AI can support academic writing without replacing your thinking, arguments or writing voice. For a scoping review, you can use AI selectively for factual checking, testing the strength of arguments, identifying weaknesses and reviewing clarity while keeping the core writing and analytical decisions your own. 

How can you use AI ethically when writing a scoping review?

Using AI ethically in academic writing means deciding which parts of the process should remain yours and where AI can provide useful support. For this scoping review, the writing was fully manual, but AI was used heavily for specific purposes where it could add value.

I did not use AI much for planning the document because I already had an idea of how I wanted the paper to look. It became more useful during the later stages of writing.

It is very easy to get AI to write everything for you, particularly when English is not your first language or academic writing is not your strongest area. However, there is a sense of accomplishment in completing a first full draft and reading the arguments you have developed yourself.

I did not think AI would have produced the same arguments or made the same critical observations from the data, particularly in the discussion section. Those arguments extended from my own understanding and knowledge of the topic.

How can AI help you write the introduction of a research paper?

The introduction was more difficult to start because this was a review and I was not coming into it with extensive knowledge of the topic. I needed to decide how much detail to include and how technical the writing needed to be.

I knew I wanted to introduce the importance of AI in research and provide background information about how AI can produce themes. This would give the reader the context needed to understand how AI was being used in practice later in the paper. The background would then lead to the rationale for the review.

Most of the writing involved simply getting words onto the page. In the first draft, I was not concerned about grammatical mistakes, sentence structures or even citations. I needed to write without constantly editing and agonising over individual sentences or paragraphs.

AI was more useful when I needed to write about how different AI models work and how they can generate themes. There were many terms I wanted to confirm so that I was not misrepresenting them or using the wrong terminology. I prompted AI for summaries of areas I was uncertain about and checked my paragraphs for factual errors.

How can AI be used for the methods and results sections?

The methods and results sections were approached with a more structured process. In a scoping review, there is considerable structure around how these sections should look, so much of the work involved filling out the relevant sections.

The main challenge was locating information across Rayyan, ASReview, Elicit AI, Excel spreadsheets, Zotero and other sources.

The methods section was largely an explanation and justification of each research step. It was actually one of the easiest sections to write because it was mainly descriptive. The key focus was transparency and ensuring that everything was well supported, particularly because I was conducting a review using AI and this remains relatively new in the field.

The results section was also straightforward. In a scoping review, the results involve conveying what was contained in the Excel spreadsheets in an understandable way. It mainly took time to organise the information.

AI was not helpful here because of the mistakes it had already made during data extraction. It was likely to misunderstand the papers again, potentially wasting the work involved in tidying and checking the extracted data.

Why should you write your own discussion rather than ask AI to write it?

As I wrote the results section, I was already planning my discussion mentally. This happens in the background when you put your own research together.

Using AI to write the discussion can take this process away from you. It can become a slippery slope where the next section becomes more difficult to write because you have not developed the thinking yourself, making AI increasingly tempting.

The discussion also contains a core contribution of the paper. I needed original arguments that I could develop and synthesise from the findings of the scoping review.

My discussion was constructed around three interrelated arguments. Because this involved generating creative arguments, there were several ways they could go wrong, including fallacies, poor logic and under-substantiated points.

This was where AI was particularly useful. I uploaded my three arguments separately and asked AI to suggest improvements and identify weaknesses. I also asked whether my discussion was too critical or analytical for what a scoping review could conclude. I asked for further reading to help me improve.

How can AI help you check and edit academic writing?

AI was used heavily during the editing stage. My problem as a writer is that after looking at my writing for a long time, I can stop seeing its problems. At this point, another academic would normally be useful, but I was not ready for the paper to be reviewed yet.

I first uploaded the whole draft to ChatGPT and asked for general feedback. It provided positive feedback and areas for improvement.

I then started with the abstract, asking how well it reflected the paper and where the writing could be clearer. Some observations were useful, while others were inappropriate. For example, AI suggested pushing the critique further in the abstract, which went beyond what a scoping review should do.

I also asked AI to identify verbosity and then prompted it for specific areas. After redrafting the abstract, I asked for further feedback on the updated version.

I repeated this process for every section of the paper, checking whether AI identified the same issues I had identified myself and asking it to assess each section.

How should you challenge AI feedback on your academic writing?

I felt more comfortable rejecting and challenging AI feedback than I would normally feel when responding to another academic. When another person reviews your work, you may also want to show that you value their input. With AI, I could question its judgement directly.

Some of my prompts were also designed to explain how I wanted AI to read my writing. I briefed it on my intentions for the section, what I wanted it to convey, and then asked how well I had achieved that.

I did not simply accept every suggestion. Parts of the conversation involved questioning AI's judgement and deciding whether I was actually convinced by the feedback.

I also approached the editing process step by step. Each section received individual attention rather than asking AI to jump around the paper. This helped me focus on the specific purpose of each section without confusing myself or the AI.

What can you learn from using AI when writing academic research?

Writing can be enjoyable, and AI can easily take that experience away from you. It can be tempting to let AI take over when you are staring at a blank page and struggling to start.

There is a genuine achievement in writing something original and putting forward an argument that no one else has developed based on your own findings. Research is not simply about producing the final document. The process of developing your own thinking and writing can also be rewarding.

The way I incorporated AI into this writing allowed me to harness its capabilities while retaining my writing style and the intrinsic rewards of writing.

AI helped with factual checking, identifying weaknesses in arguments, assessing clarity and providing feedback. I kept control of the writing, the arguments and the decisions about which feedback to accept.

This allowed me to use AI as part of the writing process without outsourcing the thinking that made the research my own.