canvas. If you use Facebook or Twitter, you are also using a platform. Our starting point is that digital platforms have become central to interaction and participation in contemporary societies. Platforms are basically a set of services, products and tools that combine new forms of governance, technical elements, computation and economics. Platforms are kinds of infrastructures, made by people to organise social life.
The Google Classroom platform is not just a delivery mechanism for content, it is emerging as an infrastructure for pedagogy. By this we mean that it has features and properties that channel and organise the work teachers and students do. All sorts of tasks are now offloaded on to the platform, on to third-party integrations, and on to parents and guardians. Teachers often no longer have a say about what functionalities get integrated into their classrooms. A system administrator now makes that decision. Teachers are required to accept it. They become a cog in this infrastructure. Now, this may not be that much different from how teachers feel about endless policy changes, or data entry requirements. However, what might be different is that teachers are now connected through Classroom to a larger global infrastructure, much of which has little to do with education.
Platforms operate by creating frameworks for other tools to work together and for users to engage with the platform. Whether it’s Facebook, Twitter, Google, there are certain predetermined ways you can engage. Those are determined through a design process. An API is part of that process. APIs are formal collections of programming conventions and data restrictions that allow external applications to integrate into a platform and provide interoperability in distributed computing environments. That is, how different software packages can talk to each other across schools and systems. One feature of Google Classroom is the Google Classroom API. With the Google API, third-party entities (for example, small developers, large vendors and service providers) are enrolled in the platforms as a source of innovation, while ‘end-users’ and their interactions become sites of data extraction.
What do we mean by the pedagogy of Classroom?
Popular understandings of pedagogy emphasise its role as a framework for educational praxis, grounded in certain philosophical traditions, such as John Dewey’s moral philosophy. Pedagogies are therefore theoretical paradigms, at least they are taught that way in initial teacher education, that underpin practices and values in education. In more prosaic terms, pedagogies are what educators do as part of their jobs, often (but not always) under conditions of employed labour and in specific institutional settings. This more mundane, yet far-reaching, connotation of pedagogy is what we investigated in our research. Without doubt there are a number of benefits brought by using Classroom during the COVID-19 crisis. However, taking a longer-term perspective, the way the Classroom platform is structured takes away a degree of teacher pedagogical agency.
In Classroom, the API determines what counts as a legitimate user action. We call it a ‘data ontology.’ It determines what is actually ‘real’ in a particular context. But this ontology is actually arbitrary. Developers and corporations make those decisions in the interests of efficiency. For teachers, this means that you end up primarily doing what the platform allows. If certain teaching activities do not fit within that particular framework, if they do not fit the pedagogical framework that Classroom encourages, then, to be modified, they require additional work, technical skills, time, all things teachers may not have. The risk is that teachers just adapt and go with the flow of what Google allows rather than challenge it with something more pedagogically meaningful.
Also, there is a degree of platform literacy that is now required to teach and learn. A lot of pedagogy becomes about how to engage with the platform correctly. The ability to engage meaningfully with the platform increasingly cannot be separated from actual teaching and learning. That benefits Google first and foremost. It gets users used to the Google environment, so when they leave Classroom, they will keep engaging with the Google ecosystem. (Disclaimer: We wrote earlier versions of our research using Google Docs).
Or we could take the example of literacy, of learning how to read and write. Google Classroom now automates the process of originality checking,[ii] so it can be carried out by Google Docs itself. Teaching students how to engage appropriately with original material and explaining originality in a way that students can understand is a pedagogic process. But if that becomes automated, and it is just Classroom telling students what is original and what is not, it takes away the pedagogical dimension and becomes a matter of surveillance. Mistakes get flagged as a problem – another source of data – rather than being treated as a teachable moment, an issue of pedagogy.
But should we really worry about Google Classroom?
Schools deal with so much data everyday that it is hard to see why anyone should be worried about something like Classroom, that, especially in a time of lockdown, has provided a ready-to-use, readily available interface. Like other platforms, they make life easier and their efficiencies are undeniable. And it is understandable that teachers, and parents like us with kids using them, should just go along with them without questioning the problems.
