Teacher perceptions on Generative AI in programming classes

Click here for the full article

In September and October 2023, the teachers within Utrecht University’s Information and Computing Sciences faculty were asked to respond to a short survey based on the ITiCSE 2023 paper The Robots are Here: Navigating the Generative AI Revolution in Computing Education. The survey explored teachers’ opinions about allowing Generative AI in programming classes, ethical boundaries when using Generative AI, and perspectives on the importance of teaching or using Generative AI. The survey also captured information about the teachers’ experiences in teaching and with Generative AI.  

The survey garnered 41 complete responses, the majority of whom had over 10 years of teaching experience, had taught classes in the last year with over 100 students, and taught both Bachelor's and Master's courses that taught or required programming skills. Most teachers do not use Generative AI regularly. The responses show that Generative AI is mostly used for textual purposes, followed by images and code. 75% of respondents have never used Generative AI when working with code, and less than 13% of respondents regularly use Generative AI for code.

When teachers were asked if they had a policy in their course syllabus about the use of Generative AI, only 38% of respondents stated that they had a Generative AI policy in their syllabus. This number is surprisingly low when looking at the thought of respondents about the Generative AI policies of the university. 

More than half of respondents indicated that the Generative AI policies of the university and the faculty were unclear to them. The following infographics shows that, for teachers, these policies become more unclear in higher institutional levels. There should be at least a clear faculty wide policy were teachers can base their own course policies on. So that every course has the rules about the use of Generative AI incorporated in the course syllabus and that the level of allowance of AI tools is clear for both teachers and students.

The majority of teachers thought that ‘some’ or ‘many’ students were using Generative AI unethically. The vast majority of respondents believed that auto-generation or translation of complete solutions was unethical, but only a handful of respondents stated that having Generative AI find bugs, ‘explain’ how to solve the problem, and generate small portions of code was unethical.

Teachers didn’t believe that Generative AI will replace human teachers, but they do expect to use Generative AI in their teaching practices in the future. There is a majority of respondents that believes there should be restrictions on the use of Generative AI tools in course work. However, around half of the responds do not actively check on the use of Generative AI tools and can not detect code that is written by Generative AI. Teachers are are concerned that Generative AI tools will harm students’ learning and that they will become over reliant on these tools.

These are just a few of the observation made from the results of the Teacher survey, a more elaborate analysis can be found here