Always check and follow your local guidelines about whether this use of generative AI is permitted. If it is permitted, you may need to acknowledge the use of AI
Generative AI tools that are not connected to the internet or research databases are likely to suggest fake sources. Ensure you use a tool that is connected to real source databases.
It is critical for you to develop and maintain your own critical and original perspective. Ensure that AI does not overshadow or replace your own analysis and synthesis of literature.
When using generative AI to find search terms, write down your ideas first to ensure you find your own voice and don't get influenced by AI. Once you generate a list of search terms, make sure you double check them to make sure all are relevant and related to your topic.
Ensure you do a close read of literature that AI finds so that you pick up on specific nuances and articles, and independently appraise the articles to decide if they are relevant to your review.
Literature searches with AI tools like Elicit or Semantic Scholar may only find articles that are open access and freely available online. Supplement searches by searching the Library Catalogue or academic databases to ensure you find as much relevant information as possible.
It is important that you critically engage deeply with key texts in your field.
Read articles yourself before getting an AI to summarise them. Making decisions on the relevance of an article yourself utilises your subject knowledge and ensures that all included articles are appropriate for your review.
and Semantic Scholar
can be used to find journal articles on a topic.
, and Research Rabbit
use citation chaining to find articles that have referenced your key papers.
You might consider using AI for proofreading your work in addition to looking for literature sources and summarising sources.
Generative AI might be helpful in expanding your perspectives on a topic. Consider these prompts:
"What are the key theories related to [your topic] in [your field]?"
"What are some potential gaps in the current research on [your topic]?"
"How might I synthesise these conflicting findings? [Study A findings] and [study B findings]?"