Integrating ChatGPT into a written assignment plan

Laurent Béland, Smith Engineering This is an integration of GenAI into established assessments in MECH 479. This class has roughly 30 students.

Article Category

AI-Augmented Learning Activities

AI integration

Assignment

ChatGPT

This assignment integrates ChatGPT into an assignment that was done in previous years without the use of AI. Students use ChatGPT to refine their plans and technical writing for their 2000-word report.

It met the learning outcomes of the established critical assessment (report). My experience is that the distribution of critical assessment quality is roughly unchanged as compared to previous years. As far as I can tell, the work is taking the students about one-third/one-quarter of the time that it used to take them, with results of similar quality (i.e. some are excellent, some are good, some are bad).

For most students, this course is the first MECH course in which they are expected to search for and read peer-reviewed scientific articles. Most of them do not know how to use Google Scholar or Web of Knowledge. They spend 4-6 hours in-class writing (with my help and my TA’s).

I received very positive feedback from the students in MECH 479 about this activity. They mention that they felt like they were learning a useful skill, and that it was their only time in MECH getting guidance and practice on how to use GenAI in a professional/scientific setting.

Additional Information

I help the students come up with a good plan and good points to bring up in their critical assessment (doing a synthesis of 4-6 peer-reviewed papers and assessing the engineering potential of an emerging technology is not easy! They then put this high-level plan through ChatGPT, and refine it. I help them iterate and improve the output, both through correcting the AI output and providing it with better prompts. I also provided them with 15-20 minutes crash-course in how to get the maximum out of GenAI in terms of technical writing.

None.

This activity is best suited to an upper-year undergraduate or graduate technical course with a literature review, critical assessment, or technology-evaluation component. For this activity to work well, students need enough background knowledge to distinguish strong arguments and reliable sources from weak or superficial ones. In other words, it is most appropriate at a stage where students are expected to synthesize information across multiple peer-reviewed sources and already have some working knowledge of the field.

The assessment itself was a critical assessment report/essay on an emerging nanostructured materials technology. AI was encouraged as a tool to help students plan, structure, revise, and improve their scientific writing, but students remained responsible for finding, reading, synthesizing, and critically evaluating the peer-reviewed literature. The learning objectives addressed include locating and interpreting scientific articles, synthesizing evidence across multiple sources, critically assessing the engineering potential of an emerging technology, developing a coherent technical argument, and learning to use generative AI responsibly in a professional/scientific context.