I’m facing an issue with my weekly reading reflections, as most submissions appear to be produced by AI. While in-person instructors have switched to having students handwrite their reflections during class, this option isn’t viable for my online asynchronous students this term. I’m looking for innovative alternatives that ensure submissions are the students’ own work. What methods or assignment designs can I adopt to uphold academic integrity in this digital format?
One approach I’ve found effective is designing assignments that require context-based and reflective responses tailored to the course content. For instance, including questions that call on recent discussion points or specifics from online sessions forces students to connect their personal insights with the material. This method prevents a straightforward generic answer often produced by AI. Another effective method is encouraging drafts followed by peer-review or self-assessment processes to track the evolution of thought and ensure that the work is indeed original.
had a go with assignments that require rough drafts and personal journaling before submitting final work. the step-by-step process makes it harder for ai to be the main driver and ties work to personal experiences, keeping things more genuine
Hey everyone, I’m really intrigued by this topic! I was thinking that maybe another option could be to have students engage in a kind of ongoing dialogue throughout the course. For example, you could set up assignments where each week they have to respond not only to the reading but also interact with feedback from their peers or even update a running commentary as the course progresses. This could make it a bit of a creative project rather than a one-off submission, which might help differentiate their individual thought processes. I wonder if mixing in some current happenings or even asking them to relate course themes to something personal from their week could spark more unique answers. Does anyone have experience with dynamic, multi-part assignments that evolve over time? I’m really curious if this approach could add an extra layer of authenticity and make it harder for AI to chip in. What are your thoughts?