Developing Practice Material

It is well known that many new students are often unsure about the task of commenting on, or acting as formal reviewer of texts, written by their peers. Allowing students to practice on less successful texts from previous courses is not always a good idea, even if they are anonymised. Students may be uncomfortable with the idea that their own texts might be used in a similar way. Instead, selecting “VG-texts” (passed with distinction) can easily give students too narrow an idea of what a good essay should look like.

With the help of AI, teachers can easily generate several different texts on any subject, including references. Teachers can refine the texts in various ways, through follow-up instructions or by supplementing them themselves. Students - who are, of course, informed that the texts have been generated by AI - can then comment on and discuss the texts, online or in the classroom, and teachers can follow up on the students’ discussions. After a few rounds of initial exercises of this type, one should, of course, move on to discussing real texts.

In addition to training students in basic critical skills and academic discussions, it is also an excellent opportunity to discuss AI-related issues with students: academic integrity, independent work, etc.

Generative AI can quickly produce material that can be used in various types of scenario exercises. It can create a storyline for a role-play and even write dialogues for the different participants. It can formulate proposals for different cases if students are to work with case methodology.

Test it for yourself, and see how good the suggestions are for your specific subject area - the tools may not handle all areas equally well! Formulate your prompts carefully. Specify all important parameters that should be included, provide context and background, and specify the desired scope. Refine the result if necessary with additional prompts, or on your own. Always check the result to ensure it does not contain irrelevant or misleading details. If you request a script for a dialogue, remember that you can also specify the desired tone (humorous, angry, worried) and language level (formal spoken language, slang).

If you have old cases or role-plays that you want to develop, you can include them in the prompt and ask the tool to process them in the direction you want.

With the help of generative AI, you can easily produce material that can serve as a basis for discussions and various types of analytical exercises. For example, it is possible to generate, through well-chosen prompts, material that may contain errors, ambiguities, or weaknesses that students can practice quickly identifying, correcting, or improving. This could involve translating a text, checking programming code, formulating an application for a research project, providing an interpretation of a legal case, describing a historical event, carrying out calculations, analysing images or diagrams, and so on.

Such exercises may be best suited for beginner students who are still relatively inexperienced. After being trained to work on AI-generated material, they can more confidently tackle authentic material.

Giving students the opportunity to take diagnostic tests, which are not graded, can have many benefits. It supports students’ learning, can prepare them for the actual examination, and can also give teachers an indication of the students’ knowledge level before the examination.

Generative AI can formulate multiple-choice or free-text questions in any subject in just a few minutes. Specify in the prompt which knowledge area and approximate level it concerns, how many questions are desired, the number of answer options, whether the correct answer (or for free-text questions, a suggested correct answer) should be provided, as well as other necessary parameters. As always, it is the teacher’s responsibility to check both the proposed questions and answers to ensure that no errors or ambiguities are included when creating the diagnostic tests and making them available to students.