One of the main parts of my current job involves teaching academic writing to second-year Japanese students who are studying English. The students are tasked with writing a five-paragraph argumentative position paper on a specified theme, and they have no choice about the topic. The essay involves presenting their argument clearly, while also refuting the arguments of the other side. For example, students might take a position on the use of nuclear power, either for or against, and then present their case, followed by a refutation of the opposing side’s arguments.

As you can imagine, this kind of essay writing is quite a tough ask of young adults working in a foreign language. Whilst over the years I’ve been teaching, there have been clear improvements in some areas of student writing, progress has been uneven. For example, the adoption of Google Suite and Google Docs has largely eliminated many surface-level grammatical errors, but the use of sources to support claims within essays remains a persistent challenge.

It is difficult to put a precise number on what constitutes enough citations, but during the first semester the mean number of references per essay was 3.8. When giving feedback, one of my most common comments related to missing citations. Despite clear instructional focus on the need for references, students often failed to use citations to support their claims or acknowledge the sources of their ideas. Conversations with students revealed that, although they understood why citations were necessary, the time required to research and back up points was, to say the least, somewhat arduous.

Since all students were writing essays on the same theme, I decided to try a different approach at the start of the second semester and asked students to share their sources before writing. As homework at the start of each topic unit, each student researched online sources related to the essay theme. They were required to briefly summarise the information they found in simple English.

In the following class, students discussed what they had found, what kind of information it was, whether general background, pro, or con, and whether they thought it was useful for their essay and why. After these discussions, students pasted their research into a shared class document that everyone could access. They were then free to use any of their classmates’ sources when writing their essays.

The results were quite impressive. All students increased the number of references they used compared to the first semester. By the third essay, the mean number of references had risen to an average of around 6.1 per essay. This was a notable increase. Importantly, this increase was not driven by a small number of students using far more references, but reflected a general rise across the whole class.

At the end of the semester, I asked students to reflect on the usefulness of the sharing exercise by writing a short reflective paragraph during the final class. Students highlighted several benefits. They felt the activity saved time, strengthened their arguments, and gave them access to a broader range of information than they would have found on their own. Many also mentioned that seeing their classmates’ perspectives before writing helped them think more deeply about the topic. The only real drawback noted was that some arguments gleaned from the shared document were repeated frequently in other classes, such as debate class.

I think the combination of quantitative data, specifically the increase in the number of references used, and qualitative student comments makes a fairly strong case for this being a successful intervention. That said, there are some limitations worth acknowledging. This was a small-scale action research project involving just 30 students, and other factors may have contributed to the improvement. These include conceptually easier topics or students gradually becoming more aware of academic writing requirements. It is also possible that my repeated reminders about reference numbers simply took until the second semester to sink in. Overall, though, I consider this a success.

As a final addendum, it is difficult to write about academic writing, or writing done outside the classroom, without mentioning the massive game changer that is AI. Last year, I felt fairly confident in spotting fabricated AI-generated references. However, recent iterations of ChatGPT have become increasingly sophisticated at producing accurate reference lists and in-text citations. I also noticed that some students were using AI to summarise articles for them, which was not quite what I had hoped for. Still, technology moves on, and this is clearly something I will need to think about more carefully in future iterations of the course.