Skip to Main Content
askus Ask us

Scholarly use of AI tools

How to use generative AI tools like ChatGPT, Bing Search, DALL-E 2, and others in academic settings ethically and in accordance with standards of academic integrity. How to reference content created by them or with their assistance.

1. Frameworks - Header




'Prompt Engineering' is the design of written instructions (prompts) for GenAI tools that are optimized to generate and/or manipulate those tools' outputs in order to meet the user's desired criteria as closely as possible. The ability to create such prompts is considered to be one of the essentials of AI literacy.

The following growing collection of tools (guides, frameworks, articles, etc.) on Prompt Design/Prompt Engineering consists of extracted pieces from a variety of resources. Please explore the full text of those resources for a deeper understanding of their respective approaches to prompt design and engineering.

← Click the arrows below the content to navigate between different Prompt-Engineering Frameworks →

CLEAR (2023): Prompt Engineering Framework for Academic Librarians

From the abstract (p.1):

[...] The framework encompasses five core principles—Concise, Logical, Explicit, Adaptive, and Reflective—that facilitate more effective AI-generated content evaluation and creation. [...] By integrating the CLEAR Framework into information literacy instruction, academic librarians can empower students with critical thinking skills for the ChatGPT era and adapt to the rapidly evolving AI landscape in higher education.

From the article (p.2):

Prompt engineering is an invaluable skill for academic librarians, as it enables them to utilize the maximum potential of AI language models for information literacy instruction. By understanding the principles of prompt engineering and mastering the art of formulating effective queries, librarians can ensure that AI-generated content is not only accurate and coherent but also pertinent to their students’ particular needs and learning objectives.

Lo, L. S. (2023). The CLEAR path: A framework for enhancing information literacy through prompt engineering. The Journal of Academic Librarianship, 49(4), 102720.

UNESCO (2023): Guidance for generative AI in education and research.

UNESCO provides the following definition of Prompt Engineering:

‘Prompt-engineering’refers to the processes and techniques for composing input to produce GenAI output that more closely resembles the user’s desired intent. 

Their specific recommendations include:

  • Use simple, clear and straightforward language that can be easily understood, avoiding complex or ambiguous wording.
  • Include examples to illustrate the desired response or format of generated completions.
  • Include context, which is crucial for generating relevant and meaningful completions.
  • Refine and iterate as necessary, experimenting with different variations.
  • Be ethical, avoiding prompts that may generate inappropriate, biased or harmful content.

From chapter 1.3: Prompt-engineering to generate desired outputs (p.10-11). In: Miao, Fengchun, & Holmes, W. (2023). Guidance for generative AI in education and research (UNESCO Digital Library, p. 44). UNESCO.

OpenAI (2023): Guide to Prompt Engineering: Six strategies for getting better results

From the introduction:

This guide shares strategies and tactics for getting better results from large language models (sometimes referred to as GPT models) like GPT-4. The methods described [...] can sometimes be deployed in combination for greater effect. We encourage experimentation to find the methods that work best for you.

The six strategies described in the guide include tactical approaches and example prompts:

1. Write clear instructions

2. Provide reference text

3. Split complex tasks into simpler subtasks

4. Give the model time to "think"

5. Use external tools

6. Test changes systematically

Note: In the introduction, OpenAI points out that some of the examples demonstrated currently work only with their most capable model, GPT-4.

OpenAI. (2023). Prompt Engineering. OpenAI Developer Platform: Guides.

The 'OpenAI Cookbook' provides an extensive collection of resources for further exploration of prompt design and engineering. These resources include growing collections of... 

Sanders, T., & Fishman, S. (2023, January 19). [Prompt Engineering]: Related resources from around the web. OpenAI Cookbook.

Harvard University (2023): Getting started with prompts for text-based Generative AI tools

Harvard University provides guidance for text-based GenAI and image generatorsTheir main recommendations on prompt design for text-based tools are: 

Be specific

  • Generic prompts lead to generic outcomes.
  • Specify details of what you would like to see in an output, e.g.:
  • What genre of text do you want?
  • What is the audience?
  • Length?
  • etc.
  • This will help the AI with generating more useful outputs, and reduce the likeliness of inaccurate responses. 

Role play: “Act as if…”

  • Using role-play prompts with AI can enhance the relevance and specificity of its responses.
  • For example, prompting the AI to "act as a climate scientist" when asking for analysis on climate change's agricultural effects can produce insights that are more relevant to the field's specific considerations.

Tell it how you want your output to be presented

  • Specifying the desired output format improves generative AI's results (e.g., code, stories, reports).
  • Use specific instructions like "Present this in the form of…" or "Create a [format] about / that contains…" to guide the AI's output.

Use “do” and “don’t”

  • Directly guiding AI with "do" and "don't" clarifies desired content and boundaries.
  • For example: "Act as a climate scientist. Analyze impacts on agriculture. Do include effects on crop yields and weather patterns. Don't include unrelated economic factors."

Consider tone and audience

  • Tailoring AI prompts with audience and tone considerations improves academic content creation.
  • Specifying "a detailed, accessible lecture on climate change impacts for undergraduate students" enhances relevance and engagement compared to a generic academic content request.
  • Adding specificity about the lecture's tone and intended audience ensures the AI-generated material meets educational objectives effectively.

Build on previous prompts

  • You don’t have to get everything into your first prompt.
  • Try starting with a basic question and adding to it over time.
  • Change the wording or tone or add more context and specificity to guide the AI toward the output you’re looking for.

Correct mistakes and give feedback

  • Chat with the AI as if it’s a colleague or teammate and you’re working on a project together.
  • Give feedback – tell it which parts of the output were useful and which parts could be improved.
  • If you notice it got something wrong, tell it so it can correct its mistake.

Ask it to create your prompts or what else it needs from you

  • Ask AI to help by creating a prompt for you!
  • Start with a basic idea of what you want and ask the AI to expand on it for you

Creative Commons License
This work by The University of Victoria Libraries is licensed under a Creative Commons Attribution 4.0 International License unless otherwise indicated when material has been used from other sources.