This is the second post on our series dedicated to the use of AI techniques in the three contexts explored in Patterns in Practice, this time we focus on the education sector.
Artificial Intelligence in education has existed as a research field since the 1980s. After periods of fluctuating interest, it regained attention in 2010. Decades of work on AI research and development in academic institutions, the growing influence of global tech corporations in the education sector, and the birth of data-driven policy and governance are factors that have contributed to the contemporary development of AI in education.
Today the presence of digital technologies in the classroom is stronger than ever before and the use of Learning Analytics (LA) is rapidly growing. Data mining and data analytics packages embedded in educational platforms are increasingly being used to track what students do in digital environments, evaluate their performances, predict their future outcomes, and evaluate how satisfied they are with the ‘student experience’ and learning provisions. Student data harvested from such platforms are also being used as a proxy to assess the performance of academic staff, programmes, and institutions.
Solutions powered by AI – tools to revolutionise education?
Philanthropists with power and financial resources, EdTech companies, and policymakers have promoted the idea that big data, learning analytics software, and adaptive learning systems have the potential to revolutionise education. Supporters of the use of AI tools have tended to emphasise that AI-based solutions can improve the efficacy of teaching and learning.
It has been argued that AI-based solutions help increase student achievement as these tools can help students to develop mastery in a number of knowledge areas, enhance student success and minimise inequalities. It has also been said that by identifying patterns in student data, educators can promptly identify students’ needs or weaknesses and act based on these data to improve their learning experience.
However, as will be explored in more detail later in this blog, while AI-based tools have often been promoted as avenues that lead to empowerment, flexibility and personalisation of education, some have contested these claims.
Ethical issues – an important part of the conversation
Ethical issues associated with the adoption of technologies are an important topic of discussion in the education context. For example, in 2021 JISC launched the National Centre for AI in Tertiary Education in the UK with the objective of addressing ethical challenges concerning the use of AI in higher education. Similarly, Surf, the equivalent to Jisc in the Netherlands, created the Special Interest Group AI in Education with the aim of gathering and expanding knowledge and insights about the ethical uses of artificial intelligence in education.
What could go wrong?
Despite the intense promotion of the big promises of AI for the education sector, a growing body of research (e.g. Jarke & Macgilchrist, 2021; Perrotta & Selwyn, 2020; Prinsloo, 2020; Williamson, 2017; Yu & Couldry, 2020) has reflected on the potential negative implications of adopting AI-based solutions without careful consideration, and raised a number of concerns about developments in the field.
Marketisation of education
Research suggests that the introduction of certain AI-based solutions to the education sector, has played an important role in furthering its marketisation. Some players in the EdTech industry have been seeking to exploit commercial opportunities in the education sector.
While their public discourse is that the use of data can bring enormous benefits to universities interested in using technology to measure their performances, they are also increasingly using data to further improve and refine their products and services. These dynamics have raised concerns about the monetisation of student data and, in particular, about how businesses have benefited from the use of these data to train their ML systems.
The creation of surveillant environments
The collection of data via educational platforms can have a positive impact if it is implemented in a consensual way. However, it is important to consider what might happen if the discourse of these platforms leads to the naturalisation of surveillance and the extraction of data in the educational process.
Corporations have tended to frame the surveillant environment created through these platforms as the foundation of a ‘new educational value’ and at the same time to treat the agency of teachers and learners as less important than meeting the demands of systems for collecting and processing data. This is concerning because it could potentially numb students’ self-development, obstruct their autonomy and violate their privacy.
Shaping the future of students, transforming the role of educators
Some issues have also been identified regarding the use of predictive analytics. While these techniques have the potential to bring benefits, Jarke and Macgilchist argue that they are transforming the role of students and teachers and could negatively influence the future of students.
Studies show that educators and other actors tend to modify their behaviour based on the data outputs produced by these systems. For example, teachers had considered the possibility of suggesting some students leave school and pursue a different path only because the system tagged them as ‘students at risk’.
Patterns in practice
As we can see, the growing use of AI-based solutions in education has raised critical questions regarding the implications, with researchers paying attention to issues such as surveillance and loss of privacy, bias and discrimination, and the marketisation of education.
In this context, it is also important to explore practitioners’ cultures in this sector. In the coming months, the Patterns in Practice team will be exploring how the beliefs, values and feelings of practitioners in the UK higher education sector (i.e. learning technologists, academics, senior managers) interact to shape how they engage with Learning Analytics. Stay tuned for further updates.