How to use GUIDE-LLM

Purpose. The GUIDE-LLM checklist offers a clear and consistent way to report how large language models (LLMs) are used in behavioral and social science research. It should help ensure transparency, reproducibility, and ethical accountability. Researchers can complete the checklist as they design, implement, and document their studies.

How to use the checklist.

  • Provide short, specific answers for each item.
  • You may reference sections or appendices of your manuscript instead of repeating text.
  • If an item does not apply, write “N/A” and, if useful, add a brief explanation.
  • If your study uses more than one LLM, fill out the relevant sections separately for each model.

When to complete the checklist.

  • We recommend complete the checklist during (rather than at the end) the research process. This helps ensure accurate reporting of details such as model versions, prompts, and configurations, which can otherwise be overlooked later. It should also help you reflect on specific decisions like which LLM to choose, and why, etc. to make an informed, deliberate choice and the defend this choice against readers and reviewers.
  • Ideally, all items are reported. However, this is not a prescriptive requirement. Research designs and LLM use cases differ, and some items may not apply. In those cases, write “N/A” and, if helpful, explain briefly why the item does not fit your study.

Submitting the checklist.
When your manuscript is ready, include the completed checklist as supplementary material so that reviewers and readers can clearly see how LLMs were used.

📚 Example from published research

Below are a couple of examples showing how the GUIDE-LLM checklist is used in research
(Note: the list is still under construction).