Artificial intelligence in education: where are the educators?

Artificial intelligence (AI) has become a very popular topic in recent years, with ongoing questions as to how it can be used to enhance learning opportunities and outcomes. In order to find out more, Melissa Bond and Olaf Zawacki-Richter recently conducted a systematic review of the latest research.

Our review covers research articles published between 2007 and 2018. The focus is on AI in higher education, partly due to the sheer lack of research conducted within K-12 school settings. The findings do, however, still have relevance for the schooling system.

The articles in our review were grouped into four main categories:

1. profiling and prediction (e.g. student dropout and retention),
2. assessment and evaluation (e.g. automated grading),
3. adaptive systems and personalisation (e.g. using academic data to monitor and guide students), and
4. intelligent tutoring systems (e.g. diagnosing strengths and automated feedback).

Whilst the studies provided a range of positive examples of AI in education, they also showed AI to be less suitable for smaller classes, such as those found in schools. One reason is that a large number of writing samples is needed in order to train computers to assess accurately. Furthermore, studies found that not all writing genres were suitable for automated essay scoring, and therefore careful consideration would need to be given to the types of tasks where AI might be applied.

Our study also found that very few research articles on AI in education have been written by actual educators (8.9%), with the majority of authors coming from computer science and STEM backgrounds. This raises the question of how much reflection has occurred about appropriate pedagogical applications of AI.

Teachers are therefore strongly encouraged to engage in conversations around AI, including forging links with higher education institutions, so that future research into AI might be directed in a more pedagogically minded way, keeping students and ethical considerations at the forefront. Teachers are also encouraged to engage in practitioner-based research themselves, which can lead to greater reflection and insight, alongside improved outcomes for students.

Read the results of the study here.

---

Melissa Bond (@misc_nerd) is a PhD candidate and Research Associate, and Prof. Dr. Olaf Zawacki-Richter (@Zawacki_Richter) is a Professor of Educational Technology. Both work at Carl von Ossietzky Universität Oldenburg (Germany) and are members of the Center for Open Educational Research (COER).