Applying ChatGPT and AI-Powered Tools to Accelerate Evidence Review and Synthesis
Artificial intelligence (AI) tools have significantly enhanced the efficiency of evidence review and synthesis processes over the years. Since the emergence of tools like EPPI-Reviewer and Abstrackr in 2016, there has been a notable improvement in research productivity. However, the introduction of ChatGPT by OpenAI in late 2022 marked a significant milestone in the utilization of generative AI for research purposes, particularly in text-based data analysis.
Through a critical reflection approach, we explored and evaluated the capabilities of various generative AI tools such as ChatGPT, GPT for Google Sheets and Docs, Casper AI, and ChatPDF in facilitating the initial phases of a rapid evidence review. Our findings indicate that these tools play a crucial role in enhancing research productivity, particularly in tasks like formulating search strings and screening literature. Despite their efficacy, it is important to acknowledge that these tools may exhibit certain weaknesses, including generating inconsistent results and occasional errors.
Recommendations and Considerations
Based on our analysis, we advise researchers to exercise caution when utilizing generative AI technologies. It is essential to develop a comprehensive research strategy and review protocol to ensure effective monitoring and quality control throughout the process. By implementing robust protocols, researchers can mitigate the risks associated with the use of AI-powered tools and maximize the benefits they offer in terms of research efficiency and output quality.
Keywords:
Artificial intelligence, ChatGPT, evidence synthesis, large language models, rapid review, research methods, literature review, generative AI, ChatPDF, GPT for Sheets and Docs
For further reading, you can access the full study here.
