Skip to Main Content

Using Generative Artificial Intelligence for Research

Writing Effective Text to Text Prompts

While generative AI tools may not be the best for straightforward information retrieval, they can be used as search assistants to support the research process to brainstorm, explore and refine ideas.

Text-based generative chatbots like ChatGPT, Bing or Bard can be used in the first stages of research to help you with topic exploration, narrowing down ideas and gathering background information.

Brainstorming & Narrowing Down Topics

Steps:

  1. Provide context for your information need in the prompt.
    Prompt: "I'm writing a research paper for [course] and I need help coming up with a topic. I'm interested in topics related to [subject]. Provide a list of 10 topic ideas related to that."

    Example: "I'm writing a research paper for Architectural History and I need help coming up with a topic. I'm interested in topics related to the Renaissance. Provide a list of 10 topic ideas related to that."
     
  2. Select of of the topics that interests you and ask for some sub-topics to narrow down the scope of the topic.
    Prompt: "I like the topic of gardens and landscapes in Renaissance architecture. Suggest some sub-topics to explore further."

    Note: If you don't see any that interest you, ask it to suggest more related topics until you find one that you want to use.

Writing Research Questions

Steps:

  1. Using one of the sub-topics of your choice, ask for potential research questions.
    Example: "Provide some research questions for the topic of Renaissance gardens and urban planning."

    Note: You can get feedback on research questions you have in mind by asking it to act as a research expert and provide constructive feedback on your question. 

    Example: "Act as a research expert. Provide some feedback on the following research question: How did the location and design of gardens reflect the social hierarchies and cultural values of the Renaissance? Provide some suggestions for improving the research question."

Identifying Keywords

Steps:

  1. Using the research question of your choice, ask for some keywords to use when searching library databases.
     
  2. You can use keywords and phrases it suggests in Page 1+ (Library search engine), library databases, or other search engines like Google Scholar.
     
  3. For further assistance, you can ask for some Boolean search strings to use in library databases.
    Prompt: "Construct a few Boolean search strings I can use when researching this topic in library databases."

Here are some tips to keep in mind when crafting effective prompts that LLMs can handle well for text-based tasks:

  • Use simple and natural language: Use straightforward language and a conversational tone.
  • Be clear and specific: Provide a clear, descriptive, and specific task or request you need the AI tool to complete. 
  • Context matters: Relevant contextual information can help the AI tool understand the task fully.
  • Add examples, rules, or constraints as needed: Provide some examples of what you expect, or rules or constraints for the AI tool to follow, which can lead to better results.
  • Test and refine: Try with different words and structures, and refine them based on the results you get.
  • Always verify: AI tools like ChatGPT can sometimes provide incorrect or made-up information. So it is essential to verify or fact-check the responses.

Connect with us