However, the flip side is that in education systems around the world, we see a focus on measurement, accountability and high-stakes tests. The negative effects on teaching from these regimes of accountability, such as the narrowing of curricula, have been widely documented. We are interested in whether platforms like Classroom will continue a de-professionalisation of teaching, while adding to the burden accompanying the encroachment of administrative and accountability-related duties that have repeatedly undercut the educational dimension of teacher work.
Furthermore, the use of Google Classroom raises issues of data privacy and transparency in education, that are arising in other areas such as health and policing. Google is clear that any data they collect through Classroom is not being used to profile users or target them with advertisements. But Google has unprecedented scale, and is a company primarily set up on a search and advertising business model of extracting and using data from users.
For teachers and students, the moment users step out of Classroom, the traditional extractive model still applies – that is, even the data collected within the confines of Classroom is still used to refine Google’s tools. They use all the data collected from Google Docs, for example, to train the algorithms for the company’s AI models. Anyone who uses Google Docs is contributing to that process. If a teacher assigned a YouTube video to watch, that extractive model applies.
As Classroom becomes more embedded in schooling, we think there needs to be a broader political debate about regulation. Classroom is not really a fully closed environment. We call it a leaky pipe. There are gaps and holes, and current regulatory frameworks are unable to keep up with patching them all. This is hard to do in moments of crisis, but we suggest this framework should change to make Google more accountable as an educational actor that is shaping these dynamics in an active way.
Overall, Google Classroom allows teacher and students to undertake activities that seem to span what we understand as the key aspects of schooling – curriculum, pedagogy and assessment. But it seems that the Classroom platform ends up narrowing what we understand as education to a fairly narrow set of options to interact. It is not that online teaching itself narrows options but using Google Classroom may not be the best way to expand our online options.[iii]
To that end, more work needs to be done about how Classroom is used, and the ways it may or may not be changing teachers’ day-to-day practices. We would be interested in hearing from teachers if any of the above points ring true (or false).
Acknowledgement: The above is based on: Perrotta, C., Gulson, K. N., Williamson, B., & Witzenberger, K. (2021). Automation, APIs and the distributed labour of platform pedagogies in Google Classroom. Critical Studies in Education, 62(1), 97-113. [Please contact the authors if you would like a copy of this paper].
References:
Herold, B. (2020, December 17). How Google Classroom is changing teaching: Q&A with researcher Carlo Perrotta. Education Week. https://www.edweek.org/technology/how-google-classroom-is-changing-teaching-q-a-with-researcher-carlo-perotta/2020/12
Perrotta, C., Gulson, K. N., Williamson, B., & Witzenberger, K. (2021). Automation, APIs and the distributed labour of platform pedagogies in Google Classroom. Critical Studies in Education, 62(1), 97-113.
Kalervo N. Gulson is a Professor in the School of Education and Social Work, University of Sydney, Australia, and an Australian Research Council Future Fellow. His research investigates whether new knowledge, methods and technologies from the life and computing sciences, including Artificial Intelligence, will substantively alter the processes and practices of education policy.
Carlo Perrotta is Senior Lecturer in Digital Literacies in the Faculty of Education, Monash University. He is interested in the sociological and psychological ramifications of digital technology in education. His current research focuses on data-driven educational processes, digitally automated pedagogies and AI-driven education.
Ben Williamson is a Chancellor’s Fellow at the Edinburgh Futures Institute and the Centre for Research in Digital Education at the University of Edinburgh. His current research focuses on digital technologies and data infrastructures in higher education, and on the role of data science in the production of policy-relevant knowledge.
Kevin Witzenberger is a PhD candidate at the University of New South Wales in Sydney. He researches the automation of governance in education and the potential impact of artificial intelligence on education policy.
[i] For a wide-ranging set of offerings on online learning see: https://mitpress.mit.edu/books/manifesto-teaching-online
[ii] https://edu.google.com/products/originality/
[iii] https://blog.google/outreach-initiatives/education/classroom-roadmap/