CASE STUDY

The Matlab Cheat Sheet

Overview

I have started to experiment with ChatGPT. For example, when questioned about a topic within my expertise, the tool provided quite good responses for general overviews but weaker responses for specific questions. Furthermore, when asked for academic references, I found the output to be poor, with all references being false or made up.  

To enable students to get acquainted with ChatGPT, I have started to introduce it and allowed its use in an assignment. 

My approach

An assignment for the above-mentioned module requires students, working in pairs, to carry out theoretical exercises (experiments) during a set time period distributed across two lab sessions. 

The devices in the lab are operated with the MATLAB software. Students use this software to verify their results.  

All students must use MATLAB to some extent, although this aspect of the coursework is not graded. Still, many students worry about their ability to use MATLAB. 

To help alleviate their fears, I produced a list of instructions – the ‘cheat sheet’. Furthermore, they were allowed to use online documentation or ChatGPT for coding assistance, e.g. how to plot data). I ensured there was a clear distinction that this was permitted for the non-assessed coding only and not for the written part.

What benefits has AI had on my practice?

To help students: 

  • Become acquainted with such AI tools 
  • Start developing critical skills 
  • Use such tools appropriately with care and awareness 
  • Alleviate anxiety 

What have I learned? 

Surprisingly, the students didn’t use the ChatGPT alternative. Apparently, they preferred to use the instructions and explanations provided by the lecturer and/or textbooks to complete their coursework. 

My future plans

My plans for the future are quite extensive and focus on module MECH556 Autonomy Principles and Ethics on our MSc Autonomous Systems programme, level 7. 

This module introduces some essential but general aspects of autonomous systems. I want to help student to become mindful users of AI  tools and to be able to identify strengths and weaknesses of ChatGPT and in particular, to be critical of output from such systems.  

I will use ChatGPT (or a similar AI tool) to implement a flipped classroom dynamic during the module. The module content topics/concepts are general, and there is plenty of good information available online. Students will investigate the critical aspects of a specific element of autonomous systems (e.g., microcontrollers) through a ‘conversation’ with ChatGPT.  

During students’ interaction with ChatGPT, they will be able to ask what they think is important about the topic. So, they have the freedom to choose what to investigate. The amount of available information can be overwhelming, so the ‘Conversation with ChatGPT’ enables students to carry out more focused research. 

Then, during the lecture, I will organise round table discussions about the key points they learnt on e.g. microcontrollers and their role in autonomous systems. One of the reasons I’m using round-table discussions is the lack of such interactive methods in engineering courses and to develop students’ communication skills.   

During the round-table discussions, students should be able to determine the reliability of the information and reference sources. 

Potential benefits are to raise awareness of the flaws of existing tools such as ChatGPT, encourage appropriate and ethical use, and be more inclusive in my approaches as some students might feel more comfortable having a ‘conversation’ with ChatGPT than with the lecturer. Even if I encourage my students to participate, they are sometimes reluctant to ask. Gradually scaffolding interaction and communication by discussing outputs from ChatGPT will, it is hoped, help students become more confident and interactive. 


This case study has been captured by the Academic Development team and showcases current practice from BEng Mechanical Engineering, BEng Marine Technology and MEng Mechanical Engineering (CONT222 Engineering Mathematics and Control). We would like to thank Dr Pablo Borja (School of Engineering, Computing and Mathematics) for participating in this case study.

💻 Technologies: ChatGPT, MATLAB