AI can be used in research manipulation to assist in tasks such as brainstorming, extracting information from research papers, scientific writing and editing, conducting peer reviews, and formatting citations.
Researchers use both general purpose chatbots (such as ChatGpt, Gemini, Copilot) and research and writing-only (such as Grammarly and Jenni). These uses of AI to help researchers improve research efficiency and reduce the burden on research, can improve research efficiency. However, similar to using AI, concerns about accuracy, privacy and plagiarism should be carefully considered when choosing to use these tools.
It is clear that AI holds a great promise in educational research, but we also need to be aware of the potential risks to data privacy, accuracy and transparency. We propose three important best practices to effectively use AI tools.
- Understand what the tool or method you have chosen will do. Understanding how machine learning and AI models work is key to implementing them effectively and identifying possible sources of bias or errors.
- Use manually generated comparison data to benchmark or review the results of AI-generated analysis. By verifying the results of AI-generated conclusions, you can gain more confidence in your research accuracy.
- Think systematically about AI risks Using the AI Risk Framework. Before using AI tools (both research and practice), teams should consider applying risk frameworks to identify and mitigate AI-related risks.
footnote
[1] There are also many applications of AI for pedagogy, but here we focus on AI as a research tool.
