Sorry! De informatie die je zoekt, is enkel beschikbaar in het Engels.
This programme is saved in My Study Choice.
Something went wrong with processing the request.
Something went wrong with processing the request.

Design your course with ChatGPT or MS Copilot in 5 steps

Last updated on 12 August 2024
Reviewing a paper, gathering information or structuring notes - students already make use of ChatGPT and other AI. But how can you, as a teacher, use this tool as your assistant? In this teaching tip, you will discover how to use ChatGPT or Microsoft Copilot and how it can support you in your course design.

Step 1: first fully understand what generative AI is 
To take full advantage of ChatGPT or Microsoft Copilot, it is important to understand what the tool is and how it works. AI supports us in all sorts of things in everyday life. From the algorithms in Netflix, to automatically enhancing a photo on your smartphone. Generative AI is a separate branch that can create new content such as text, images, audio and video. ChatGPT and Microsoft Copilot are such generative AI systems in the form of a chatbot. It is an advanced language model trained on a dataset of internet information. If you give the chatbot an assignment, the tool answers your questions, writes an essay and can also support you in your teaching practice. Such an assignment is also called a prompt. The best way to get to know chatbots is to simply try it. 

Step 2: write clear and specific prompts
The quality of a chatbot's answer depends greatly on how you wrote your prompt. The more information and context you provide to the chatbot in your prompt, the better the AI tailors the answer to your situation. For example, look at the difference between these two prompts: 

  1. Make a biology assignment on DNA 
  2. Create a biology assignment for second-year university students on DNA, specifically on gene regulation. Give the assignment three different steps, with increasing difficulty. Begin with a knowledge question, then an analysis question, and end with an application question.  

At the first prompt, the chatbot does not know what level the assignment should be written for, and the answer may just be a secondary school level assignment. In addition, the topic is very general; the answer will reflect that as well. In the second example, in addition to the level, we have also added a learning objective: knowledge, analysis and application of knowledge on gene regulation. Curious about both answers? Try the prompts out for yourself in a chatbot.  

Step 3: give the chatbot a specific role 
In addition, you can give a specific role to the AI. By doing so, you nuance the task and the chatbot becomes your assistant and sparring partner. For example: 

Take the role of an educational expert and assistant in course design. I teach psychology at the university and need your help in creating the curriculum for the Social Psychology course for first-year students. Ask me any questions needed to establish the structure and content of the course, with clear learning objectives and challenging assignments.   

When you give this prompt, the chatbot will create a list of questions that will further help you build your teaching programme, and examples that fit the topic. If you answer the questions, the AI will process them back into a new answer. In addition to teaching assistant, you can give a chatbot any other role that is helpful to you in your work. For instance: sparring partner, proofreader, a student or a colleague from another subject. It also helps to share what your own role is.   

Step 4: ask follow-up questions 
Despite a clear and comprehensive prompt, the chatbot's answer may still disappoint you, or even make you question it. Realize that AI is not flawless: you usually don't get the best result all at once. For example, the AI might give you some good ideas in the first answer, and you might let the tool develop one of those ideas into a more concrete result via a follow-up question. Or there are certain features to the output that don't fit your practice, for which you ask for a modification again. This is how you get the best results after 'chatting' back and forth with the chatbot several times.  

Also, chatbots often make content errors. So always make sure to fact check. With a follow-up question you can get an addition or change in your result, and you can always point out that certain facts are incorrect - chatbots always apologize profusely.  

Tip: Use chatbots best in the morning when the tool's capacity is not yet overloaded by the American public. In the morning, the tool works fastest and makes the fewest errors.  

Step 5: have control questions asked 
In addition to asking questions yourself, you can also instruct a chatbot to ask you control questions at various points in the process. That way, the AI can assist you in not losing sight of key goals. You can add this in your first prompt, for example: 

Chatbot [Copilot or ChatGPT], as teaching assistant, help me come up with five different activating blended work formats for my course on Microeconomics for first-year university Economics students. In addition, ask me questions to check for constructive alignment between the activities, learning objectives and assessment.  

Source 
Website – ChatGPT: your teaching assisstant (only in Dutch); Created by the I-coaches of the ROC of Amsterdam-Flevoland. For questions, tips or knowledge sharing, please contact: j.devre@rocva.nl or s.heemskerk@rocva.nl 

Want to know more? 
See the extensive VU information on ChatGPT and Copilot on Canvas VU Tools for Education. Need more prompts? Check out ChatGPT and generative AI: 25 applications in teaching and assessment. Also give the experimental system of the VU Education Lab a try. Or come to one of the VU CTL shorts to learn more quickly.

Quick links

Research Research and Impact Support Portal University Library VU Press Office

Study

Education Study guide Canvas Student Desk

Featured

VUfonds VU Magazine Ad Valvas

About VU

About us Contact us Working at VU Amsterdam Faculties Divisions
Privacy Disclaimer Safety at VU Amsterdam Colofon Cookies Web archive

Copyright © 2024 - Vrije Universiteit Amsterdam