Skip to Main Content

Resources for Faculty

The Jung-Kellogg Learning Center works with faculty to support teaching and research. This guide highlights our faculty services.

Critical Thinking - Prompt Engineering

A prompt is the human-created input an AI model uses to generate an output.

Integrating prompt engineering techniques when using an AI chatbot tool (such as ChatGPT, Claude, Copilot, Gemeni, etc.) can improve outputs because well-crafted prompts yield more accurate, relevant, and useful AI-generated responses. As a bonus, writing effective prompts requires clarity, specificity, and an understanding of how AI processes language, which strengthens your critical thinking, analytical, and communication skills. It must be stated, however, that the discourse about the effectiveness of prompt engineering, like AI technology itself, is rapidly changing to the point that some question the future usefulness of prompt engineering.

When teaching prompt engineering skills, best practices suggest there are a few components to a prompt deemed somewhere between essential (TASK) and important but not mandatory (PERSONA, FORMAT, TONE).

Task: What you want the AI chatbot to do

Context: Give the chatbot enough information to constratin the endless possible responses)

Example(s): Including a relevant example greatly improves the quality of the output.

Useful, NOT mandatory:

  • Persona: Give the AI a role to embody for more specific results, but remember that it helps if the persona is someone well-known.
  • Format: Visualize what you want the chatbot to create; bullet points, charts, graphs, paragraphs, etc?  Give it instructions to limit the options.
  • Tone: Options could include casual, formal, pessimistic, etc. You could select a 'feeling,' prompt the chatbot for possible tone keywords you might choose from, or check out this table of tone descriptors at the bottom of the page.

Another example of a prompt engineering framework often referenced in LIS circles includes the acronym CLEAR (Concise, Logical, Explicit, Adaptive, Reflective), popularized by Lo, L. S. (2023). The CLEAR path: A framework for enhancing information literacy through prompt engineering. The Journal of Academic Librarianship, 49(4). https://doi.org/10.1016/j.acalib.2023.102720

When creating prompts, imagine that you're explaining a task to a smart assistant. Here is a step-by-step guide:

  • Provide full context: begin by setting the scene for your request.
  • Detail your specific needs: clearly articulate what you need from GenAI.
  • Specify the desired format: indicate the format in which you'd like the response.
  • Include relevant parameters: mention any specific requirements or constraints.

Key principles of effective prompt design:

  • Clarity: use simple, straightforward language.
  • Specificity: provide detailed instructions and parameters.
  • Context: offer background information and relevant details.
  • Structure: organize prompts logically and coherently.
  • Iteration: refine prompts based on initial outputs.

This section has been adapted from GenAI in Teaching and Learning Toolkit by Gwen Nguyen used under a CC-BY-NC 4.0 International License.

Better Prompting_crop.jpg

"Getting Started with prompting" from Writing Prompts, a section of AI in Education by University of Sydney used under a CC-BY-NC license.